Euphorbia in Southern Africa: Volume 1 3030494012, 9783030494018

Citation preview

Peter V. Bruyns

Euphorbia in Southern Africa

Volume 1

Euphorbia in Southern Africa

Peter V. Bruyns

Euphorbia in Southern Africa Volume 1

Peter V. Bruyns University of Cape Town Rondebosch, South Africa

ISBN 978-3-030-49401-8    ISBN 978-3-030-49402-5 (eBook) https://doi.org/10.1007/978-3-030-49402-5 © Springer Nature Switzerland AG 2022 This work is subject to copyright. The Author has marked certain illustrations in the manuscript in the figure legends as follows: ‘© PVB’. whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the author give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Contents

Volume 1 1 Introduction���������������������������������������������������������������������������������������������������������������    1 1.1 Historical Sketch�����������������������������������������������������������������������������������������������    2 1.2 Classification of Euphorbia�������������������������������������������������������������������������������    5 1.2.1 Relationships among and within the Subgenera �����������������������������������    8 1.2.2 Species Concepts in Euphorbia�������������������������������������������������������������    8 1.3 Morphology of Euphorbia (mainly in Southern Africa)�����������������������������������   10 1.3.1 Rootstock�����������������������������������������������������������������������������������������������   12 1.3.2 Shoots�����������������������������������������������������������������������������������������������������   14 1.3.3 Spines�����������������������������������������������������������������������������������������������������   19 1.3.4 Leaves ���������������������������������������������������������������������������������������������������   22 1.3.5 Stipules���������������������������������������������������������������������������������������������������   23 1.3.6 The Organization of Flowers�����������������������������������������������������������������   24 1.3.7 Fruit, Seed and Seedlings�����������������������������������������������������������������������   28 1.4 Chromosome Numbers �������������������������������������������������������������������������������������   35 1.5 Pollination Biology �������������������������������������������������������������������������������������������   35 1.5.1 Flowering Times �����������������������������������������������������������������������������������   35 1.5.2 The Mechanics of Pollination ���������������������������������������������������������������   37 1.5.3 Pollinators ���������������������������������������������������������������������������������������������   37 1.6 Biogeography and Ecology�������������������������������������������������������������������������������   37 1.6.1 Aspects of Overall Distribution�������������������������������������������������������������   37 1.6.2 Aspects of the Distribution of Euphorbia in Southern Africa���������������   43 1.6.3 Habitat and Ecology�������������������������������������������������������������������������������   45 1.7 Cultivation���������������������������������������������������������������������������������������������������������   47 1.7.1 Propagation �������������������������������������������������������������������������������������������   48 1.7.2 Diseases, Parasites and Predation ���������������������������������������������������������   49 1.8 Uses�������������������������������������������������������������������������������������������������������������������   51 1.9 Systematic account���������������������������������������������������������������������������������������������   55 1.10 Key to the subgenera of Euphorbia in Southern Africa�������������������������������������   55 2 Euphorbia subg. Athymalus �������������������������������������������������������������������������������������   57 2.1 Sect. Anthacanthae���������������������������������������������������������������������������������������������   58 2.1.1 Subsect. Dactylanthes ���������������������������������������������������������������������������   60 2.1.2 Subsect. Florispinae�������������������������������������������������������������������������������   91 2.1.3 Subsect. Medusea�����������������������������������������������������������������������������������  234 2.1.4 Subsect. Platycephalae���������������������������������������������������������������������������  409 2.1.5 Subsect. Pseudeuphorbium �������������������������������������������������������������������  413 2.2 Sect. Crotonoides�����������������������������������������������������������������������������������������������  460 2.3 Sect. Lyciopsis���������������������������������������������������������������������������������������������������  462

v

New synonyms  published in this work: Euphorbia grandialata R.A. Dyer is reduced to synonymy under E. grandicornis A. Blanc, E. halipedicola L.C. Leach under E. bougheyi L.C. Leach, E. decliviticola L.C. Leach under E. graniticola L.C. Leach, E. stenocaulis Bruyns under E. plenispina S. Carter. Lectotypes are designated here for Euphorbia benguelensis Pax, E. caerulescens Haw., E. cucumerina Willd., E. enopla Boiss., E. fleckii Pax, E. genistoides var. leiocarpa Boiss., E. genistoides var. major Boiss., E. grandicornis K.I. Goebel, E. grandicornis J.E. Weiss, E. involucrata var. megastegia Boiss., E. latimammillaris Croizat, E. melanosticta E. Mey. ex Boiss., E. nodosa N.E. Br., E. platymammillaris Croizat, E. polygonata G. Lodd., E. proteifolia Boiss., E. trichadenia var. gibbsiae N.E. Br. and Tithymalus zeyheri Klotzsch & Garcke. A type is also designated for the genus Tirucalia Raf.

vii

1

Introduction

Euphorbia L., with around 1840 species, is one of the largest genera of flowering plants and is by far the largest genus in the Euphorbiaceae. Common to all species of Euphorbia is the copious milky sap (transported to all organs of the plant by unarticulated laticifers), the possession of an unusual but typical ‘inflorescence’, the cyathium and the absence of a perianth in the male florets. Along with its almost cosmopolitan distribution comes a bewildering variety of vegetative forms. Many of the species are small to medium-sized herbs. In stark contrast to these are some spectacular species in Africa, the Arabian Peninsula and Peninsular India, which may form enormous succulent trees reaching 30 m tall (as in E. ampliphylla, E. cussonioides and E. ingens). Even among the succulent species there is remarkable variation, from these huge trees to dwarf spherical or globose plants that often do not exceed 50 mm tall at maturity (such as E. gymnocalycioides and E. obesa) or small geophytes with a rosette of leaves flat on the ground (such as E. acaulis, E. rubella and E. tuberosa) and even tiny annuals not exceeding 10 × 100 mm (as in E. prostrata). The larger members form a conspicuous and characteristic, sometimes dominant component of the vegetation of many of the tropical, semiarid areas of Africa, the Arabian Peninsula and Peninsular India. Outside the tropics, ­succulent species of Euphorbia are an important component in semi-arid, temperate areas receiving winter-rainfall in southern Africa (in the Greater Cape Flora) and in north-west Africa in some parts of coastal Morocco and on the Canary Islands. The last revision of Euphorbia for southern Africa was contained in N.E.  Brown’s accounts in Flora of Tropical Africa Vol. 5 (2) and Flora Capensis Vol. 5 (2) (Brown 1911– 12; 1915). The succulent species received further detailed attention from R.A. Dyer (1931) for the Eastern Cape and for the region as a whole in The succulent Euphorbieae (Southern Africa) by Alain C. White, R. Allen Dyer and Boyd L. Sloane, a sumptuous pair of tomes which appeared in 1941. This monograph was largely based on the field-experience of Dyer, who travelled widely to collect and familiarise himself with the species in habitat. Subsequently, regional accounts were

Fig. 1.1.  Milky sap oozing from cut branch of Euphorbia grandidens, PVB 6870a, Springs, NE of Uitenhage, South Africa, 16 Jul. 2018 (© PVB).

published by P.G. Meyer (1967) for Namibia and by Carter & Leach (2001), where Botswana was included within the area of the Flora Zambesiaca. The 80 years since 1941 have seen much more extensive exploration of southern Africa. During this period some new species have come to light (most of these in formerly inaccessible areas such as Kaokoveld,

© The Author, under exclusive license to Springer Nature Switzerland AG 2022 P. V. Bruyns, Euphorbia in Southern Africa, https://doi.org/10.1007/978-3-030-49402-5_1

1

1 Introduction

2

Namibia), but the taxonomy of Euphorbia in southern Africa has become progressively disorganised, with many names applied in different herbaria in South Africa to quite different species. Further exploration has clarified the distributions of and variation in most of the species. This led to the realisation that some of the ‘species’ discussed by White et  al. (1941) were not distinct (though in many such cases they expressed reservations about their distinctness), while a few others mentioned there were undescribed and a few that they had reduced to synonymy were distinct species. Therefore, The succulent Euphorbieae (Southern Africa) is now considerably out-ofdate and is difficult to use for the identification of recent collections, so that there is an urgent need for a replacement. This book presents a new account of the southern African species of Euphorbia. For our purposes, southern Africa is taken to be most of Africa south of 17°S, including the whole of Botswana, Lesotho, Namibia, South Africa and Swaziland (but leaving out Moçambique and Zimbabwe). The island of Madagascar, with its rich and almost exclusively endemic euphorbiaceous flora, is also not included. In this area Euphorbia is represented by 172 species, which are distributed across all four subgenera. Although none of these subgenera is endemic to the region, there are several sections and subsections that are only found in southern Africa. Of the 172 species found in southern Africa, 128 are endemic. The 44 non-endemic species mainly extend beyond southern Africa into Angola and the southern parts of East Africa but a few, such as E. ingens, are found from subtropical South Africa to southern Ethiopia and Somalia. Although fewer than 10% of the species in Euphorbia are treated here, 74% of these are endemic and so this book will have little overlap with any account of the species from further north. This book brings together the results of the exploration and research that has taken place over the past 80 years. The first section introduces the reader to the diversity in Euphorbia, especially in the vegetative parts, but also in how the flowers are arranged. The terminology peculiar to the floral organization of Euphorbia is explained and illustrated with examples. The second section is the systematic account, where the species are arranged into their sections, subsections and series as these have recently been defined. Keys are provided to all the species within each of these subdivisions, while keys to the subspecies and varieties are provided under the respective species in the systematic account. Of the 172 species in southern Africa, 157 are presented in detail here. Each is illustrated with several colour photographs, with a map showing its known distribution and with line drawings in which some of the minute details of the plant and floral parts are highlighted. This book would not have been possible without the generous help received over many years from many individuals. My early interest in succulents was particularly fostered by excursions into the field with Walter Wisura, formerly the curator for succulents at Kirstenbosch and by M.  Bruce Bayer, formerly the curator of the Karoo Botanic Garden,

Worcester. Succulent species of Euphorbia were collected and recorded during my earlier collecting work on stapeliads, which led to the development of a large data-base of collections in otherwise poorly explored areas. Many kind and helpful farmers have allowed access to their properties to look for Euphorbia and other plants. I particularly wish to thank the following for their hospitality and help: Johann and Odile Becker, Rolf Becker and Alma Möller, Susanne Bell, Susan and Richard Dean, Elke Erb, Wilfried Friedrich, the late Johan Geldenhuys, Michael and Joanne Kroon, Frans and Duberette Labuscagne, Molly, Chris and Marina Lochner, Paul and Linda Loffler, Douglas McMurtry and Shane Burns, Steven, Alicia and Pieter Theron, James and Regina van Vuuren and Gordon and the late Ada Whittal. David Cumming, Sean Gildenhuys and P.J.D. Winter shared valuable information on several little-known species. Paul E. Berry of the University of Michigan involved me in the Planetary Biodiversity Inventory-working-group on Euphorbia and gave financial assistance for visits to some European herbaria. He also provided very helpful criticism of some papers which contributed towards this work and commented on most parts of this work itself. Erich van Wyk and Jean J. Meyer provided regular assistance during visits to the herbarium at PRE, the largest collection of Euphorbia in southern Africa. Christiane Anderson and Rafaël Govaerts have frequently assisted with nomenclatural questions, while Gill Challen at Kew (K) and Arne Anderberg at Stockholm (S) arranged for many scans of specimens in their respective institutions. I am also grateful to the Keeper of the Herbarium, Royal Botanic Gardens, Kew for permission to publish the illustrations by George Bond and Thomas Duncanson of several types of Haworth’s species, as well as several of the lovely water-colour sketches by Nell Lugard. All of these were scanned for me with great care at Kew by Patricia Long. Dmitry Geltman assisted with the world-wide distribution of subg. Esula. John Burrows, Steven P. Fourie, Graham & Kate Grieve, Pavel Hanáček, Niels Jacobsen, J.A.R. de Paiva (Coimbra), Ralf Peckover, Ernst Schmidt and Geoff Tribe each allowed me to use several of their photographs. Finally, I wish to thank Cornelia Klak for her support and encouragement and also for many useful suggestions, which have helped to improve this book substantially.

1.1

Historical Sketch

Since Euphorbia is well-represented in Europe and Asia, it has been known as long as interest in plants has existed. Members of Euphorbia have long been held to have medicinal value and the name itself is derived from ‘Euphorbos’, ostensibly a ‘physician’ to Juba, ruler of the ancient kingdom of Mauretania in West Africa. Nevertheless, the species of the northern hemisphere are rarely succulent. The succulent species have an unusual appearance and were reputed to yield sap

1.1 Historical Sketch

that had great curative value. The first such succulents were recorded from Morocco, the Canary Islands and southern India as early as 1570 (Croizat 1934), as European exploration of these areas began to gather steam. Around 1680 the distinctive and characteristic plants of Euphorbia of southern Africa began to appear in European accounts of exploration at the Cape. Several were cultivated at Cape Town (especially in the garden of Colonel Robert Gordon) and soon cuttings and seed of these were taken to and cultivated in Europe. It was immediately obvious that these succulent species exhibited greater diversity than those known from elsewhere. How this diversity was to be dealt with taxonomically was not quite understood. This one sees from the account of Euphorbia in Linnaeus’ Species Plantarum of 1753, which contains 56 species and is surprisingly confused about the southern African succulent species (see for example what he included under E. caput-medusae). Some of this confusion arose in the doctoral dissertation from 1752 of his student Johannes J.F.  Wiman and was taken over into the Species Plantarum (Croizat 1934). The documentation of Euphorbia in southern Africa can really be said to have begun in earnest with the explorations of C.P. Thunberg and Francis Masson between 1772 and 1795. Thunberg spent nearly 3 years at the Cape and during this time he and Masson explored some areas together. Masson spent a total of about 12 years at the Cape during two visits and left for England for the last time early in 1795. Thunberg made specimens of several species of Euphorbia, including several succulents. Such novelties as Euphorbia meloformis were among the many remarkable species that they encountered and which Masson sent back as live specimens to England. Nevertheless, neither seems to have had a particular interest in Euphorbia and the same is true of the famous collectors Ecklon & Zeyher and the brothers Drège, who discovered many other species before 1840. These were gathered during their general collecting activities, which were so comprehensive that the first floristic accounts for South Africa could be compiled from them. Several new species were described by Haworth from among the many succulents sent back to Kew by James Bowie, who explored at the Cape from 1816 until 1823. During his preparation of the accounts of Euphorbia for the Flora of Tropical Africa and the Flora Capensis, N.E.  Brown (1911–12; 1915) was especially aided by the collecting activities in South Africa of Rudolf Marloth. Marloth was the first botanist who took a particular interest in Euphorbia, discovering many new species and describing many of these himself. Marloth also tried to clarify the identity of some of the lesser-known species, such as E. virosa. He and others in South Africa assisted Brown with material of some of the succulent species, which were generally poorly represented in herbaria. Marloth knew and documented most of the widely distributed species and completed the basic exploration for Euphorbia in South Africa. Nevertheless, significant discoveries have been made subsequent to his work. In particular, many quite local species

3

have been discovered since then, especially in the north-east of South Africa (where many were discovered by F.Z. van der Merwe) and in the north-west of Namibia, where Major Hahn and Max Otzen found several for the first time. Most of these were described by R.A.  Dyer and a few more were named later by L.C. Leach.

Fig. 1.2.  Rudolf Marloth in 1887 (© Stellenbosch University Archives).

Fig. 1.3.  Rudolf Marloth in the field near Cape Town in 1912 (© Stellenbosch University Archives).

4

After Marloth, possibly the person with the greatest interest in Euphorbia was Robert Allen Dyer. Dyer worked from 1925 until 1930 at the Albany Museum in Grahamstown. Initially he was the Botanical Survey Officer and later, after the retirement of S.  Schönland, he became curator of the Museum’s Herbarium. This is an ideal place for someone interested in succulents and particularly in Euphorbia, which is especially diverse in that area. His investigations there led to an account of Euphorbia in the Eastern Cape (Dyer 1931), where he placed on record many of his discoveries. From 1931 to 1934 he worked at the Herbarium at Kew, England and this experience turned his attention away from ecology towards taxonomy. After returning to South Africa he continued collecting while employed at the Botanical Research Institute in Pretoria. He documented the distribution of Euphorbia widely over South Africa and made more specimens of the succulent spe-

1 Introduction

cies than any previous collector, with the intention of producing a monograph of the southern African species (Dyer 1979). However, Alain Campbell White (1880–1951) and Boyd Lincoln Sloane (1886–1955), after completing their first monograph The Stapelieae of 1937, decided that succulent Euphorbia was their next project. For this, they approached Dyer, who later joined them in it, but insisted that the monograph be restricted to the southern African species and that he was made a co-author. For this work he supplied many photographs of plants in habitat and in cultivation in South Africa and he also provided the keys and most of the detailed discussion of the species. Together, the three authors produced the now famous two volumes on the southern African representatives of the Succulent Euphorbieae (White et al. 1941). Further volumes were planned, together with P.R.O.  Bally, but they never materialized (Sloane 1952; Dyer 1979).

Fig. 1.4.  R.A. Dyer in the field. Left: on Tristan da Cunha, 1937 (© Albany Museum). Right, camping on the Blouberg, Jan. 1955 (© Tristan Dyer).

A further major contributor to the study of the succulent species of Euphorbia in southern Africa was the amateur botanist Leslie Charles (Larry) Leach (1909–96). Born in Essex in England, Leach worked as an electrical technician in the army in Britain until he emigrated to Southern Rhodesia (later Zimbabwe) in 1938. There he set up a business supplying electrical equipment, especially batteries for vehicles. His business was successful enough for him to sell it off and ‘retire’ in 1956 before he was 50. This gave him more time to pursue his programme of collecting and documenting the succulent members of Euphorbia as well as stapeliads and Aloe in Zimbabwe and other parts of Africa south of the equator. He had begun these studies in 1950 and continued them in Zimbabwe, working as Honorary Botanist at the National Herbarium in what was then Salisbury until he emigrated to South Africa in December 1981. This

research and exploration in areas that were previously littlecollected (such as Angola and Moçambique) resulted in many papers on the taxonomy of Euphorbia. Most of these involved descriptions of new taxa, but there were also several that clarified the identity of previously misunderstood species, such as E. berotica, E. candelabrum and E. virosa. Once in South Africa, he briefly established himself at the Botanical Research Institute in Pretoria and then was employed at the ‘Karoo Botanic Garden’ at Worcester from 1982 to 1989. After this he worked as Honorary Research Fellow in the Department of Botany at the University of the North in what was then Pietersburg (now Polokwane) until his death on 18 July 1996 at the age of 86. From 1982 onwards he described several new species from southern Africa, but not all of these have stood the test of further collecting.

1.2 Classification of Euphorbia

5

The relatively thorough exploration of the Namibian flora, from 1897 until 1935, by Moritz Kurt Dinter brought to light hardly any species of Euphorbia that had not already been discovered in South Africa or were not found around the same time in Namibia and described by Marloth. The extensive collecting by Merxmüller, Giess and Volk between 1956 and 1968, prior to the publication of Prodromus einer Flora von Südwestafrika, also revealed almost no new species of Euphorbia. In the remote north-west (the Kaokoveld), which neither Dinter nor Marloth visited, there were species whose identities were uncertain. In several cases, these were first noticed by Max Otzen (4 Jan. 1871–10 Aug. 1948). One is even mentioned in White et al. (1941, with figures 1095 and 1096). Some (such as E. eduardoi) were recorded later again by the botanist P.G. Meyer during his preparation for the account of Euphorbia in the Prodromus but Meyer did not describe any of these and they were later mostly described by Leach. White et al. (1941) gave a very thorough account of people who had discovered species of Euphorbia and the general progress of this discovery. This is not repeated here and, for additional biographical information, the reader is referred to Gunn & Codd (1981) and Glen & Germishuizen (2010).

Fig. 1.5.  L.C. Leach, Nov. 1962 (© Larry Leach Herbarium, University of the North).

Fig. 1.6.  L.C. Leach, c. 1962, finding directions out of Magude, Moçambique, with his trusted VW-Kombi close by (© Larry Leach Herbarium, University of the North).

1.2

Classification of Euphorbia

With some 5000 species, the Euphorbiaceae is one of the larger families of flowering plants. It is divided into five subfamilies, with five tribes making up the subfamily Euphorbioideae (Webster 1994). Of these tribes, the Euphorbieae includes somewhat under half the total number

of species and consists of Euphorbia and several very small genera that are mainly found in the tropical parts of both the Old and the New Worlds. A feature common to all members of the tribe Euphorbieae is the greatly reduced flowers that are tightly gathered together inside flower-like structures known as cyathia (for more details of these structures, see below).

6

Webster (1975) divided the tribe Euphorbieae into three subtribes, of which the nearly cosmopolitan Euphorbiinae is by far the largest. In this subtribe the involucre is made up of five fused bracts (rather than four), the male flowers lack a perianth and the female flowers are also mostly devoid of a perianth. In the past (as, for example, in Webster 1994), the subtribe Euphorbiinae consisted of the seven genera: Chamaesyce (with ± 300 species), Cubanthus (3 spp.), Endadenium (1 sp.), Euphorbia (± 1450 spp.), Monadenium (± 80 spp.), Pedilanthus (14 spp.) and Synadenium (4 spp.). The genera Poinsettia (24 spp.) and Elaeophorbia (4 spp.) have also occasionally been recognised. In Africa Chamaesyce, Elaeophorbia, Endadenium, Euphorbia, Monadenium and Synadenium occur naturally and all of these except Chamaesyce and Euphorbia are endemic. Chamaesyce was distinguished by vegetative features (apical abortion of the main shoot followed by sympodial growth from buds lower down, the presence of interpetiolar stipules and opposite pairs of often asymmetrical leaves), the infrequently recognised Elaeophorbia was distinguished by its indehiscent fruit (a feature also found in several species of Euphorbia). Endadenium, Euphorbia, Monadenium and Synadenium were separated purely on details of the cyathium (Carter 1988; Carter and Leach 2001). The very large, almost cosmopolitan genus Euphorbia has been subdivided in many different ways. For example, Wheeler (1943) recognised eight subgenera: Agaloma, Chamaesyce, Eremophyton, Esula, Lyciopsis, Poinsettia, Rhizanthium and Tithymalus. For the naturally occurring tropical and southern African species of Euphorbia, Gilbert (1987) recognised four subgenera: Chamaesyce, Esula, Euphorbia and Lacanthis. This was modified to six subgenera in Gilbert (1995). On the other hand, eight subgenera were recognised by Carter (1988) for the naturally occurring Tropical East African species and by Holmes (1993) for the naturally occurring Somalian species: namely Chamaesyce, Eremophyton, Esula, Euphorbia, Lacanthis, Lyciopsis, Tirucalli and Trichadenia. These arrangements differed considerably, since Carter’s concepts of the subgenera Eremophyton, Lyciopsis, Trichadenia and Tirucalli fell under a broadly defined subgenus Esula of Gilbert (1987). In their account for Flora Zambesiaca, Carter & Leach (2001) abandoned these subgenera and instead recognised 14 sections, though they mentioned that sect. Euphorbia ‘would be better regarded as a Subgenus’. Clearly then, the subgeneric classification of Euphorbia is complex and confusing (Leach 1976c). This confusion is much aggravated by (a) the very large number of species involved. (b) the broad distribution of the genus as a whole and of each of the subgenera. (c) the complexity of previous classifications and.

1 Introduction

(d) the extent of convergence in many vegetative features and the consequent uncertainty as to which features define natural groups. The problem of convergence in Euphorbia is especially acute. Convergence is now known to have taken place in vegetative features such as annual habit, geophytic habit, succulence of the shoots, development of thorns (even its development from the same organ, e.g. the tips of the shoots, has occurred several times independently), succulents with photosynthetic shoots and conspicuous, deciduous leaves, succulents with terete photosynthetic shoots with reduced leaves, the so-called ‘pencil-plant’ which developed in Australia (in E. sarcostemmoides), several times independently in South Africa (as in E. burmanni, E. dregeana, E. gregaria and E. mauritanica for example, each belonging to a different subgenus but all considered to be closely related in White et al. 1941), in South America (as in E. appariciana) and in North America (as in E. antisyphilitica), tuberculate stems with leaves borne on tubercles or phyllopodia and the phenomenon of shiny, peeling bark on the thicker and older shoots. As in many groups of plants (as well as other organisms), analysis of the wealth of new data derived from DNA-­ sequences has provided entirely new insights into relationships within the Euphorbiaceae. As a consequence of these new results, the Euphorbiaceae was split into three families: Euphorbiaceae, Phyllanthaceae (including Phyllanthus) and Picrodendraceae (including Hyaenanche) (APG IV). Of greater interest here is that these new techniques significantly sorted out the previously intractable problems of relationships between the major groups within Euphorbia (Leach 1976c) and enabled a new classification of Euphorbia to be developed. These techniques were first applied to Euphorbia by Steinmann & Porter (2002) who assembled and analysed data from DNA-sequences from the nuclear ITS and plastid ndhF gene-regions of 192 species from Chamaesyce, Endadenium, Euphorbia (including a few from Elaeophorbia and Poinsettia), Monadenium, Pedilanthus and Synadenium. Other preliminary studies reconstructing the phylogeny of Euphorbia with DNA-data (Bruyns et al. 2006, 2011; Park and Jansen 2007; Zimmermann et al. 2010) yielded similar results, often with much improved resolution and support, but all entirely different from the results obtained when morphological characters were analysed (Park 1996; Park and Elisens 2000). Subsequently data from all three different genomes (mitochondrial, nuclear and plastid) was used to try to strengthen the support for relationships among major groups in the genus revealed by previous work (Horn et al. 2012). Steinmann & Porter (2002) and Horn et al. (2012) found that three subtribes of the Euphorbieae of Webster (1975) were very well-supported. Surprisingly, the subtribe Euphorbiinae broke into four clades (initially termed A, B, C and D), which did not correspond to any taxa that had been

1.2 Classification of Euphorbia

recognised before. Three of them (A, C and D) contain a remarkable mix of subgenera and sections of Euphorbia, while the fourth (clade B) was mainly made up of Euphorbia subg. Esula (in the sense put forward by Wheeler in 1943). Even this clade exhibited some novel features, since a small group of African and Arabian succulents usually placed in ‘sect. Tirucalli’ (e.g. Carter 1988; Holmes 1993; Gilbert 1995; Carter and Leach 2001) was nested within it, while E. tirucalli itself was not part of it (corroborating suggestions along these lines made by Leach 1975a, 1976c). At the generic level, the universally accepted segregate genera Cubanthus, Endadenium, Monadenium, Pedilanthus and Synadenium as well as the less widely recognised genera Chamaesyce, Elaeophorbia and Poinsettia were all nested among species traditionally placed in Euphorbia. At the subgeneric and sectional levels, most of the previously recognised taxa were not monophyletic. Also, the relationships among the species have turned out to be quite different to those suggested before (e.g. for sect. Euphorbia, compare the relationships found in Dorsey et al. 2013 with those suggested by Carter 1994). This new information made it clear that major taxonomic re-organization was necessary within Euphorbia to achieve a classification that reflected these results. The process of establishing a monophyletic Euphorbia was initiated by Steinmann (2003), who placed all species of Pedilanthus in Euphorbia, where they now reside in sect. Crepidaria of subg. Euphorbia. It was taken a step further in Bruyns et al. (2006), who recognised Clades A–D of Euphorbia as four subgenera and moved all species of Endadenium, Monadenium and Synadenium into subg. Euphorbia (clade C). Steinmann et  al. (2007) completed this process and moved the species of Cubanthus into Euphorbia, where they now form sect. Cubanthus in subg. Euphorbia. Under the auspices of the Planetary Biodiversity Initiative project on Euphorbia based at the University of Michigan, USA, investigations were conducted on each of the four subgenera in much greater detail, to infer relationships among the species and to classify each subgenus into sections. These investigations by Yang et al. (2012) for subg. Chamaesyce (where 291 of the 566–574 species were sampled), Dorsey et al. (2013) for subg. Euphorbia (216 of 661 species sampled), Peirson et al. (2013) for subg. Athymalus (88 of 148 species sampled) and Riina et al. (2013) for subg. Esula (273 of 457 species sampled) placed virtually all known species in their respective subgenera, in the process obtaining sequences for 868 species out of 1840 or around 47% of the total. The monophyly of each of the four subgenera was confirmed and no species has yet been found that does not fit into these four subgenera. The notion that Euphorbia is ‘too large’ has inspired various attempts to break it up. Early efforts were those of Haworth (1812) and Klotzsch & Garcke (1860a, 1860b). In the latter case Euphorbia was split into seven genera, while Haworth had slightly more. All the spiny species remained in Euphorbia and this included such distant relatives as E.

7

polygona and E. canariensis, for example. A. Berger (1906) and N.E. Brown (1911–12; 1915) placed all of them back in Euphorbia except for a few, such as Monadenium, which were considered to be florally sufficiently distinctive to remain separate from Euphorbia. The splitting of Euphorbia began again with Koutnik (1984b, 1987, for example), who re-established Chamaesyce (sect. Anisophyllum in the present arrangement). This directly contradicted Webster (1967: 420–421), who assessed the pros and cons of recognising Chamaesyce as a genus and showed that there were no significant features to distinguish it from Euphorbia. For this reason attempts to split off Chamaesyce from Euphorbia did not find general favour, though Carter & Leach (2001) still maintained (without substantiation) that the evidence for recognising their ‘sect. Chamaesyce’ (sect. Anisophyllum in the present arrangement) as a genus ‘has been accepted by almost all workers’. In the same manner, it was suggested by Gilbert (1987, 1995) and Carter (1994) that some of the groups that they recognised in Euphorbia could be split from Euphorbia into separate genera. For example, Gilbert (1995) stated that each of the six subgenera he recognised could be treated as a genus and that the species bearing spine-shields (now recognised as sect. Euphorbia), were ‘a distinctive group lacking obvious relatives elsewhere in the genus’. Analysis of DNA-data (as well as better graphical methods for displaying the relationships revealed and a statistical assessment of the confidence in these relationships) has made it clear that the groups singled out for being split off (such as sect. Euphorbia or sect. Anisophyllum) and others that were split off previously (like Cubanthus, Monadenium, Pedilanthus and Synadenium) are deeply nested within Euphorbia. For example, the species of Monadenium and Synadenium were found to be closer to E. antiquorum (the type of Euphorbia) than such ‘typical’ species of Euphorbia as E. mauritanica or E. clavarioides. Here, one possibility is to break up Euphorbia and recognise many smaller genera. If the spine-shield-bearing species (i.e. sect. Euphorbia) were recognized as a genus, this would involve anything up to 1500 name-changes (since E. antiquorum, the type of Euphorbia, is a spine-shield-bearing species and so all but the ± 350 species of sect. Euphorbia would remain in Euphorbia). Such a treatment would replace the relatively easily recognised, widely known and almost cosmopolitan Euphorbia with many new and often small genera. Even more importantly, these new genera would be difficult to separate from one another, especially now that the extent of convergence within Euphorbia is better understood. Following Steinmann & Porter (2002), recent work has moved even further from the scenario of carving up Euphorbia and placed all members of the Euphorbiinae within the single genus Euphorbia. Euphorbia is then recognised as a fairly ancient radiation that has evolved into many species-rich lineages over most of the globe. This endorses the view of Dyer (1957: 493) that Euphorbia provides a per-

8

fect and beautiful phylogenetic example of how great diversity in plant forms can be derived ‘from a common stock’. It also substantiates the opinion of Croizat (1972), who believed that Monadenium made more sense phylogenetically as part of Euphorbia.

1.2.1 R  elationships among and within the Subgenera The relationships between the four subgenera in Euphorbia that were revealed first by Steinmann & Porter (2002) were

1 Introduction

refined with successively increasing support by later work (Bruyns et  al. 2006, 2011; Park and Jansen 2007; Zimmermann et al. 2010). Horn et al. (2012) used nine loci from all three genomes (chloroplast, mitochondrial and nuclear) to place this on a firmer basis, once more finding the relationships between the four subgenera as in Fig. 1.7. Here the earliest diverging branch is subg. Esula, with subg. Athymalus arising after that and the other two the most recent. Each of the four subgenera spread into southern Africa between two and six times, so that Euphorbia in southern Africa is not a monophyletic entity but is made up of survivors from many distinct invasions.

Fig. 1.7. Relationships between the four subgenera in Euphorbia as derived from analysis of DNA-data. The total number of species, the numbers of species in southern Africa and the number of southern African endemics is given for each. The length of the vertical side of each of the triangles is proportional to the number of species in that subgenus (© PVB).

1.2.2 Species Concepts in Euphorbia As with many plants, a simple, intuitive ‘folk concept’ of species (Cronquist 1988) has mostly been applied, where species are groups of ‘essentially similar’ individuals. In this sense, as with most genera, Euphorbia contains many easily recognised species that are impossible to confuse with any other (e.g. E. hallii, E. phylloclada and E. stellispina). It also contains complexes, where the limits of the species are harder to discern. Examples are the species around E. procumbens in subsect. Medusea and the complexes around E. rhombifolia and E. spartaria in sect. Articulofruticosae. In the past, where only small amounts of material were available and variability was impossible to assess, this led to the recognition of ‘species’ that do not occur in nature as distinct elements. This phenomenon is especially noticeable in sect. Articulofruticosae. Here, in several widely distributed species of the winter-rainfall region one finds a distinct dwarfing and thickening of branches as one proceeds northwards into the more arid Northern Cape. This is not clear without the considerably expanded herbarium-record that now exists and such information on variability was not available to N.E. Brown

and earlier workers. Especially obvious examples of this are in E. burmanni (with the thick-branched forms often known as E. karroensis), in E. rhombifolia (where the dwarfed coastal forms were known as E. angrae and the much larger ones were often called E. chersina) and also in the distinctive E. stapelioides (where the more slender southern forms were named E. lumbricalis). Outside this section, the same phenomenon is found in E. celata of sect. Anthacanthae, where the names E. miscella and E. namuskluftensis were applied to the stouter-branched forms occurring to the north of more slender ones. In the complex surrounding E. rhombifolia, N.E. Brown (1911– 12; 1915) recognised at least nine different names which cannot be located as distinct species in nature and exist only as the types of his names. White et al. (1941) did not know these species well and so did not reduce many of these names. In the complex around E. heptagona, where White et  al. (1941) recognised the three species E. atrispina, E. enopla and E. heptagona, they (i.e. Dyer) knew the plants well in habitat and expressed scepticism as to their distinctness. Here their scepticism has been corroborated by further collections and observations in habitat, which have shown the presence of only a single entity

1.2 Classification of Euphorbia

across a relatively wide range. Field-­work has also shown that the many names applied to E. caput-medusae do not express the actual relationship between the plants involved, where there is continuity between the largest, most robust plants of the Olifants River valley (formerly E. tuberculata) through to the smaller, more slenderstemmed plants on the Cape Peninsula (E. caput-medusae), the even more slender-stemmed, often rhizomatous ones in the sandy areas around Cape Town and further east (E. marlothiana and E. muirii) to the very dwarfed plants found in coastal Namaqualand (E. ramiglans etc.). Over this range there is considerable vegetative variation in this species (mainly in the length and thickness of the branches) but the floral parts are indistinguishable in all of them. Among the succulent Aizoaceae, revisions based on extensive field-work revealed as many as five times (and sometimes up to 20  times) as many names as species (Ihlenfeldt and Gerbaulet 1990; Hammer 1993; Klak and

9

Linder 1998). L.C. Leach found that certain species of stapeliad (such as Ceropegia mixta  =  Orbea variegata and C. lutea = O. lutea) also have many synonyms. In Euphorbia, Leach established important synonymy for such species as Euphorbia virosa. While he was also instrumental in showing how to separate the members of sect. Tirucalli in the Namib Desert of Angola, Namibia and South Africa, he used the existence of differences in the material before him to define species. Where he worked on material from previously unexplored regions (such as Angola and northern Moçambique), this was not problematic. In the much better-­ explored southern Africa, this relatively simplisitic view was not always backed up by adequate field-experience and knowledge of the variability of the species. Differences that could be found among the relatively few collections before him did not necessarily exist in the wild and led him to describe too many taxa. Examples of this are E. miscella and E. namuskluftensis = E. celata; E. mira = E. silenifolia; E. glandularis = E. exilis and several others.

Fig. 1.8.  Variation in E. silenifolia. The plant on the left would be E. silenifolia, while that on the right would be ‘E. mira’. However, it was clear from its lack of flowers and small tuber that the plant on the left is a juvenile, while that on the right is mature. PVB 13737, just north of Arniston, South Africa, 29 Jun. 2019 (© PVB).

How populations and individuals can be grouped into species was summarized in detail by Luckow (1995) and it is widely accepted that persistent discontinuities in at least two ‘good’ characters define species (e.g. Stebbins 1950; Hedberg 1957; Wiley 1981; Sidwell 1999). Among flowering plants, a ‘good’ character is taken to be one which is readily observed (i.e. usually a morphological character for which a magnification of ×10 is enough) and where variation in these characters is assessed by measuring or counting. These discontinuities are believed to result from reproductive isolation, since interbreeding normally causes such differences to disappear. Species are generally defined here in this way. Nevertheless, in a few cases taxa separated by consistent differences in only one character have been found to co-exist but maintain their distinctness. A good example of

this is E. rhombifolia and E. spartaria. These differ in the shape of their leaf-rudiments but are florally indistinguishable and occur together frequently, without any sign of hybridizing. Consequently, one must assume that they are reproductively isolated (despite their floral similarity) and so they are recognized as distinct species. For taxa which are spatially separated but which differ only very slightly in morphological features, some authors have taken the view that they could interbreed if they co-­ existed and so they ought to be treated as the same species (Mayr 1964, but Cracraft 1992 took a very different view). Here such taxa are often recognised at levels below the rank of species. In this account the rank of subspecies is applied where two geographically complementary taxa occur which differ in only one ‘reasonably reliable’ character. An example

1 Introduction

10

is the pair of ‘species’ E. obesa and E. symmetrica. In nature they occur far apart and cannot interbreed. However, they differ only in minor (and somewhat unreliable) features of their floral parts and in cultivation they interbreed readily to produce fertile offspring. Here they are treated as subspecies. The rank of variety is only rarely used (as in E. lydenburgensis): where taxa that occur together differ in a single, often quite distinctive character, but where intermediates may be found.

1.3

Morphology of Euphorbia (mainly in Southern Africa)

It has been suggested (Dressler 1957; Steinmann and Porter 2002; Horn et al. 2012) that the ancestors of Euphorbia were non-succulent, woody shrubs or trees. When mature, such

Fig. 1.9.  Ancestral growth-form of Euphorbia, here seen in Euphorbia guerichiana, tree ± 4  m tall, on sparsely wooded slope overlooking Namib Desert and Hartmann Mountains, SE of Camp Synchro on Kunene River, Namibia, 27 Dec. 2014 (© PVB).

plants consist of roots, stem (or trunk), branches, leaves, flowers and fruit. Although the size of the plant varies in Euphorbia, most species conform to this basic structure. There is a particularly wide range of sizes within subg. Chamaesyce and subg. Euphorbia (from minute herbs 10–20 mm tall or small geophytes not exceeding 20 mm tall above the ground to trees 20 m tall or more). This variety of sizes is less pronounced in subg. Athymalus (which contains trees 3–5 m tall to small geophytes rising 1–5 cm above the ground) and the range is the least within subg. Esula, where most species are shrubs. In Euphorbia the ancestral shrub or tree has evolved in many directions. Among the southern African species the following are among the most striking: (1) Reduction of the plant in size. This has developed further along at least five lines: (a) shoots are short and packed into dense mounds or mat-like plants, as in E. clavarioides and E. polycephala in subg. Athymalus or E. mosaica in subg. Euphorbia from NE Africa. (b) geophytic habit (see below under ‘Rootstock’) (c) horizontally spreading, underground rhizomes up to 0.3 m long (as in E. stapelioides of subg. Chamaesyce, shorter in E. patula and E. polygona of subg. Athymalus; E. knuthii and E. namuliensis of subg. Euphorbia). Subterranean rhizomes are slender and without angles or tubercles and become thicker, erect, often tuberculate and even angled on emerging from the soil. (d) loss of branches with the plant reduced to a single, often quite small and short stem, as in E. meloformis and E. obesa of subg. Athymalus; E. gymnocalycioides and E. turbiniformis of subg. Euphorbia from NE Africa. Here the juvenile form, consisting of a stem only, has become reproductive without developing the additional shoots of a fully-­sized adult. (e) development of a ‘medusoid habit’: thick stem often greatly reduced in length, with slender often short branches radiating from it. This is well-­known in many members of subsect. Medusea in subg. Athymalus. In subg. Euphorbia, the succulent trees and robust shrubs with a trunk in sect. Euphorbia have the same form (thick, many-angled stem, slender, fewer-angled branches) and here the ‘medusoid habit’ varies from these large forms to dwarfs such as E. schizacantha in NE Africa, E. clavigera and E. stellata in South Africa.

1.3 Morphology of Euphorbia (mainly in Southern Africa)

11

(2) Increasing succulence. In Euphorbia this is usually associated with reduction in the size of the leaves, which become increasingly ephemeral and their photosynthetic function is taken over by the shoots, where the epidermis is long-lived and not rapidly replaced by bark. In southern Africa, of the 172 species of Euphorbia, 135 or 79% are succulent (Table 1.1). This includes all 47 members of subg. Euphorbia, three of the 11 species of subg. Esula, 67 of the 80 species of subg. Athymalus (where the other ten are only slightly succulent or woody shrubs) and 18 of the 34 species of subg. Chamaesyce. Fig. 1.10.  Modified growth-form into dense low mound in Euphorbia clavarioides, PVB 13079, Ubala, north of Kokstad, South Africa, 21 Dec. 2015 (© PVB).

Fig. 1.11.  Rhizomatous growth-forms. A, rhizomatous branches excavated in Euphorbia stapelioides (green parts to ± 5 mm thick), PVB 3947, foot of Kortdoringberg, east of Alexander Bay, Namaqualand, South Africa, 13 Jul. 1997. B, rhizomatous branches excavated in Euphorbia patula subsp. patula (green parts to ± 12 mm thick), PVB 12550, NE of Addo, near Port Elizabeth, South Africa, 16 Dec. 2012. C, Euphorbia knuthii subsp. knuthii, rhizomatous growth and small tubers excavated (green parts to ± 10 mm thick), PVB 4466, near Ndumu, South Africa, 11 Jan. 2004 (© PVB).

Fig. 1.12.  Plant reduced to single short, squat stem partly embedded in the ground. Left, in subg. Athymalus, Euphorbia meloformis, ± 10 cm diam., PVB 12255, NE of Peddie, South Africa, 5 Dec. 2012. Right, in subg. Euphorbia, Euphorbia gymnocalycioides, ± 4 cm diam., south of Negele, Ethiopia, 18 Nov. 2009 (© PVB).

12

1 Introduction

Fig. 1.13.  Medusoid growth-form with thick partly subterranean stem and slender branches radiating near its apex. Left, in subg. Athymalus, Euphorbia crassipes, just south of Britstown, South Africa, 25 Mar. 2013. Right, in subg. Euphorbia, Euphorbia schizacantha, PVB 12749, between Garissa and Garsen, Kenya, 19 Jun. 2014 (© PVB).

Whereas shoots in cacti have been elaborately modified in almost all aspects, succulents in Euphorbia are anatomically similar to their non-succulent relatives and lack many of the specialized features associated with other desert-adapted plants (Mauseth 2004a, b). In Euphorbia as a whole, nearly 500 species out of the total of 1840 exhibit succulence, so that the preponderance of succulents in southern African is not typical for the genus elsewhere in the world. Most of these nearly 500 succulent species belong to subg. Euphorbia in the Old World of tropical Africa, Madagascar, the Arabian Peninsula and SE Asia and to subg. Athynalus in southern Africa. In contrast, there are only a few succulents in subg. Esula (these are only found in Africa, Macaronesia and Madagascar) and in subg. Chamaesyce. Although Euphorbia is well represented in the New World (i.e. the Americas), only relatively few species in subg. Euphorbia and subg. Chamaesyce in Brazil, Chile and Mexico are succulent (Yang et al. 2012; Dorsey et al. 2013).

1.3.1 Rootstock In Euphorbia the base of the stem often continues into a strongly-developed tap-root anchoring the plant firmly in

the ground. Most roots arising from the tap-root are fine and fibrous. However, in several species, such as E. hallii, E. inermis, E. oxystegia and E. quadrata in subg. Athymalus (usually with relatively conspicuous, deciduous leaves), in E. espinosa of subg. Chamaesyce and also in some tropical members of sect. Monadenium, the taproot and some side-roots may be swollen and fleshy. These are not to be confused with geophytes. In true geophytes, the stem forms part of an underground tuber whose apex is below the surface of the ground. Here the lower part of the tuber is the rootstock, the upper part is the stem and usually the two are not clearly separated either within the tuber or externally. In the rainy season leaves develop above the ground from the apex of the tuber, often in rosettes on the surface. They are lost during dry periods, when the plant ‘retreats’ underground. Geophytes have evolved many times independently in subg. Euphorbia (in several lineages, as in E. rubella from NE Africa and E. acaulis from SE Asia), in two lineages in subg. Athymalus, in E. gueinzii of subg. Chamaesyce and in E. striata and allies of subg. Esula. Most geophytes are reduced to a short stem bearing leaves, but others may develop an almost shrublike structure underground with leaves at the tips of the shoots (as in E. tuberosa).

Number of species 147

± 570

± 460

± 660

± 1837

Subgenus Athymalus

Chamaesyce

Esula

Euphorbia

Total

cosmopolitan (mainly Eurasia) cosmopolitan (especially Africa, Madagascar) cosmopolitan

Distribution Africa, Arabian Peninsula to Iran, Macaronesia cosmopolitan

64 (14)

21 (3)

21 (2)

15 (6)

Number of sections (southern Africa) 7 (3)

172 (128) 74%

34 (21) 62% 11 (9) 82% 47 (28) 60%

% endemic 80 (70) 88%

Southern Africa (endemic)

55 (12) 22%

20 (5) 25% 4 (1) 25% 13 (3) 31%

% endemic 19 (2) 11%

Namibia (endemic)

Table 1.1.  The subgenera of Euphorbia, showing the numbers exhibiting different growth-forms.

138 (85) 62%

22 (6) 27% 9 (4) 44% 35 (20) 58%

14 8%

0

10 29% 0

South Africa (endemic) Annuals % of % endemic total 72 (55) 4 76% 5%

9 5%

3 9% 5 45% 0

% of total 1 1%

Perennial, non-succulent herbs

4 2%

0

2 6% 0

% of total 2 3%

Woody shrub to tree

135 78%

18 53% 3 23% 47 100%

% of total 67 84%

succulents

10 6%

1 3% 3 23% 0

% of total 6 8%

geophytes

1.3 Morphology of Euphorbia (mainly in Southern Africa) 13

1 Introduction

14

Fig. 1.16.  Deciduous succulent leaves in the geophytic Euphorbia acaulis (subg. Euphorbia) among short grasses, leaf-litter and chunks of granite, PVB 11461, Nandi Hills near Bangalore, Karnataka, India, 16 Aug. 2009 (© PVB).

Fig. 1.14. Swollen roots emanating from the tap-root in Euphorbia inermis, PVB 11005, Port Elizabeth, South Africa, 29 Dec. 2008 (© PVB). Fig. 1.17. Three plants of the geophytic Euphorbia ecklonii (subg. Athymalus) removed to show leaves radiating on surface of ground just above tuber, PVB 11162, Swellendam, South Africa, 22 Aug. 2008 (© PVB).

1.3.2 Shoots

Fig. 1.15.  Small geophyte ±  3  cm tall in flower among pieces of limestone and leaf-litter, Euphorbia rubella (subg. Euphorbia), Mt Achim, Harrar, Ethiopia, 29 Jan. 2015 (© PVB).

Most of the southern African species of Euphorbia have fleshy shoots with persistent photosynthetic epidermis that may live and function for decades, where the development of bark is retarded or wholly prevented. The shoots may be cylindrical or distinctly angled. Species with angled shoots all belong to subg. Euphorbia and subg. Athymalus and are

1.3 Morphology of Euphorbia (mainly in Southern Africa)

not found in the other two subgenera, where succulent stems are always cylindrical (though they may sometimes be somewhat longitudinally ridged).

Fig. 1.18. Cylindrical branches in subg. Athymalus and subg. Euphorbia. A, Euphorbia dregeana, Harrasberg, Namaqualand, South Africa, 16 Jul. 1995. Right, Euphorbia carunculifera, NE of Namibe, Angola, 18 Mar. 2017 (© PVB).

15

So, in semi-arid parts of southern Africa it is not unusual to see representatives of Cynanchum (especially the former Sarcostemma, Apocynaceae), Euphorbia (sometimes from two different subgenera) and Kleinia (Asteraceae) growing socially, all exhibiting this growth-form and all looking quite similar vegetatively.

Fig. 1.20. Pencil-like branches of Euphorbia lignosa (subg. Athymalus) gathered into a dense low shrub, PVB 10067, East of Namusberg, Rosh Pinah, Namibia, 15 Jul. 2005. Behind this are several yellowish plants of E. rhombifolia (subg. Chamaesyce) also with pencil-like branches (© PVB).

Tuberculate shoots are common in succulent members of Euphorbia, mainly in subg. Athymalus and subg. Euphorbia. With increasing succulence and reduction of the leaf, the base of the leaf swells and raises the leaf out of the surface of the stem on a tubercle (sometimes referred to as a phyllopodium or podarium). These tubercles are often longitudinally elongated (as in E. bubalina) but may also be polygonal to nearly circular at their base. In the highly succulent members of subg. Athymalus (in sect. Anthacanthae) tubercles cover the shoots (very spread

Fig. 1.19.  Angled branches: Left, in Euphorbia pentagona of subg. Athymalus, PVB 6892, Fort Beaufort, South Africa, Sept. 2019. Right, in Euphorbia avasmontana of subg. Euphorbia, PVB 12874, Otavi Mountains, Namibia, Sept. 2019 (© PVB).

A widespread type of cylindrical shoot in Euphorbia is the so-called ‘pencil-stemmed’ form. In these the plant consists of slender, cylindrical, photosynthetic branches that are usually erect and gathered together into a shrub, mostly also with reduced leaves. This has evolved in each of the four subgenera, sometimes even more than once and is an adaptation to semi-arid conditions that is found in several families.

Fig. 1.21.  Tuberculate branches of Euphorbia hypogaea, PVB 6684, Juriesfontein, Loxton distr., Great Karoo, South Africa, 8 Apr. 1996 (© PVB).

16

1 Introduction

out in most species of subsect. Pseudeuphorbium, densely packed together in subsect. Medusea and others) and they are often spirally arranged along the stem and branches (as in E. clandestina and many others). In ser. Meleuphorbia they are fused vertically into angles along the stem and branches and here the individual tubercles are demarkated by a darkened outline if they are not still partly separated. The angles are mostly relatively low and mostly between four and eight per shoot, which is then square or polygonal in cross-section. They are most pronounced as slender, flattened (often undulating) wings with deep furrows separating them in E. polygona, where the shoots may be particularly stout (to 200 mm thick) and up to 20-angled (reaching nearly the same number in the not quite so stoutly branched E. stellispina) and here shoots are star-like in cross-section. Each tubercle is tipped by a leaf (often rudimentary and soon lost, leaving a small to prominent apical scar) and the axil of the leaf (from where sterile or fertile short-shoots may arise) remains against the

shoot. The angles along the shoots combined with spines lend the plants a distinctly cactoid appearance. In species with thicker shoots, such as E. avasmontana, more than half the thickness is occupied by soft, white pith, which is divided up into large inter-connected cavities separated by horizontal layers of tissue well supplied by latex (Pearson 1914: 43). Worsdell (1914) found that these spaces were created by groups of cells dissolving and that the cavities assist in the movement of gases within the shoot. Tubercles bearing the leaves are universally present in sect. Euphorbia. They may be separate and vertically arranged into spirals (usually densely clustered on the shoots, as in E. unicornis and E. unispina (here very low) from tropical Africa), or partially fused into low spiralling rows (as in E. neriifolia from SE Asia). Most commonly they are laterally flattened and vertically fused into often slender and sometimes wavy angles along the shoots, where the individual tubercles are not distinguishable. Their

Fig. 1.22.  Inside of branch showing pale pith with transverse cavities, Euphorbia tetragona, PVB 13531, west of Fort Beaufort, South Africa, 26 Oct. 2018 (© PVB).

Fig. 1.23.  Branches with low tubercles near tips, but becoming almost smooth away from their tips, Euphorbia unispina, PVB 12604, east of Vom, Jos Plateau, Nigeria, 14 Sept. 2013 (© PVB).

1.3 Morphology of Euphorbia (mainly in Southern Africa)

17

Fig. 1.24.  Tubercles on branches arranged into spiralling rows in Euphorbia neriifolia. Left, big shrub nearly 2.5 m tall during dry season, Shirwah, Maharashtra, India, 16 Mar. 2001. Right, stem of young plant with prominent tubercles, Trimbak, Maharashtra, India, Sept. 2019 (© PVB).

Fig. 1.25.  Tubercles fused into prominent, continuous angles along branches, Euphorbia eduardoi, PVB 13410, south of Catengue, Angola, 13 Mar. 2017 (© PVB).

approximate centre is indicated by a (usually) minute leaf which is surrounded by a patch of hard, brown to grey, rigid surface, the ‘spine-shield’ and is subtended by two spines. Two more spines often develop in the stipular position at the bases of the margins of the minute leaf and the upper end of the tubercle is indicated by a small hemispherical axillary bud. In sect. Euphorbia the number of angles starts off at two in tiny plants, corresponding to the two cotyledons. If this two-angled stage persists, a flat-stemmed seedling arises (as in the SE Asian E. antiquorum). Usually as the seedling elongates new angles are introduced with the terminal bud giving rise to three or four tubercles rather than two. This often happens immediately after the first pair of leaves after the cotyledons so that usually within 20 mm of the cotyledons the stem is 3- or 4-angled. Some species have six or more angles on the branches, but this is a relatively rare phenomenon that evolved more than once and is confined to certain lineages. This is usually achieved by, say, whorls of three leaflets or tubercles being shifted to alternate with the previous whorl of three and being joined only to the next whorl up which is in that position and not joined to the whorl immediately above it (Troll 1935–7). This leads to an increase from three angles to six in such a case. In the taller tree-forming species this takes place in the trunk (and occasional forks in it, as in E. grandidens) but mostly not in the branches. Many of these trunks gradually become cylindri-

18

1 Introduction

Fig. 1.26.  New angles on stem beginning by adding a new row of tubercles in sect. Athymalus and sect. Euphorbia. A, stem changing from 4- to 5-angled, ± 40  mm thick, Euphorbia eduardoi, (subg. Euphorbia) PVB 13410, south of Catengue, Angola, 13 Mar. 2017. B, changing from 4- to 8-angled, ± 40  mm thick, Euphorbia polygona, (subg. Athymalus) east of Joubertina, South Africa, 11 Nov. 2018 (© PVB).

cal through swelling of the tissue between the angles (by secondary thickening with wood). They also gradually lose their spines and the green epidermis is progressively replaced with thick grey to brown bark. Angled shoots are not unique to Euphorbia (though they are not found elsewhere in the Euphorbiaceae) and are ­present in stem-succulents in several other families. They are especially widespread in the Cactaceae, where they are only absent in some of the ‘basal lineages’ such as the Pereskioideae and some members of the Opuntioideae. Angled shoots are found in several distinct lineages in the Asclepiadoideae of the Apocynaceae, they also occur in the Geranianceae, Vitaceae and in a few Cucurbitaceae in Madagascar (Troll 1935–7; Rauh 1967; Mauseth 2004b). Generally, they are associated with the increased photosynthetic activity of the shoots replacing that of the reduced leaves. Angles also increase the surface area of the shoots and they add strength to these structures in the larger species, helping them to survive dry periods without collapsing. In succulents in Euphorbia, which lack cortical bundles to transport substances across the cortex (Mauseth 2004b), the angles also reduce the distance between the epidermis and the transporting tissue.

Fig. 1.27.  Bark, which has replaced green, photosynthetic tissue, formed on cylindrical trunk ± 20 cm thick, Euphorbia confinalis ssp. confinalis, PVB 12061, eastern Soutpansberg, South Africa, 4 Nov. 2011 (© PVB).

The surfaces of the branches are mostly smooth but, especially in sect. Articulofruticosae, there are several with variously papillate, ridged, warty or convoluted surfaces. Those of E. muricata and E. verruculosa are shown here. Similarly convoluted surfaces were also observed in E. restituta of subg. Athymalus. All these are species of the arid west coast of southern Africa where fog from the sea adds to the precipitation and these complex surfaces may assist in absorbing this moisture.

1.3 Morphology of Euphorbia (mainly in Southern Africa)

19

Fig. 1.28.  Surfaces of branches viewed with SEM: A, B, Euphorbia muricata, PVB 6026, road to Kalkgat, NE of Vanrhynsdorp, South Africa; C, D, Euphorbia verruculosa, PVB 12542, Lüderitz, Namibia (© PVB).

1.3.3 Spines Often inaccurately referred to as ‘the cacti of the Old World’, Euphorbia has many spiny members. In subg.

Chamaesyce only E. spinea is spiny and here the spines are formed by the branch-tips drying out rapidly to form a hard and sharp spike (E. espinosa and E. guerichiana were also said to become spiny at the branch-tips, e.g. Yang et  al.

Fig. 1.29.  Spikes formed by the tips of the shoots have arisen many times independently in Euphorbia. Left, in E. spinea of subg. Chamaesyce, PVB 13499, just west of Alheit, near Kakamas, South Africa, 23 May 2018. Middle, in E. cuneata of subg. Athymalus, Teita Hills, Kenya, 6 Oct. 2015. Right, in E. lignosa, also of subg. Athymalus, PVB 12836, east of Wlotzkabaken, Namibia, 21 Dec. 2014 (© PVB).

20

2012: 776–7 but this is unknown in both species in habitat and it is not mentioned in either White et al. 1941 or Carter & Leach 2001: 390–1, so it is discounted here). In subg. Esula spines are absent. There are many spiny species in subg. Athymalus and also many in subg. Euphorbia. In subg. Athymalus spines are derived from hardened and sharp shoot-tips or special slender and sharp-tipped shortshoots that arise in the axils of tubercles and are particu-

1 Introduction

larly prominent in subsect. Florispinae. These short-shoots are morphologically different from the other shoots: they are much more slender, non-succulent and taper to a fine tip, they bear only vestigial tubercles and tiny leaf-rudiments and dry out quickly to harden into a rigid spike. As they are morphologically very similar to short-shoots bearing cyathia, they are often referred to as ‘sterile peduncles’ (e.g. White et al. 1941).

Fig. 1.30.  Spikes formed by axillary short-shoots. Left, dense armature of spines, one per leaf-axil, in Euphorbia ferox ssp. ferox, PVB 11946, south of Steytlerville, South Africa, 7 Jun. 2011. Right, more densely clustered spines, several per leaf-axil, in Euphorbia polygona, PVB 12727, Toorwaterpoort, South Africa, 27 May 2014 (© PVB).

The spiny members of subg. Euphorbia are all succulent and they are confined to the Old World in Africa and Madagascar, the Arabian Peninsula and in SE Asia from Pakistan to China. They are typically tropical and subtropical but a few species venture into temperate zones in Morocco, in the Himalaya of SE Asia and in South Africa. In South Africa they are found as far south as 33° around Port Elizabeth and Calitzdorp. In southern Africa all members of subg. Euphorbia are spiny except the four species of sect. Tirucalli (where spines are only known in E. stenoclada from Madagascar and are formed by the rigid tips of the shoots drying into spikes, as in E. spinea). In sect. Euphorbia each leaf is usually surrounded by a hardened, brown to grey ‘spine-shield’ which covers the tip of the tubercle and may extend downwards towards the leaf below. In some species the shields fuse to form a continuous hard margin along the angle, while in others the shields remain separate. Two stout ‘dorsal spines’ are situated slightly behind each leaf on the spine-shield (one spine in the case of some species, such as E. unicornis and E. unispina, more rarely with one spine forked towards its apex, as in E. glochidiata or E. marrupana). Two much smaller gland-like to scale-like or spine-like structures (usually referred to as

prickles) are found on the spine-shield near the base of the margin of each leaf. When they are similar in shape and in size to the dorsal spines, the spine-shield then has four spines surrounding each leaf. The interpretation of the structures on the spine-shield has varied in the literature: Berger (1906: 6) referred to the dorsal spines as thorns or stipular thorns and Troll (1935–7: 900) also called them stipular thorns but Brown (1911: 471; 1915: 223–4) pointed out that, because they are located behind the leaf, they could not be of stipular origin, though he admitted that he was not quite sure what they were; White et al. (1941: 20–21) repeated that their origin was ‘not definitely understood’ and that they were often referred to as ‘stipular outgrowths’. Dyer (1957), not wanting to enter into complex morphological arguments, preferred to refer to them as ‘stipular spines’ and in Dorsey et  al. (2013: 313) they were again referred to a ‘stipular spines’ (see also Uhlarz 1978: 56). Ontogenetic investigations (Shah and Jani 1964; Uhlarz 1974) showed that the smaller structures at the base of the leaf-margins arise when the leaf blade begins to develop in the apical bud (of the shoot) and keep pace with the leaf to reach their full size quickly. Only once these are fully developed do the meristematic zones giving rise to the dorsal spines begin to swell

1.3 Morphology of Euphorbia (mainly in Southern Africa)

21

Fig. 1.31.  Shape and degree of vertical fusion of spine-shields. A, spines-shields fused together into tough, continuous ± uniformly broad margin along the angles (at first pinkish brown hardening to grey), Euphorbia virosa, PVB 13523, north of Pella near Orange River, South Africa, 29 May 2018. B, spine-shields (here almost white) remaining separate along margin of the angles, broader around axillary bud (with red primordia of synflorescences) and around spines, Euphorbia triangularis, PVB 9407, near Bathurst, South Africa, 24 Apr. 2016. C, spine-­shields separate, ± elliptical, spines absent here, from branch out of reach of grazers, on large tree-like species, Euphorbia tetragona, PVB 13531, west of Fort Beaufort, South Africa, 26 Oct. 2018. D, spine-­shields separate, very slender beneath spines and slightly broader around axillary buds and spines, Euphorbia cataractarum, PVB 9599, Zambia (© PVB).

near the dorsal base of the tiny leaf. After the leaf has left the apical bud, the dorsal spines develop, reaching their full size in many cases only after the leaf has dried up and fallen off. Since the smaller structures arise very early in the development of the leaf, as they are positioned near the base of the leaf-margins and since they are present at every leaf on a plant, they are interpreted as stipules. As the dorsal spine or spines arise later and also well behind the ‘stipular region’ at the base of the leaf-margins, they are not part of any stipular ‘complex’ but are considered to be dorsal emergences of the leaf-base (Uhlarz 1974). Unlike stipular structures, these dorsal spines are not constant in their presence or shape on the plant: in several species they are present (often in very robust form) in the young plant but, as it ages, smaller and smaller dorsal spines are produced and eventually there are none below each leaf-base (as in many of the tree-like species, Fig. 1.31); in those species with just one dorsal thorn, this uniqueness may vary on the plant. For example, in the Nigerian E. unispina two dorsal spines are present below each leaf-base in young plants with a gradual transition via lateral fusion to a single dorsal spine per leaf-base as the plant matures (Rauh et  al. 1969: 217, fig. 9; Uhlarz 1974: 43). In the Moçambican E. marrupana, on a single plant some leaf-bases have two dorsal spines, others have only one spine which is forked below its apex and yet others have only one unforked spine (Fig. 6.82 and 6.85, vol. 2). In a few species further small prickles may project alongside the axillary bud. These are most notable in E. grandicornis, where they can reach 7 mm long, but they also occur in E. barnardii, E. restricta, E. pseudocactus and occasionally in E. knobelii. Their origin is uncertain, but it is possible that

Fig. 1.32.  Spines changing from paired (lower right) to solitary (centre) in young seedling ± 2 cm tall of Euphorbia unispina, PVB 12604, east of Vom, Jos Plateau, Nigeria, 14 Sept. 2013 (© PVB).

the small bracts that initially shelter the axillary bud may each have a single dorsal spine-like outgrowth or may have stipular prickles. In sect. Monadenium tiny prickles are present as rudimentary stipules alongside the bases of the leaf-margins in E. lugardiae (Uhlarz 1978: 52). In certain species from tropical Africa these prickles may be much larger. There may be also a dorsal spine below the leaf as well as complex spiny outgrowths raised on papillae in a row below between the leaves. Bally (1961: 63) referred to the spines in such species as Euphorbia guentheri (M. guentheri) and E. neospinescens (M. spinescens) as ‘three sharp prickles’, so did not distinguish the dorsal spines from the stipular prickles. Uhlarz

22

1 Introduction

Fig. 1.33.  Spines around the leaves in sect. Monadenium. A, leaves with small flattened stipular scales and no spines (branch ± 10 mm thick), Euphorbia bisellenbeckii, PVB 12763a, N. Kenya. B, each leaf-­scar with single spine below it (branch ± 10 mm thick), Euphorbia torrei, PVB 9701, north of Nantulo, northern Moçambique. C, each leaf-scar with three spines, one below and two alongside scar (branch ± 10 mm thick), Euphorbia biselegans, PVB 8733, Tanzania. D, each leaf-scar with several spine-tipped teeth in row below it and sometimes somewhat fused, with pair of much smaller irregular stipular spines alongside it (branch ± 20 mm thick), Euphorbia magnifica, PVB 9647, south of Mpwapwa, Tanzania (© PVB).

(1975b) showed that the spines below the leaf in sect. Monadenium are homologous to the dorsal spines in sect. Euphorbia. The presence of dorsal spines that are not stipular in both sect. Euphorbia and sect. Monadenium corroborates the close relationship between these two groups that was later revealed by DNA-data. This shows also that, even morphologically, sect. Euphorbia is not as isolated within Euphorbia as Gilbert (1995) suggested.

1.3.4 Leaves In Euphorbia, leaves are generally prominent. However, in southern Africa leafy plants such as in E. transvaalensis are

Fig. 1.34.  Leafy shrub of Euphorbia transvaalensis, PVB 12089, south of Ellisras, South Africa, 28 Dec. 2011 (© PVB).

the exception: only eight species of sect. Esula and 14 members of subg. Chamaesyce are herbaceous. The remaining species are more or less succulent and in most of them the leaves are reduced to minute rudiments. Succulent leaves are uncommon generally in Euphorbia except in subg. Euphorbia where they are usually combined with fleshy green shoots so that photosynthesis takes place both in the shoots and in the leaves (a rare phenomenon among succulents generally, Mauseth 2004b). Some Madagascan species such as E. elliotii (sect. Denisophorbia) bear succulent leaves, but these are evergreen and the shoots are not photosynthetic. Succulent leaves are typical of sect. Monadenium (as in E. lugardiae) and may reach 150 mm long or more (even larger in E. neoarborescens), but they are always deciduous. Members of sect. Euphorbia also have succulent leaves, but these are mostly reduced to tiny rudiments that are only visible around the terminal bud when it is actively growing. Nevertheless, prominent, fleshy, deciduous leaves occur in SE Asian species such as E. caducifolia and E. nivulia (also in the closely allied geophytic E. acaulis and its relatives), in several West African species and in the south tropical African geophytes related to E. decidua (Leach 1976b). Euphorbia ampliphylla from tropical Africa bears prominent, fleshy, deciduous leaves 10  cm long or more and its close relative in southern Africa, E. ingens, also has prominent leaves on young plants, though these rarely exceed 8 cm long. However, in E. ingens (unlike in E. ampliphylla), the leaves decrease in size as the plant ages and in mature trees they are reduced to minute, scale-like rudiments on the new shoots.

1.3 Morphology of Euphorbia (mainly in Southern Africa)

23

1.3.5 Stipules Stipular structures in Euphorbia may take the form of thorns or small rudiments or may be absent (Uhlarz 1974; 1975a; 1978) and are not elaborate as in some members of Jatropha. In Euphorbia they are of irregular occurrence among groups of related species: a good example is sect. Articulofruticosae, where fairly prominent gland-like stipules are present only in E. burmanni (examined in detail by Uhlarz 1975a), E. exilis and E. suffulta out of the 15 species in this section, while in sect. Tirucalli they are found in three of the four species in southern Africa (being absent in E. gregaria). Nevertheless, in species where they are found, their occurrence is very regular. Stipules are absent in all members of subg. Esula (Riina et al. 2013), they are occasional in subg. Athymalus and subg. Chamaesyce in southern Africa and are almost always present (though small but variable in shape in different species) in subg. Euphorbia among the structures around the leaf-rudiments on the spine-shields.

Fig. 1.35.  The prominent succulent deciduous leaves in the shade-­ loving Euphorbia neocannellii, PVB 13413, east of Catengue, Angola, 13 Mar. 2017 (© PVB).

Fig. 1.36.  Leaf-rudiments on growing branches in members of sect. Euphorbia. A, leaflets ± 5 mm broad around the terminal bud in Euphorbia radyeri, PVB 10539, 3 km towards Matjiesvlei, west of Calitzdorp, South Africa, 5 Oct. 2008. B, ± 4 mm broad in Euphorbia triangularis, PVB 9407, near Bathurst, South Africa, 24 Apr. 2016. C, small scale-like leaflets ± 2 mm long in Euphorbia grandidens, PVB 6870a, Springs, NE of Uitenhage, South Africa, 16 Jul. 2018 (© PVB).

24

1 Introduction

Fig. 1.37.  Stipular structures and spines associated with leaf-rudiments in sect. Euphorbia. L = leaf-rudiment, D = dorsal spine, S, stipular structure (irregular and scale-like or spine-like), A = additional spines around axillary bud. A, (scale 2 mm, as for B) and B, Euphorbia confinalis subsp. confinalis, PVB 11764, Komatipoort, South Africa. C, (scale 4 mm), Euphorbia grandicornis subsp. grandicornis, PVB 4467, near Ndumu, South Africa. D, (scale 2 mm, as for E) and E, Euphorbia schinzii subsp. schinzii, PVB 9390, Kaalrug, east of Nelspruit, South Africa. F, (scale 4 mm) and G, (scale 1 mm), Euphorbia triangularis, PVB 12157, between Kirkwood and Enon, South Africa (© PVB).

1.3.6 The Organization of Flowers Mature specimens of Euphorbia (and all members of the Euphorbieae) produce flower-like ‘units’ in which five groups of anthers are clustered densely around a single ovary + style and are tightly enclosed in a cup-like container with (mostly) five differently coloured, spreading glands on its exterior. This ‘unit’ is known as a cyathium. The huge diversity and vast distribution of Euphorbia suggest that the cyathium has been an ‘astounding performer’ as a reproductive device (Croizat 1937: 525) and is one of the key innovations leading to the diversity in Euphorbia. A. Internal organization of the cyathium. Linnaeus (1753) believed that the cyathium in Euphorbia was a bisexual flower, but Lamarck (1786) suggested rather that it was a false flower or ‘pseudanthium’ derived from a greatly contracted inflorescence. This theory was investigated further by R. Brown (1814), Röper (1824) and others and, more recently, by Hoppe & Uhlarz (1982) using the Scanning Electron Microscope. They showed that the cyathium is derived from an inflorescence by the fusion of the bases of the early stages of five bracts into a cup-like structure called an involucre. In the mature cyathium only the tips of these bracts remain visible as the so-called ‘lobes’ between the ‘glands’. The involucre surrounds five c­ yme-­like groups or ‘partial inflorescences’ of male florets. Each male

floret consists of a pedicel and a stamen and there is usually a filiform bracteole associated with it as well. The male florets surround a single terminal female floret which consists of a pedicel and a gynoecium. The pseudanthial theory of the cyathium received support from interpretation of the stamens as individual but greatly reduced male flowers and the ovary + style as a greatly reduced female flower. In the males this is particularly based on the presence of a distinct join along their stalks (suggesting where the perianth was lost and that the part below the join is a pedicel and that above it is a filament) and in the females by the very occasional presence of a rudimentary ‘calyx’ just beneath the ovary. The lack of a perianth on the male florets separates Euphorbia from the other Euphorbieae. Prenner & Rudall (2007) and Prenner et  al. (2008) have shown that there are problems with this interpretation and suggested that the cyathium may be a ‘hybrid’ structure combining a terminal female flower with male partial inflorescences. They found that the abscission zone separating the ‘male florets’ from their ‘pedicels’ arose well after the development of the anthers, that the rudimentary ‘calyx’ also arose after the ovary was already quite well developed and that it might be made up of bracts rather than tepals. Within a bisexual cyathium, the female floret matures first. In all species it becomes receptive in an erect position. In some species its pedicel then elongates further, bends downwards to project horizontally from the cyathium between the glands. This pushes the ovary well beyond the

Fig. 1.38.  Parts of the cyathium. A, cyathium from side (scale 2 mm, as for B). B, cyathium from above. C, dissected cyathium from side (scale 1 mm). Euphorbia braunsii, PVB 11363, SE of Warmbad, Namibia (© PVB).

Fig. 1.39.  Ovaries with ‘calyx’ in several species of Euphorbia. A, Euphorbia confinalis subsp. confinalis, without any visible ‘calyx’, PVB 11764, Komatipoort, South Africa (scale 1 mm, as for B–E). B, Euphorbia celata, with small protrusions of ‘calyx’ around base of ovary, showing also cutaway involucre, filiform bracteole and small, rudiments of male florets, PVB 4638, north of Lekkersing, South Africa. C, Euphorbia radyeri, ‘calyx’ small but present, PVB, Jansenville, South Africa. D, Euphorbia lividiflora, with more prominent ‘calyx’, PVB 7684, south of Mambone, Save River, Moçambique. E, Euphorbia ingens, ‘calyx’ consisting of several filiform parts equalling the ovaries, ex hort. NBG, South Africa (© PVB).

26

cyathium and minimizes the chances of self-fertilization. The stigmas are usually receptive for up to a week or more, during which they are sticky and widely spread out. Once receptiveness is over, they become red or brownish and dry out. The males mature after the female, ripening from the centre of the cyathium towards the circumference. The male florets elongate to project from the cyathium, dehisce to release their pollen and fall off within a few hours. In southern Africa the male florets have glabrous filaments and their pedicels are often pubescent towards the top

1 Introduction

in subg. Athymalus and Chamaesyce (though never in subg. Esula or sect. Euphorbia). Pubescent filaments are rare in Euphorbia generally (Prenner and Rudall 2007) and are known only in a few American succulents belonging to sect. Crepidaria. The female pedicel is occasionally pubescent and this may be true of the ovary too. The outside of the involucre is ornamented with 1–5 spreading ‘cyathial glands’ and these may further be ornamented by petal-like appendages (as in the Mexican E. antisyphilitica and many other New World species, some

Fig. 1.40.  Shapes of cyathial glands. A, Euphorbia quadrata, PVB 4044, Cornellsberg, NW of Eksteenfontein, South Africa (scale 3  mm). B, Euphorbia polycephala, PVB11083, west of Cradock, South Africa (scale 3 mm). C, Euphorbia monteiroi subsp. ramosa, Gildenhuys 657, SW of Steelpoort, South Africa (scale 2 mm). D, Euphorbia dumosoides, Grieve 2428, Umtamvuna Reserve, near Port Edward, South Africa (scale 1 mm). E, Euphorbia guerichiana, PVB 3297, foot of Oemsberg, north of Eksteenfontein, South Africa (scale 1 mm) (© PVB).

of which produce the closest mimics of a true flower in Euphorbia). Roeper (1824) suggested that each gland is made up by fusion from two stipules. Detailed investigation by Hoppe & Uhlarz (1982) and Hoppe (1985) showed that the cyathial glands are indeed derived from fusion of the pairs of interpetiolar stipules from adjacent bracts that formed the involucre. This is also confirmed by their vascular supply, which shows clearly that each gland is linked to the vascular system of the two cyathial lobes that flank it (Genç and Rauh 1984: Fig. 1 III). As they mature, these fused stipules lose all outer indications of their separate origins and develop a massive flattened apex where secretory tissue is lodged, forming the principal secretory part of the gland. Hoppe (1985) found that the outer edge of the gland is always smooth and that all outer appendages of the glands develop from meristems that lie just beneath this smooth edge. This applies to the petaloid appendages

of American species like E. antisyphilitica, to the teeth on the outer edge of the gland that are common in subg. Athymalus and to the two dorsal horns often found in subg. Esula. Secondary secretory zones on the appendages of the glands appear to be common but are not easily detected (Genç and Rauh 1984; Schnepf and Deichgräber 1984). In sect. Monadenium the usual five glands are apparently reduced to two. The larger of these has an outline like a horseshoe, with a secretory zone around its apex. This enlarged gland is derived from fusion of four of the five glands which are present in very early stages of the cyathium (Hoppe 1985). B. Arrangement of cyathia into ‘cymes’ and ‘synflorescences’ on ‘flowering shoots’. In southern Africa, cyathia in Euphorbia may either terminate shoots or they may arise in the axils of tubercles or

1.3 Morphology of Euphorbia (mainly in Southern Africa)

leaves. Those that arise in axils are borne on short-shoots that are mostly morphologically distinct from the other shoots, but even a vegetative shoot terminated by a cyathium may change morphologically just below the cyathium.

27

others) arise when additional cyathia repeatedly develop on peduncles from the axils of bracts just beneath the first, terminal one. In subg. Euphorbia (in southern Africa) the synflorescences arise in the axils of tubercles near the tips of the branches (usually on growth from the previous growing season). One to three (rarely more than five) short-shoots or ‘cymes’ develop in each axil. In sect. Euphorbia each cyme consists of a short peduncle (occasionally up to 30  mm long, as in E. eduardoi) with a single internode and a pair

Fig. 1.41.  Few cyathia, one of which terminates the branch, Euphorbia gariepina, PVB 11115, 10 km east of Jakkalswater, South Africa (© PVB).

In subg. Athymalus many species (such as E. lignosa and E. polycephala) have solitary cyathia or small groups that terminate some of the branches (though not the stem). In many other species the cyathia develop at the tips of special short-shoots arising in the axils of the tubercles on the branches (or the stem in unbranched species such as E. obesa) that are morphologically different from the branches and stem (usually more slender with reduced tubercles and leaves). So they again terminate certain shoots, though this time not the branches themselves. In most members of ser. Meleuphorbia these short-shoots differ only from the spines by the terminal cyathium. More complex ‘synflorescences’ (as in E. bubalina, E. globosa, E. monteiroi and

Fig. 1.42.  Relatively long inflorescences in Euphorbia eduardoi. South of Iona, Angola, 8 Jan. 2007 (© PVB).

of bracts at the apex subtending a male cyathium (occasionally bisexual, as in E. vandemerwei). Two further bisexual cyathia develop from the axils of these bracts and it is ­uncommon for more to be produced (though this occasionally happens in E. excelsa, E. ingens and E. waterbergensis). In many species (especially those with thicker shoots) the minute, undeveloped cymes are already present under

1 Introduction

28

Fig. 1.43.  Inflorescences covered by bracts. Left: only one cyme per axil, Euphorbia zoutpansbergensis, PVB 7099, above Soutpan, Soutpansberg, South Africa. Right: three cymes per axil, Euphorbia avasmontana, PVB 12872, 10 km west of Outjo, Namibia (© PVB).

protective bracts in the axils in the season before they mature. In sect. Euphorbia the lateral cyathia arise in two essentially different orientations relative to the central cyathium: those where they and the central cyathium are arranged in a line perpendicular to the axis of the branch (referred to as ‘transversely disposed’ and sometimes as ‘horizontally disposed’) and those where they are all arranged in a line in the direction of the axis of the branch (or ‘vertically disposed’). This difference in their orientation seems to have been used first in White et al. (1941) and was illustrated clearly by Dyer (1957). Examples are shown in Fig.  1.44. In sect. Tirucalli each cyme bears only one cyathium. In sect. Monadenium the peduncle of the cyme may reach 100 mm long or more and the cyme can be richly branched to bear many cyathia, though E. lugardiae is similar to sect. Euphorbia and the cyme consists of just three bisexual cyathia on a short peduncle. In subg. Esula many of the shoots are terminated by a cyathium. After this has matured, more cyathia develop from the axils of bracts just beneath it to form a rosette around it. This process can be repeated in herbaceous species to produce very complex ‘synflorescences’ with large numbers of bracts and cyathia. In the succulent species of subsect. Africanae the ‘synflorescences’ are less prolific: the terminal cyathium is usually male and a single whorl of bisexual cyathia develops from the axils of bracts immediately beneath it, each of these on a short peduncle, often referred to as a ‘ray’. In subg. Chamaesyce in southern Africa many branches are also terminated by a cyathium, with others developing from the axils of bracts around it. In species of sect. Tenellae

Fig. 1.44.  Disposition of the lateral cyathia in the cymes relative to the central cyathia and the axis of the branch in sect. Euphorbia. A, transversely disposed, with lateral cyathia not yet mature, E. louwii, PVB 12051, south of Marken, South Africa, Aug. 2019. B, transversely disposed, with lateral cyathia in final male stage and central cyathium already dried up, E. barnardii, PVB 11896, west of Steelpoort, South Africa, Aug. 2011. C, vertically disposed, with three cymes arising in each axil, each with the lateral cyathia before female stage and central cyathium full of anthers, E. avasmontana, PVB 10063, east of Namusberg, Rosh Pinah, Namibia, Nov. 2011 (© PVB).

and sect. Anisophyllum the stem usually aborts growth after a few nodes above the cotyledons and growth continues from shoots arising in the axils of the leaves. The whole plant resembles the richly branched ‘synflorescence’ of subg. Esula, with most of the ‘leaves’ actually bracts, small cyathia terminating the cymes and others arising in the axils of these bracts (Webster 1967: 420).

1.3.7 Fruit, Seed and Seedlings In Euphorbia the fruit is a capsule usually made up of three locules, where each locule contains one seed. In southern

1.3 Morphology of Euphorbia (mainly in Southern Africa)

Fig. 1.45.  Capsules on Euphorbia fasciculata, PVB 11422, northern Knersvlakte, Namaqualand, South Africa, 25 Sept. 2012 (© PVB).

29

Fig. 1.46.  Triangular capsules in Euphorbia stellata ssp. micracantha, PVB 11920, SE of Cradock, South Africa, 16 Oct. 2012 (© PVB).

Fig. 1.47.  Spherical capsules: in Euphorbia stellispina of subg. Athymalus, PVB 11102, NE of Laingsburg, South Africa, 2 Aug. 2011; in Euphorbia virosa of subg. Euphorbia, ± 20 km south of Vioolsdrift, South Africa, 26 Dec. 2005 (© PVB).

Africa only E. damarana, E. gregaria, E. gummifera and E. virosa regularly have four to six locules per capsule (and then with 4–6 seeds in the capsule). The capsule is formed from laterally and centrally fused carpels around a slender central column, with a shallow groove along the middle of each carpel on the outside. It is mostly obtusely three-­ angled but occasionally the angles are more prominent (as in E. knuthii or E. stellata) or it may be spherical (as in E. stellispina and E. virosa). The capsules vary greatly in size across the genus. They are particularly small in many of the tiny-flowered, annual members of subg. Chamaesyce. They

are often relatively small in the tall succulent trees but among them there is a close correlation between the diameter of the cyathium and the size of the capsule. The largest capsules are in species such as E. scheffleri from East Africa. In the Madagascan E. antso of subg. Athymalus the capsules reach 20–35  mm in diameter (despite the fairly small cyathia at 4–6  mm diam.) and 40  mm in sect. Pachysanthae (of subg. Euphorbia), also in Madagascar. This is slightly larger than any found in southern Africa, where they reach a diameter of 26 mm in E. gregaria and 30 mm in E. damarana.

30

1 Introduction

Capsules of Euphorbia mature fairly quickly (usually within a month, though up to 3  months in E. lugardiae) after fertilization. In species where the capsule is ± sessile and erect, it dehisces in this position. Where the capsule is held on a longer, decurved pedicel, it matures in this position, after which the pedicel becomes erect and then the capsule dries out rapidly to dehisce in this erect position. Capsules in Euphorbia are schizocarps (i.e. a fruit that is dry when ripe and breaks into single-seeded parts) and most of them dehisce explosively. As the capsule dries out, tensions mount along the groove along the middle of each locule and the locules also begin to separate from each other. These tensions are suddenly released when each locule violently breaks into two down the groove along its

Fig. 1.48.  Large capsule (here ± 3 cm diam., later drying smaller and dehiscent) in Euphorbia scheffleri, PVB 13611, Mto wa Mbu, west of Arusha, Tanzania, 30 Jan. 2019 (© PVB).

Fig. 1.49. Large pendulous indehiscent capsules in Euphorbia gregaria, PVB 13501, near Maraisvlei, west of Kakamas, South Africa, 5 Jun. 2018 (© PVB).

Fig. 1.50.  Capsules about to dehisce. Left: some erect and about to dehisce, others still pendulous, Euphorbia stellata ssp. stellata, PVB 10983, north of Carlisle Bridge, Eastern Cape, South Africa, 11 Jan. 2008. Right: Capsules held erect and beginning to dry out just before dehiscing, Euphorbia mlanjeana, PVB 9739, Mount Namuli, Moçambique (© PVB).

1.3 Morphology of Euphorbia (mainly in Southern Africa)

Fig. 1.51.  One capsule still maturing and pendulous, other pedicel erect after dehiscence of its capsule (and one fertile cyathium), Euphorbia herrei, PVB 3943, Bloeddrift, east of Alexander Bay, South Africa (© PVB).

31

Fig. 1.52.  Capsule drying out (here ± 8 mm diam.), with the locules beginning to separate before explosive dehiscence in Euphorbia platycephala, PVB 6966, north of Pandamatenga, Botswana, 30 Jan. 2019 (© PVB).

Fig. 1.53.  Capsules all gone, with pedicels remaining in erect position. Left: Euphorbia stellata ssp. stellata, PVB 10983, north of Carlisle Bridge, Eastern Cape, South Africa, 11 Jan. 2008. Right: Tips of shoots of female plant with central columns of capsules remaining on pedicels after release, Euphorbia heptagona, Klein Waterval, NE of Laingsburg, South Africa, 19 Jun. 1989 (© PVB).

32

1 Introduction

Fig. 1.54.  Predation on shoot-tips of Euphorbia pulvinata which may lead to dispersal of the seeds, PVB 11056, west of Witmos, south of Cradock, South Africa, 4 Aug. 2011 (© PVB).

middle and then the locular walls and the seeds are flung away, usually only leaving the central column of the capsule intact on the pedicel. In a few species of sect. Euphorbia (E. ingens and E. virosa) and certain species of sect. Tirucalli such as E. carunculifera, E. damarana, E. gregaria and E. gummifera (but not E. neochamaeclada, where seeds are released explosively), the capsule is not explosive, but dries out and falls to the ground. In the process of falling down, it usually breaks up and releases the seeds, though sometimes the capsule remains intact and becomes buried (possibly by rodents) and the seeds germinate in a small dense clump. On a hot day among plants bearing mature capsules, the violent dehiscence of the capsules can be heard clearly. These little explosions may fling the seeds to between one and ten metres from the parent-plant. Other methods of ­dispersal also play a role. For example, it has often been observed that the apices of the shoots are chewed off, even in such spiky species as E. ferox, E. pulvinata and E. stellispina, in the very acrid E. damarana and E. mauritanica and even in the supposedly deadly E. virosa. If this happens when capsules are almost ripe, they will be transported in the gut of the animal and, after passing through its digestive tract, dumped possibly many kilometres from the parent. Animals responsible for this are various buck, especially kudu and eland but also smaller species such as steenbok and duiker. The black rhinoceros, too, relishes such species as E. damarana and E. virosa (Loutit et al. 1987; Jacobsen 1988: 62). These larger mammals may have been instrumental in the wide and scattered occurrence of populations of such species as E. pulvinata and E. stellispina. In addition, the seeds of members of ser. Meleuphorbia are sticky and some

of them could adhere to the feet of passing animals and be moved further away from the parent. A further aid in dispersal is the caruncle on seeds of members of subg. Esula, as well as a few other species in subg. Chamaesyce and sect. Tirucalli. They are very attractive to ants which will even chew holes into paper packets to remove carunculate seeds stored there. These seeds are taken into their burrows, where they germinate when moistened. It has also been noted (Leach 1964: 212) that doves have a great liking for seed of Euphorbia once they are scattered on the ground and doves have been known also to peck the capsules off plants of E. obesa in cultivation (E.  Heunis, personal communication). These and other birds may also contribute to their dispersal in nature.

Fig. 1.55.  Predation on shoots of Euphorbia damarana. PVB 11262, West of Usakos, Namibia, 6 Nov. 2008. Here a brush-like cluster of fibrous material is left behind (© PVB).

Fig. 1.56.  Fruits and seeds of Euphorbia virosa. Left: dried-out capsules and seeds, PVB 12832, 5 km south of Solitaire, Namibia (4 larger seeds from cultivated material in Windhoek: scale 3 cm, one subdivided). Here we see five of the parchment-like dry capsules that are easily broken up by handling. Those in face view vary from three- to five-locular (and may have even more locules). The hard, nut-like seeds also vary greatly in size and colour. The four largest ones are possibly the largest found in any species of Euphorbia. Right: opened seed, PVB 12837, Khorixas, Namibia. Embryo on left, endosperm in middle and seed-shell on right (scale 2 cm, one subdivided) (© PVB).

Fig. 1.57.  Examples of seed in southern African Euphorbia (the largest seeds, found in E. virosa, are shown in Fig.  1.56), all to roughly the same scale and, in each, the scale bar represents 1 mm. A–D, subg. Athymalus. A, E. oxystegia, PVB 6342, shiny, smooth and attractively mottled. B, E. ferox subsp. ferox, PVB 6311a, slightly shiny, sticky with some adherent dust. C, E. schoenlandii, PVB 6135, black and roughly papillate. D, E. hallii, PVB 896, brown and very roughly papillate. E–F, subg. Chamaesyce. E. E. exilis, PVB 6132, finely papillate, from back and front. F, E. guerichiana, PVB 13566a, smooth, with cap-like caruncle on top. G–H, subg. Esula. G, E. genistoides, slopes of Constantiaberg, Cape Town, with small apical carumcle. E. mauritanica, PVB 11351, with small, cap-like caruncle. I–L, subg. Euphorbia. Sect. Tirucalli, I, E. neochamaeclada, PVB 10602a, Angola, narrow and smooth, with relatively broad caruncle (as in E. congestiflora). Sect. Monadenium, J, E. lugardiae, PVB 11899, narrow, angled, slightly rugulose, with relatively broad caruncle. Sect. Euphorbia, K, E. grandidens, PVB 6870a, smooth, nearly black. L, E. kaokoensis, PVB 8057, warty, pale brown and small, from back and front (© PVB).

34

1 Introduction

Fig. 1.58.  Seedlings of Euphorbia. A, E. burmanni (scale 2  mm, as for B–E). B, E. schoenlandii. C, E. tubiglans. D, E. pseudoglobosa. E, E. susannae. F, E. adjurana (scale 4 mm). Drawn from: A, between Stormsvlei and Robertson, South Africa; B, PVB 6135, south of Kotzesrus, South Africa. C, PVB 4913, Steytlerville, South Africa. D, PVB 4727, north of Bredasdorp, South Africa. E, PVB 945, near Muiskraal, south of Ladismith, South Africa. F, PVB 11600, north of Mega, Ethiopia (© PVB).

Seeds of Euphorbia vary with the size of the capsules. While most are 2–3  mm long and  ±  1.5  mm broad, they range from about 0.8 mm long in some small members of subg. Chamaesyce to 6–13 mm diameter and almost spherical in E. virosa and 10 mm in diameter in E. antso of subg. Athymalus. The surfaces of the seed vary greatly, even among closely related species, from smooth and shiny to dull and finely papillate or warty (Fig. 1.57). They are may be attractively mottled with dark brown on grey, perhaps as camouflage to prevent predation (as in some bird-eggs) by animals such as birds or rodents that may crush and consume them. The warty, hard surface is likely to allow them to be carried without internal damage, especially by ants or rodents, for storage in their burrows where germination may take place. The seeds are spherical to slightly longer than broad. One end is narrower (the ‘micropylar’ end, with a small but distinct scar next to it where it was joined as an ovule to the ovary), while the other is more truncate or blunt and a fine line runs up one side between these ends. In Euphorbia each ovule is pendulous in the ovary and after fertilization the ovule swells towards the base more than near this join so that the narrow end of the seed faces the styles ontop of the capsule (Corner 1976). The longitudinal line along the seed is closest to the capsule’s central column. The mature seed has a hard, dry and brittle, usually brownish to grey outer coat (or ‘testa’) which reaches more

than 1.5 mm thick in parts in E. virosa (see Fig. 1.56), though it is mostly much thinner and even in E. virosa it is much thinner towards the micropylar end. Inside the testa is a thick, cream-coloured layer of endosperm (containing food for the germinating seedling) and embedded in this is the usually cream-coloured, straight embryo with the hypocotyl pointing towards the micropylar end of the seed (and upwards in the capsule). When the seed germinates the radicle breaks through the seed-coat near the micropylar end where the testa is thinnest. Seedlings of non-succulent species have a slender hypocotyl and comparatively large cotyledons. As can be seen in Fig. 1.58, in the succulent species the newly germinated seedling rapidly develops a swollen photosynthetic hypocotyl, which is vital for the new plant to establish itself (Von Willert et  al. 1992: 51), while the cotyledons are reduced and less leaf-like, to reduce water-loss though transpiration. The degree of succulence of the hypocotyl corresponds roughly to the degree of reduction of the mature plant, as Fig. 1.58b–e shows. The epidermis of the stem of seedlings of many members of sect. Euphorbia has a very different colouring to that in mature plants. This usually consists of pale mottling against a darker green background, presumably helping to camouflage the young plant. In the larger species (such as E. eduardoi or E. ingens) it lasts until the stem is around 1 m tall, after which the epidermis becomes a more drab green before bark develops.

1.5 Pollination Biology

1.4

Chromosome Numbers

Although chromosome numbers have proved useful in the classification of many families, especially when combined with morphological characters (Raven 1975), they are not widely available for Euphorbia. Hans (1973) listed 116 cytologically known species of Euphorbia. Seven and ten were considered the primary base numbers for Euphorbia and other numbers (such as n = 6, 8, 9) were believed to be derived from these. More counts have been done subsequently and the Chromosome Count Data Base (www.ccdb.tau.ac.il), gives counts for some 500 species of Euphorbia, most of them from Europe and North America.

35

The southern African species are remarkably poorly represented, with only 16 species sampled so far. Of these 13 belonged to E. subg. Athymalus with a single species from each of the other subgenera (E. inaequilatera (n  =  10) of subg. Chamaesyce, E. lugardiae (n = 17) of subg. Euphorbia, E. mauritanica (n  =  19) of subg. Esula). Of the 13 from subg. Athymalus all exhibited n  =  10 and this was also recorded for E. balsamifera from outside our area. Exceptions are E. ferox (n = ± 100) and E. clandestina (n = 17). The latter is doubtful, since it was obtained from material cultivated in a nursery in Sikkim, India (Srivastava et al. 1987), so that one may speculate that n = 10 is typical for subg. Athymalus. Most of the counts for species in subg. Athymalus were published by Vosa & Bassi (1991). The succulent species of subg. Euphorbia are especially poorly known. Here the only investigation was of members of E. sect. Monadenium (at that time belonging to the genera Monadenium and Synadenium) by Jones & Smith (1969). This was recommended by Bally (1961: 15) and most of these counts were made from material that he introduced into cultivation. It revealed a wide spread of values between n = 12 and n = 19, with many species, including E. lugardiae, on n = 17. These fairly high numbers relative to the base numbers of seven and ten were considered (along with the differences in the cyathia) to indicate a derived position for sect. Monadenium, an opinion that has not been corroborated by evidence from DNA-data, which places sect. Monadenium as one of the earlier branches in subg. Euphorbia.

1.5

Pollination Biology

1.5.1 Flowering Times

Fig. 1.59.  Seedling of Euphorbia eduardoi (± 12 cm tall) with brightly mottled epidermis (© PVB).

As far as they are known, flowering times are given for each species under ‘Diagnostic features and Relationships’ and at almost any time of the year some species are in flower. However, flowering depends on local conditions and may vary from year to year. With certain species (such as E. decepta and E. obesa), flowering can be induced several times during the summer months if water is withheld until plants are dessicated, after which they are watered again. Others, such as E. pseudoglobosa and E. susannae, always adhere to their autumnal flowering-time, whether they are dry or well-watered; E. gerstneriana likewise mainly flowers in spring as growth of new branches begins. In many species in sect. Florispinae, sect. Africanae and sect. Euphorbia, cyathia mature en masse over the whole of

36

1 Introduction

the top of the plant. This may occur over whole populations which then ‘flower’ almost simultaneously for a short time, producing a spectacular show of colour and making the air sweetly scented around the population (as in the sometimes spectacular flowering events of E. avasmontana or E. radyeri or the much more modest cases of E. colliculina or E. esculenta). In other cases, only a few plants in a population may flower, as often happens in E. eduardoi.

Fig. 1.60.  Plant with cyathia covering tips of each branch in Euphorbia ferox ssp. ferox, PVB 11919, near Kommandodrift, SE of Cradock, South Africa (© PVB).

Fig. 1.61.  Some plants covered with yellow cyathia, others flowering more sparingly, Euphorbia mauritanica, Kliprand, South Africa, 12 Jul. 2008 (© PVB).

Fig. 1.62.  Upper part of profusely flowering tree yellow with mature cyathia, Euphorbia tetragona, PVB 11915, Witmos, south of Cradock, South Africa, 26 Oct. 2012 (© PVB).

Fig. 1.63. Tips of branches covered with bright yellow cyathia, Euphorbia radyeri, Welgelegen, north of Jansenville, South Africa, 23 Oct. 2009 (© PVB).

1.6 Biogeography and Ecology

37

1.5.2 The Mechanics of Pollination

(a) Scent. Scent given off by the cyathium appears to emanate from the upper surface of the cyathial glands. Scents Removal of pollen from the anthers is achieved during moveare very variable across Euphorbia. They may be uniment of an insect over the top of the cyathium when pollen is form across some sections and variable across others. brushed off the anthers and adheres to the body of the insect. For example, many of the smaller members of sect. Pollen is transferred usually when the recurved parts of the Euphorbia in southern Africa have bright yellow cyathia stigmas wipe it off the body of the insect. that may be sweetly scented. In sect. Medusea a wide While certain weedy species of sect. Anisophyllum in the range occurs: cyathia with yellow glands are generally USA are suspected to be self-fertile (Webster 1967: 424), sweetly scented (E. flanaganii, E. huttoniae), green-­ among southern African species self-fertility has been glanded ones are sweet to sourish (E. procumbens) or observed in E. herrei and perhaps also in E. grandidens, but emit a goat-like smell in E. tuberosa (Vogel 1954), while is otherwise unrecorded. In most species with solitary (or those with purple glands may be sweet (E. esculenta) or well-separated) cyathia, self-fertilization is apparently distinctly foetid as in some E. procumbens. Euphorbia avoided by the male florets in the cyathium maturing after polycephala with white glands gives off an unusual the female floret, probably after it is no longer receptive. scent, which is neither sweet nor foetid. The particularly Where the female is borne on a fairly long pedicel (as in E. small cyathia of E. sect. Articulofruticosae mostly have knuthii, E. patula and others) the possibility of self-­ green glands and seem to be scentless. fertilization is further reduced by the pedicel bending out of (b) Nectar. The cyathial glands in Euphorbia are well-­ the cyathium, so that it is out of range as the males mature known to secrete nectar. This is released from pits on the and pollen becomes available. upper surface of the gland, where the peripheral cells In those species where many cyathia are produced close often have notably thin outer walls (Genç and Rauh together on each branch, other mechanisms to prevent self-­ 1984; Schnepf and Deichgräber 1984). Sometimes the pollination appear to operate. In some (such as members of glands contribute to producing the nectar but many seem ser. Meleuphorbia), the males and females are on separate to secrete products from elsewhere in the plant. Although plants. In others (as in sect. Euphorbia) each cyme of very sweet, the nectar is often somewhat irritant and this cyathia progresses through three distinct stages, first a male is carried over to honey from Euphorbia (e.g. E. radystage in the central cyathium (which is mostly male only), eri), which produces a burning sensation in the throat then a female stage (the initial stage for each of the lateral when eaten. Investigations by Hecker (1981) of honey cyathia in each cyme) and then another male stage (again in from ‘E. caerulescens’ (more likely to be E. radyeri) the lateral cyathia) probably after the females are no longer showed that the irritant elements were various mildly receptive. These stages seem to be closely co-ordinated, so poisonous diterpine esters similar to those that make the that on each plant most of the cyathia are at a given stage at sap caustic (see below). the same time.

1.5.3 Pollinators In the larger species, cyathia mature roughly simultaneously over much of the tree or shrub and the plant is then visited by quite noisy clouds of insects for several days until flowering is over. Flying visitors include bees, beetles, flies and wasps of many sizes and these are accompanied on foot by many different ants, so it appears that pollination is generally an unspecialised process in Euphorbia (as suggested by Webster 1967: 408). Struck (1992) observed 10 species of bee, beetle, fly and wasp on Euphorbia rhombifolia and 13 on E. mauritanica in Goegab Nature Reserve, east of Springbok, Namaqualand during their  ±  one month-long flowering period. Although the selection seems broad, it was around one quarter of the number of different species that visited some other succulents, such as Schlecteranthus albiflorus (formerly Polymita albiflora) in the same area. Cyathial visitors are attracted in two main ways:

1.6

Biogeography and Ecology

1.6.1 Aspects of Overall Distribution With its many species, Euphorbia has adapted to a wide range of rainfall. For example, in southern Africa species such as E. giessii, E. lignosa, E. phylloclada and E. virosa grow in the hyper-arid coastal Namib Desert with less than 50  mm of rain annually, while E. clavarioides, E. gueinzii, E. natalensis and E. striata occur in the mesic grasslands of the Drakensberg with at least 1000  mm annually. Yet wetter still are the granitic domes of northern Moçambique where the annual rainfall reaches 2000  mm (Jackson 1961) and here E. marrupana, E. mlanjeana, E. namuliensis, E. unicornis and others are locally plentiful. This wide tolerance is repeated in other parts of the world. Within southern Africa, many of the more widely distributed species are also able to tolerate a wide range of rainfall, as in E. avasmontana (< 50 mm to

38

1 Introduction

Fig. 1.64.  Euphorbia gummifera in arid conditions in the Namib Desert near Halenberg, east of Lüderitz, Namibia (© PVB).

Fig. 1.65.  Euphorbia clavarioides in mesic grasslands of the foothills of the Drakensberg, PVB 13079, Ubala, north of Kokstad, South Africa, 21 Dec. 2015 (© PVB).

± 600  mm annually) and E. clavarioides (± 250  mm to over 1000 mm annually). Africa has two main centres of diversity for Euphorbia: a north-eastern and a southern centre. The ‘southern centre’ lies in the tropical and temperate parts of southern Africa, mainly in South Africa. The ‘north-eastern centre’ is exclusively tropical and includes Ethiopia, Kenya and Somalia, together with the neighbouring parts of south-western Arabia. This is especially clear in sect. Euphorbia, which reaches its greatest diversity in the north-eastern centre and is represented by 40 species in the southern centre. Other minor African centres are coastal Morocco and Macaronesia

where subg. Esula is well-represented and where the only northern temperate members of sect. Euphorbia are found in Africa. Subg. Athymalus is restricted to the Old World, with most of its species found in southern Africa. The other three subgenera are cosmopolitan. Each of them has concentrations of species in southern Africa though, in subg. Chamaesyce and subg. Esula, these are minor and their main diversity lies in America and Eurasia respectively. Methods of dating applied to DNA-data have suggested that Euphorbia began to evolve 47–29  million years ago (mya). As a whole, Euphorbia shows a trend hypothesized

Fig. 1.66.  Approximate distribution of subg. Athymalus (© PVB).

Fig. 1.67.  Approximate world-wide distribution of subg. Esula (© PVB).

Fig. 1.68.  Approximate world-wide distribution of subg. Euphorbia (© PVB).

40

for other groups (e.g. the oaks: Axelrod, 1983) with an initial, wide spread of ancestral stocks, after which the species diversified in some of the areas occupied during this initial spread. Among these diversifications, the major succulent radiations within Euphorbia are relatively recent (Bruyns et al. 2011), with the major radiation of spiny succulents in sect. Euphorbia around 8 million years old. This fits in with the results of Arakaki et al. (2011) who showed that most of the succulent radiations are relatively recent. Most of the major continental movements took place long before Euphorbia began to evolve and spread across the world: Africa was separate from the rest of Gondwana by 100 mya, peninsular India broke off and moved away from Africa about 100–65 mya (Axelrod and Raven 1978) and Madagascar reached its present position somewhere between 121 and 60 mya (Rabinowitz et al. 1983; Flores 1970), while Australia has been isolated from Antarctica and South America for ± 35 million years. Also, according to Raven & Axelrod (1974), ‘stepping stones’ of ridges and islands between Africa and South America (for example) ­disappeared ± 65 mya. By the time Euphorbia began evolving 47–29  mya, these land-masses were mostly already separated by large expanses of sea. Consequently, the presence of closely related species of Euphorbia in Africa, America, Australia, peninsular India and Madagascar implies that their ancestors were dispersed over the sea between these land-­masses. There appear to have been several such dispersals between Madagascar and Africa, between Africa and Arabia (and Socotra) and one of the

Fig. 1.69.  Approximate distribution of sect. Euphorbia (© PVB).

1 Introduction

most notable is the major extension of the distribution of sect. Euphorbia (in the last 6 my) from Africa to SE Asia. Although not large, the seeds of Euphorbia are relatively heavy and are not readily airborne, so that the most likely explanations for their movement over large distances is by floating on sea-water (as suggested by Renner 2004 for other groups) or ingestion by migrating birds. Within subg. Athymalus the first groups to diverge have few species, are non-succulent and (apart from the monotypic Madagascan sect. Antso) are widely distributed across the African continent. Those early-divergent groups that occur in southern Africa are also widely distributed, along the margins of the Namib Desert and across the interior of the subcontinent. The more recently derived groups (of the last 5 my) have far more species and are exclusively southern African. Most of this recent diversification took place in the temperate, semi-arid southern and eastern parts of the Cape Provinces of South Africa (Klein Karoo to Grahamstown area), with minor concentrations in other dry mountainous areas (Great Karas Mountains and Rosh Pinah district in Namibia; west of the Khamiesberg, Knersvlakte and Griqualand West in South Africa). Somewhat similar patterns can be discerned in subg. Euphorbia. The earliest branches are widely distributed in the Pacific Islands and in North & South America but have relatively few species. After this, several major diversifications followed in Madagascar. The most species-rich branch is sect. Euphorbia, which is mainly African (with a few species in Macaronesia, the Arabian Peninsula and SE Asia).

1.6 Biogeography and Ecology

However, even sect. Euphorbia shows an unusual pattern, with four species (E. abdelkuri, E. canariensis, E. epiphylloides and E. lacei) branching off earlier than the others and each occurring on a separate set of islands from Macaronesia, the Socotran archipelago, the Andamans and SE Asia (Bruyns et al. 2011; Dorsey et al. 2013). Within southern Africa sect. Euphorbia is well represented, with few species along the edges of the Namib Desert and the majority along the eastern side of the subcontinent. Sect. Euphorbia is mainly tropical and in southern Africa the highest diversity is also close to the Tropic of Capricorn. Nevertheless, there is a sizeable component in the dry parts of the subtropical Eastern Cape (Fig.  5.2, vol. 2), but sect. Euphorbia is practically absent from the temperate Greater Cape Flora with its winter-rainfall regime. In the southern part of Africa, apart from the possibly natural occurrence of E. tirucalli in the granitic hills of northern Moçambique, sect. Tirucalli (Fig. 5.394, vol. 2) is

Fig. 1.70.  Approximate natural distribution of sect. Tirucalli (© PVB).

41

exclusively found along the stony margins of the Namib Desert. Here it occurs from the hyper-arid area around the lower reaches of the Orange River past the northern end of this desert in coastal Angola to peter out about 50 km north of Lobito. As with several such clades associated with the Namib, the diversity is low, it has many close relatives in Madagascar and also some relatives in NE Africa (cf. Didiereaceae, Bruyns et al. 2014, and other similar groups). It is also one of the first African branches to diverge in subg. Euphorbia. Subg. Chamaesyce and subg. Esula are both without major African diversification and are mainly American and Eurasian respectively. Nevertheless, each contains a unique African group of pencil-stemmed succulents, in both cases more diverse in southern Africa than anywhere else. The three species of subg. Esula sect. Africanae that occur in southern Africa are mainly found along the eastern edge of the Namib Desert (Fig. 4.2, vol. 2). The 15 species of sect.

1 Introduction

42

Articulofruticosae (Fig. 3.12, vol. 2) are particularly associated with the Namib Desert and its extension into western South Africa. Apart from the very widespread species, E. rhombifolia and E. spartaria, this section is most diverse in the western coastal plain, with its highest diversity on the Knersvlakte and around the mouth of the Orange River, all within the winter-rainfall, Greater Cape Floristic Region. Among the species of Euphorbia in southern Africa, phylogenetic resolution is not high enough to show how diversification has proceeded and so factors giving rise to speciation are not obvious. Mesic areas such as the SW Cape of South Africa are generally poor in species and semi-arid zones richer, though in tropical zones just outside our area a wealth of species has developed in locally semi-arid habitats embedded within mesic forests, for example in the granitic hills of northern Moçambique and the quartzitic hills of northern

Zambia. In southern Africa the highest diversity is found around Grahamstown in the Eastern Cape. Here the rain mainly falls in summer, semi-arid and mesic areas occur close together and the country is deeply dissected without being very steep, with many different types of rock exposed across the region. The relatively young diversification in Euphorbia in southern Africa and for sect. Euphorbia generally (within the last 10 million years) corresponds well with known changes in the climate towards increasing aridification and in the rainfall regime. There is evidence that the Namib Desert has been arid to semi-arid since about 80  million years ago (Ward et  al. 1983) and that it increased in aridity around 11 mya (Ward and Corbett 1990) or earlier at 17–15 mya (Pickford and Senut 2000), so that it is an older dry region than the other dry parts of southern Africa. Most of the species associated with the Namib Desert have arisen in the last 15 million years. Each of

Fig. 1.71.  Diversity in Euphorbia in southern Africa showing number of species per half-degree square (© PVB).

1.6 Biogeography and Ecology

the subgenera has one or more early-diverging branches with species that are mainly associated with the Namib Desert (and often also with NE Africa), which are older than the major radiations in these subgenera (sect. Lyciopsis in subg. Athymalus; sects. Espinosae and Tenellae in subg. Chamaesyce, subsect. Africanae in subg. Esula and sect. Tirucalli in subg. Euphorbia). These may have arisen in this area when other parts of the subcontinent were not so arid and may be part of an ancient arid flora associated with the Namib Desert and with similarly dry areas in NE Africa. Other examples with similar characteristics are the Didiereaceae, Pachypodium of the Apocynaceae, Pterodiscus and Sesamothamnus of the Pedaliaceae, Anacampseros sect. Avonia of the Portulacaceae and Tribulocarpus of the Aizoaceae. The many species of Euphorbia that are endemic to the Greater Cape Floristic Region of southern Africa (all in sect. Articulofruticosae of subg. Chamaesyce and in subg. Athymalus) have evolved quite recently. Their evolution coincides approximately with the increase in diversity of aridadapted plants in this region, that is thought to follow from the intensification of the Benguela Current around 10  mya (Siesser 1980) and from a substantial rise by 700–900 m of the eastern escarpment between 5 and 3 mya (Partridge and Maud 2000). For the western part of southern Africa, both of these events caused increasing aridity, with the rainfall concentrated in winter and severe droughts in summer (Coetzee 1980; Hendey 1982; Partridge and Maud 2000). These conditions brought about drastic changes in the flora. Increasing aridification probably released new niches for colonisation, leading to speciation and to the present-day endemic species such as E. celata and E. fasciculata of subg. Athymalus and also E. exilis and E. stapelioides of sect. Articulofruticosae. The picture is different in sect. Euphorbia, which is very widespread in Africa but most diverse in East and north-east Africa and in the tropical parts of southern Africa (with only six species in the Greater Cape Floristic Region). Only E. ingens is common to both East Africa and southern Africa (and is found from southern Ethiopia and Somalia to near Durban in South Africa), with elements closely related to E. grandicornis also occurring from Ethiopia to Kwazulu-Natal in South Africa. Between these two zones of greatest diversity much of the land was formerly covered by relatively dense forest (some of which still remains), where the relatively high rainfall of at least 600 mm annually, the plentiful cover of trees and the rapid development of tall, dense grasses in any open spaces mostly prevent succulents from surviving. In these areas, protruding granitic domes and sandstone or quartzitic outcrops with skeletal soils provide the only slightly arid habitats that are not quickly overgrown by dense bush and grass. They support many succulents, with an annual rainfall of 600–2000  mm (Jackson 1961). Species of sect. Euphorbia (Leach 1976c), Aloe (Reynolds

43

1966), Kalanchoe and Plectranthus occur widely on them. These habitats are subjected to relatively harsh climatic conditions and support a distinct flora from the surrounding denser vegetation (Porembski and Barthlott 2000). They are often considerably isolated from each other and this may have contributed to the number of localized endemic species. Phylogenetic studies have shown that these species of sect. Euphorbia are of recent origin (Bruyns et al. 2011).

1.6.2 Aspects of the Distribution of Euphorbia in Southern Africa Of the 172 species recorded in southern Africa, 128 or 74% are endemic. By far the majority, namely 139, of these species is found in South Africa (Table 1.1), of which 62% are endemic. Namibia comes a pale second with 55 species and with a much lower level of endemism (23%), despite its apparent suitability for succulents (Seely and Jacobson 1994). Far fewer species occur in Botswana (24 spp., Carter & Leach 2001 with E. oatesii and E. platycephala added), Lesotho (5 spp., Jacot Guillarmod 1971) and Swaziland (14 spp., Compton 1976) and no species are endemic to these countries except for E. venteri which is endemic to Botswana. In South Africa there are few areas which are free of Euphorbia since, where succulents are absent or rare, ­geophytes or herbs often replace them. Even at higher levels of the Cape Fold Mountains, E. tuberosa is often plentiful and several species are found in the Drakensberg, which are the highest mountains in the region and experience both high rainfall in the summer and frequent falls of snow in winter. As in many other genera of plants, the diversity of Euphorbia in southern Africa (and South Africa in particular) is high. Euphorbia is more diverse where there is considerable variation in altitude and less diverse in flat areas of uniform topography. Consequently, Euphorbia is the least diverse in the very uniform, central region covered by Kalahari sands and is more diverse around the edges of the subcontinent. The greatest concentrations of species are below 1000 m and within 200 km of the coast (an exception is Griqualand West where several groups, especially subg. Athymalus, have local maxima that lie over 500 km from the coast and at around 1500  m). Furthermore, Euphorbia shows a preference for summer rainfall (Fig. 1.71). This can be seen for sect. Athymalus, where dry winter-rainfall areas in Namaqualand have considerable numbers of species, but dry summer-rainfall areas have nearly twice as many (Fig. 1.72). Similarly, sect. Euphorbia is nearly absent from the Greater Cape Floristic Region (only six species and these only on the periphery) and most grow along the eastern side of southern Africa with summer-rainfall. In this respect, Sect. Articulofruticosae is unusual. It shows a distinct predilection for arid to semi-arid areas within the

1 Introduction

44

Fig. 1.72.  Diversity in subg. Athymalus in southern Africa showing number of species per half-degree square (© PVB).

Greater Cape Floristic Region with its mainly winter-rainfall regime (Fig. 3.12, vol. 2). The area with the highest concentration of species per half-degree square (± 30  km  ×  30  km) lies south-west and south-east of Grahamstown (in the squares 3325D and 3326B), where over 22 species have been recorded. Smaller localised centres of diversity are in Namibia in the Kaokoveld and Great Karas Mountains and in the mountains of the northern Richtersveld-Rosh Pinah district straddling the southernmost border between Namibia and South Africa; in South Africa peaks of concentration are found on the Klein Karoo, the Knersvlakte of Namaqualand, the dry parts of the Eastern Cape and southern Kwazulu-Natal (3030C), with several (mostly more minor) concentrations in Kwazulu-­Natal, the Pofadder district of Bushmanland, Griqualand West between Prieska and Kimberley in the Northern Cape, the Soutpansberg and Waterberg and the upper mountainous reaches of the Olifants River in Sekhukuniland of Limpopo Province. This variation in density is mainly brought about by the overlapping of the distributions of widespread species with

others that are very local. This is typical of the Eastern Cape, where there are many species with restricted distributions but it is also true of the Knersvlakte, the Kaokoveld, the Soutpansberg and especially for Sekhukuniland. Sometimes high diversity results from overlaps between species that are widespread in winter-rainfall areas with others widespread in summer-rainfall areas (as in the arid mountains of the northern Richtersveld-Rosh Pinah district, where there are only two local species out of a total of 16). The arid and semi-arid western side of South Africa possesses an unmatched diversity of succulents (Rutherford and Westfall 1986: 61; Coe and Skinner 1993) and these succulents make up a substantial proportion of the Greater Cape Flora (Manning and Goldblatt 2012; Snijman 2013). Figure 1.71 shows that Euphorbia is not remarkably diverse there (especially if compared with the levels of diversity in Crassula and the succulent Aizoaceae), even though there are areas where species of Euphorbia are dominant. However, of the species that occur in this region, most are endemic and

1.6 Biogeography and Ecology

45

it is in this respect that Euphorbia forms an important element of the flora of this region.

1.6.3 Habitat and Ecology In southern Africa Euphorbia is found in most habitats except in dense montane forest (though they do occur elsewhere in montane forest, e.g. E. ampliphylla in parts of tropical Africa). Species inhabit sands, heavily loamy or stony flat areas, stony slopes and shallow soils on solid granitic ‘whalebacks’.

Fig. 1.73.  Smaller and more vulnerable plants partly hidden and protected among the branches of large shrubs of Rhigozum obovatum and Euphorbia radyeri, Euphorbia tubiglans, PVB 11522, just east of Jansenville, South Africa, 3 Sept. 2015 (© PVB).

Some of the smaller species of Euphorbia benefit from the presence of protective shrubs. However, plants of Euphorbia are not wholly dependent on protection by other plants for their survival and may start off life as a small seedling in the open, possibly protected by some stones or larger rocks. Small plants of many of the taller species of sect. Euphorbia are protected from predation by robust thorns, which are later lost once they become large and out of the reach of predators. Many of them appear to be protected also by their unpleasant sap. In arid areas, this lack of need for protection enables the larger species of Euphorbia to establish populations where practically nothing else grows. Seedlings just need a few better-than-average years of rainfall to establish themselves and to reach a size at which they are able to withstand more prolonged dry periods. Unlike stapeliads, which can be difficult to find in nature and very scattered in occurrence, plants of Euphorbia are mostly easily found and they are usually plentiful where

Fig. 1.74.  Young plant ± 30 cm tall, protected against some predation by many spines, Euphorbia sekukuniensis, PVB 11773, 8  km NE of Burgersfort towards Ohrigstad, South Africa (© PVB).

they occur. Some of the small medusoid species may be difficult to locate and also sporadic in occurrence but most species are sufficiently large that they can be spotted from a passing car.

Fig. 1.75.  Cryptic appearance of one of the small ‘medusoid’ species, Euphorbia decepta, PVB 12994, NE of Klipplaat, South Africa, 4 Sept. 2015 (© PVB).

1 Introduction

46

A. Sandy habitats. These are plentiful in southern Africa. The low-lying, flat, coastal, sandy areas of the western Cape sandveld are fairly rich in species of Euphorbia. Many are widespread on these sands and may sometimes be dominant, but only E. stapelioides is endemic there, unlike the succulent Aizoaceae where there are many endemic species.

Fig. 1.78.  Euphorbia juttae (± 5 cm long) among calcritic gravel on the edge of a pan, PVB 3474, Pearson’s Hunt, north of Upington, South Africa, 9 Jan. 1989 (© PVB).

Fig. 1.76.  Euphorbia spartaria growing among scattered trees in pale pinkish ‘Kalahari sand’, PVB 12299, north of Molepolole, Botswana, 23 Dec. 2012 (© PVB).

This eastward drop-off in diversity on sands in southern Africa continues to the eastern flank of the region and only the geophytic E. gueinzii and the herbaceous E. natalensis inhabit the grassy, settled dunes of the KwaZulu-Natal and the southern Mozambican coast. B. Loamy or stony flats to slopes.

So-called Kalahari sands cover a large portion of central southern Africa in eastern Namibia, Botswana and parts of the Northern Cape of South Africa. This ‘Kalahari desert’ is not intensely dry (receiving from 200 mm annually in the southwest to more than 500  mm in the north) but Euphorbia is poorly represented in this region, with very scattered occurrence of E. duseimata, E. maleolens and E. spartaria, while E. crassipes and E. juttae occur on calcrete-outcrops around some of the shallow pans that developed along the sand-choked drainage systems (Grove 1969).

In the semi-arid, karroid parts of South Africa and Namibia species of Euphorbia are common in hard, stony ground in flat areas. They are less plentiful on the slopes of hills and are rare in steep spots. This type of habitat varies from stony stands of Rhigozum trichotomum to dwarf succulent scrub dominated by Ruschia spinosa or various dwarf Asteraceae and here many of the smaller southern African Euphorbia occur. Large populations of some species (e.g. E. esculenta) develop in these areas. Larger shrubby species of Euphorbia (E. radyeri around Jansenville, E. avasmontana

Fig. 1.77.  Euphorbia crassipes with calcritic gravel along the edge of a pan, PVB 12072, north of Hotazel, South Africa, 21 Dec. 2011 (© PVB).

Fig. 1.79.  Euphorbia avasmontana dominating the vegetation at the base of the mountains, PVB 11355, Narudas Nord, Great Karas Mountains, Namibia, 6 Apr. 2009 (© PVB).

1.7 Cultivation

47

in the Great Karas Mtns) may develop into the dominant element and many of the tree-like species (like E. grandidens and E. triangularis) form dense colonies on rocky slopes (much more rarely in flat areas, as in E. excelsa) where they can become dominant too. Euphorbia is poorly represented in the sandstone-derived, nutrient-poor, acidic soils of the mountains of the south-­ western and southern Cape. The herbaceous E. genistoides is endemic to these mountains and E. heptagona and E. polygona are found in locally semi-arid patches towards the east in them as well. C. Solid outcrops of rock or ‘whalebacks’. An important habitat for succulents in the wetter parts of south tropical Africa is provided by rounded domes of granite, which form characteristic steep to gently sloping ‘whalebacks’. This is especially true in Angola, Moçambique and Zimbabwe (also further north in Congo and Kenya as well as in eastern Madagascar) but also occasionally in South Africa. On these outcrops, which are subjected to extreme climatic conditions (such as wide fluctuations in temperatures and light-intensity), isolated patches of vegetation develop in variably thick accumulations of weathered grit and leaf-litter on the solid rock, where a dense mat of roots and black soil combine to give the ground an impenetrable, peat-like consistency. These patches of skeletal soil are often dominated by various dessication-tolerant resurrection-plants such as the tussockforming sedge Coleochloa, Myrothamnus and various species of Xerophyta. Where the patches are sufficiently isolated and not too regularly burnt, a wealth of succulents, especially belonging to Aloe, Cynanchum, Euphorbia, Kalanchoe, Plectranthus and several Periplocoideae (Apocynaceae) may develop. Many species of sect. Euphorbia are only found in these habitats, especially in Angola and Moçambique. In South Africa such domes are present on the eastern and western sides. Those on the eastern side may be inhabited by species such as E. clavigera, E. cooperi, E. schinzii and E. vandermerwei, but only the last is restricted to them. The much drier gneissic domes in western South Africa in the winter-rainfall parts of the country do not have any distinctive Euphorbia on them, though they may also be rich in succulents.

Fig. 1.80.  Euphorbia cooperi var. cooperi growing in shallow soil on a large granitic dome, PVB 11904, Gutschwa Kop, north of Nelspruit, South Africa, 27 Jul. 2011 (© PVB).

Fig. 1.81.  Euphorbia vandermerwei, in shallow soils on granitic slabs with Selaginella and clumps of grass, PVB 11905, NE of Nelspruit towards Numbi Gate, South Africa, 25 Jul. 2011 (© PVB).

1.7

Cultivation

In Euphorbia from the summer-rainfall areas growth takes place between spring and autumn and they prefer to be kept entirely dry during the cooler months of winter. Species from the winter-rainfall areas mostly prefer water in winter only and often do not respond much if watered in summer. Many of the larger species of Euphorbia prefer to grow out of doors, in a suitably sized pot in a sunny spot against a north-facing wall (in the southern hemisphere) with generous circulation of air. Smaller species, which are often more

1 Introduction

48

sensitive to direct sunlight are better kept under shelter to prevent burning. Greenhouse cultivation of the smaller species is best achieved with the pots on tables (about waist-height is best), since pots on the ground are more prone to neglect. If possible, the sides between the roof and the level of the tables should be open to allow circulation of air, but covered with fine-meshed wire to keep birds, cats and rodents out. Where such an ‘open’ house is impractical, circulation of air can be achieved with several small electric fans.

1.7.1 Propagation A. Growing media. Probably because of their fine roots, most plants of Euphorbia prefer somewhat loamy soils which prevent the fine roots from dying during prolonged dry periods. Some species that naturally occur in densely vegetated situations (e.g. E. bubalina and trees like E. triangularis) will grow easily in loose, coarse, humus-rich soil. Watering should be rationed according to their origin and those from moist areas prefer regular watering (up to twice a week) in the summer while those from drier areas can manage on once a week or less. The most difficult species to grow are the succulents from the winter-rainfall areas (such as E. fasciculata and E. restituta), which often show little tolerance for water in summer and prefer also relatively little during the winter. Those from summer-rainfall areas are rarely problematic and, once established, will continue growing for years or even decades in the same pot, provided it is large enough. B. Repotting. If a specimen starts to push the sides of its pot out of shape it has clearly outgrown it and ought to be replanted into a larger one. Once removed, some of the longer roots may be trimmed (and any dead ones pruned back), after which the plant should be left out for 2 weeks to a month in a cool, dry place to allow any damaged roots to dry off and heal. After this it may be planted again and, if this is done during the warmer weather, watering may resume soon. The plant may even be watered immediately, though then a little fungicide may be added to inhibit decay. C. Propagation by cuttings. Many species of Euphorbia are easily propagated by cuttings. This includes small species like E. patula and also most of the smaller, shrubby members of sect. Euphorbia.

In many species which have distinct adult and juvenile stages, cuttings may also root readily. However, plants grown from cuttings from an adult stage usually do not achieve the same form as a mature plant. In medusoid species like E. caput-­medusae this can be seen in many of the early illustrations of this species in cultivation in Europe (which clearly show plants grown from branches) and it is also true of small species like E. stellata. Such cuttings can be used to produce seed from which normal plants can be grown. Cuttings of large, shrubby members of sect. Euphorbia, such as E. avasmontana and E. virosa and trees like E. ingens and E. triangularis, also readily root but these also do not assume the same form as the adult plant. Leach discovered that this was a useful way to obtain flowers in large species of sect. Euphorbia: such plants may take decades to reach maturity and to flower if grown from seed, but a cutting taken of a branch of an adult plant will often flower soon after rooting and may even flower before it has rooted. Some, however, such as E. loricata and E. oxystegia are not readily grown this way without more special treatment, while cuttings of branches of some of the medusoid species such as E. crassipes and E. decepta also dry out before rooting. When a cutting is made, it must be cut off cleanly where the branch is constricted at its base and allowed to dry for 2–4 weeks, if not longer. If kept out sufficiently long, roots may begin to develop around its base. When planted, it is best to prop up the cutting firmly with some stones. This stabilises it, preventing any movement from damaging the newly formed roots and also preventing the developing roots from pushing the cutting out of the soil. At this stage plenty of water is necessary. In some cases, rooting is very rapid, but it can also take several months before roots develop. Rooted cuttings rapidly swell up and soon begin to make fresh growth, so that their establishment is usually quite obvious. D. Propagation by seeds. Although a wide variety of seeds of Euphorbia is not readily available, seeds nevertheless provide a relatively easy method of propagation when they can be found. Seeds can be planted within a month after they are released but should not be kept for more than 2 years. Various mixtures can be used to germinate seed, but I have found the hydroponic peat/perlite/vermiculite mixture (made in the ratio 2:1:3 or 2:1:4) to be effective. It is especially suitable since it retains moisture for long periods without being waterlogged (assisting the young seedlings in their first weeks) and it is loose and well aerated. To plant seed, fill the pot with this mixture to within 15 mm

1.7 Cultivation

of the rim, flatten it down lightly and then place the seeds on it. Cover the seeds with a layer 2–4  mm thick of the same mixture, so that they are not visible. After this soak the whole pot overnight (to make sure the mixture is thoroughly wet) in water mixed with fungicide. Having removed it from the water, keep it in a shady but warm place in the greenhouse. It may be necessary to water again after a week or two to prevent the upper layers from becoming too dry. Seedlings take anything between 40 h and 2 weeks to germinate, with some germinating even after a year, especially if exposed to some rain (Dyer 1935a). After germination they grow rapidly. The initial soaking with a fungicide ought to prevent any damping off in the early stages. Once several leaves have developed it is time to consider repotting the plants into normal soil, though they can survive in the hydroponic mixture for a year or more.

1.7.2 Diseases, Parasites and Predation Plants of Euphorbia are mostly relatively disease-free. In habitat scale-insects can infect them (this is occasionally seen in E. polygona and E. heptagona) but in cultivation these disappear if treated with a systemic poison containing the active ingredient Imidacloprid. The red spider-mite (which belongs to the genus Tetranynchus) is a pest to which cultivated plants of Euphorbia are prone. These minute vegetarian mites cause

Fig. 1.82.  Egg-like ‘cocoon’ of beetle belonging to the Cerambycidae (Prioninae) whose earlier stages hollow out the thicker branches of Euphorbia lignosa and kill the plants, here among the remains of a dead plant of E. lignosa, PVB 12836, east of Wlotzkabaken, Namibia, 21 Dec. 2014 (© PVB).

49

discoloration of the shoot where they penetrate the surface to feed off the sap and, once they become very numerous, they also spin a fine protective web over the surface of the plant. The discoloration of the surface is caused by the death of cells where they have penetrated and it causes the affected area eventually to turn an unsightly brown and even atrophy if the attack is sufficiently severe. Red spider-mites particularly thrive in dry, still, warm conditions. Increasing the circulation of air around the plants may reduce some of these problems. Biological control of this pest is possible. This involves introducing a predatory red mite (of the genus Phytoseiulus) which lives off the red spider-mite. However, this is not widely available and it is also very sensitive to the use of any poisons on the collection. Treatment with an acaricide such as Kelthane, Omite or Abemectin is often more effective. The last-mentioned is particularly effective in killing off eggs and preventing new generations of mites from re-infecting the collection. Red spider-mites rapidly develop resistance to sprays, because of their very short generations and then they are difficult to control. They also spread on the wind from surrounding vegetation, with the result that new infestations may appear after they were apparently eradicated. In habitat, certain of the larger species of Euphorbia are parasitized by other plants. The small but remarkable genus Hydnora (Visser 1981) consists of parasites that flourish in semi arid to very arid areas on the roots of species of Euphorbia such as E. damarana, E. mauritanica, E. radyeri and E. virosa. Other parasites with a wide range of hosts, such as Viscum and

Fig. 1.83.  Dead beetle belonging to the Cerambycidae (Prioninae) and parasitic on Euphorbia lignosa, ± 40 mm long, east of Wlotzkabaken, Namibia, 21 Dec. 2014 (© PVB).

50

1 Introduction

Fig. 1.84.  The parasitic Hydnora on roots of Euphorbia. Left: H. africana on E. mauritanica. Upper part of flower just protruding from ground, near Karoopoort, South Africa (© Geoff Tribe). Right: A species of Hydnora on E. virosa and just past flowering, south of Namibe, Angola, 13 Jan. 2007 (© PVB).

Fig. 1.86. Parasitic Tapinanthus oleifolius on branches of Euphorbia avasmontana, PVB 3558, Pieterskloof, Great Karas Mountains, Namibia, 16 Jan. 1989 (© PVB).

Tapinanthus (Loranthaceae) may occasionally occur on E. avasmontana and other larger species. Much more hostspecific is the rare Viscum minimum, which is only known to parasitize the branches of E. polygona. Similarly, in habitat, many species of Euphorbia are eaten by animals and this is discussed further under ‘Uses’.

Fig. 1.85. Parasitic Viscum dielsianum (PVB 8084) on branches of Euphorbia avasmontana, Koimasis, Tiras Mountains, Namibia, 3 Jan. 2000 (© PVB).

1.8 Uses

51

Fig. 1.88.  Viscum minimum on Euphorbia polygona near Grahamstown (© Roy Lubke).

1.8

Fig. 1.87.  Viscum minimum on branch of Euphorbia polygona. Previously unpublished painting from 1933 in the Bolus Herbarium by ‘J.A’. This parasite is reduced to a minute plant 1–3 mm long, bearing the brightly coloured fruits  ±  5  mm in diameter seen here. It is only known on this species of Euphorbia (© Bolus Herbarium, University of Cape Town).

Uses

Many species of Euphorbia are regarded as highly poisonous on account of their unpleasantly caustic, irritant and sometimes evil-smelling milky sap. Paterson (1789) recorded that the sap of E. virosa was one of the most virulent poisons known. He described how the sap of E. virosa was used on arrow-tips by the Bushmen and also to poison a pool of water, which would rapidly kill any animal that subsequently drank from it. White et al. (1941) re-iterated the belief that the sap of this species is exceptionally poisonous. While there is little doubt that the sap is highly unpleasant and irritating, Marloth (1930) noted that the tips of the branches of plants of E. virosa are occasionally eaten by wild animals (ostensibly baboons, but also possibly by the larger antelope such as kudu or eland) and I have observed the same at several localities along the Orange

52

1 Introduction

River. Loutit et  al. (1987) documented the preference of black rhinoceros in the Skeleton Coast of northern Namibia for E. virosa and found that in one area almost all plants of it were sampled and one was eaten down virtually to a stump (E. damarana was also eaten but less enthusiastically). Jacobsen (1988: 62) also mentioned its consumption by black rhinoceros. Both these accounts suggest that its toxicity has been exaggerated greatly and Fig. 1.90 shows two black rhino eating a plant of E. virosa. Euphorbia ingens was also observed to be keenly consumed by black rhinoceros in periods of drought in Malawi (Dudley 1997). Of interest in this regard is Marloth’s (1930) report of an experiment conducted on a farm near Prieska, where a goat survived for 3  weeks on a diet of branches of Euphorbia avasmontana. This established that E. avasmontana is not poisonous to stock.

Fig. 1.90.  Black rhino eating a large shrub of Euphorbia virosa, Damaraland, Namibia (© Conservancy Rhino Ranger Incentive Programme, Namibia).

Fig. 1.89. Black rhino chewing parts of a shrub of Euphorbia damarana, Damaraland, Namibia (© Conservancy Rhino Ranger Incentive Programme, Namibia).

In a similar vein, dramatic and sensational statements were also made during the twentieth century about how poisonous poinsettia (Euphorbia pulcherrima) is. In an effort to ascertain the truth of these claims, Krenzelok et al. (1996) examined reports on 22,793 medical cases that had occurred between 1985 and 1992 involving poinsettia. They showed that none of these incidents led to fatalities, even from eating its leaves and only in one case did it have any ‘major effect’, though this was also not fatal. Consequently, there is no evidence that it is dangerous. Similarly all other statements about the toxicity of Euphorbia must be regarded sceptically.

In southern Africa many species are edible (and this is true elsewhere too, as documented by Croizat 1967, for Venezuela), but edibility varies greatly between related species. Euphorbia burmanni and E. rhombifolia are edible even to humans (a fact reflected in the name ‘soetmelkbos’ for the former), though few people make use of them, while the closely related E. gentilis is highly unpleasant to us. Some of them (such as E. rhombifolia) are heavily grazed in some areas, providing an important source of food to stock and wild animals, but remain untouched in other regions. Other species are left strictly alone for most of the year but are grazed at certain times (possibly in dry periods when there little else with moisture in it). This has been noted in E. gregaria and E. damarana (also E. carunculifera in Angola), where some of the shoots are gnawed at and the toughest fibrous content is left hanging on the bush (see also Fig.  1.55), also in E. avasmontana. Euphorbia mauritanica is less fibrous and more succulent and the upper parts of the shoots are bitten off. It is widely believed (though apparently unverified) that buck tend to nibble on species of Euphorbia (such as E. mauritanica) only when they are particularly plagued by parasites in their digestive system. Few species are left completely untouched. In such low-growing species as E. clavarioides (Hargreaves 1992a), E. ferox and E. pulvinata the tips of many of the branches may be nibbled off. In others such as E. esculenta and E. rhombifolia, whole plants may be ripped apart or out of the ground (probably by porcupines) and the underground basal part of the stem eaten away while the above-

1.8 Uses

Fig. 1.91.  Small, heavily grazed plant of Euphorbia rhombifolia, Tafelberg, SW of Fraserburg, South Africa, 30 Sept. 2012 (© PVB).

53

Fig. 1.93.  Predation on shoots-tips in Euphorbia caput-medusae. Hoek van Bobbejaan, Cape Point Reserve, South Africa, 23 Dec. 2008 (© PVB).

Fig. 1.94.  Heavily eaten plant of Euphorbia esculenta, Springbokvlakte, east of Steytlerville, South Africa, 13 Apr. 2017 (© PVB).

Fig. 1.92.  Predation on Euphorbia avasmontana by baboons, leaving only the fibrous material in the branches, Elisenheim, Windhoek (© Geoff Tribe).

ground parts are left intact. Krauss (1844: 280) related how in 1839 during a serious drought in the area north of Uitenhage, the peeled stems of the extremely common ‘Fingerbollen’ (which he named Euphorbia caput-medusae, but this was probably E. esculenta) were fed to cattle and also eaten by the farmers raw, roasted on the coals or cooked with meat. Around Jansenville, where it is often dominant, Euphorbia radyeri was once regularly harvested for fodder. This process involved cutting the branches off near groundlevel, chopping them into small pieces which are left to dry for a day or two and then fed to goats and sheep. This practise still continues, albeit much more rarely. Its effect on populations of this species is unclear. Mr. H. Shoesmith, a farmer south of Cradock recorded chopping out plants of E.

54

polycephala and feeding them to his goats in times of drought. The fact that this species is highly palatable to Angora goats has led to its virtual extermination so that it is now greatly endangered and is only found where some protection is provided. Watt & Breyer-Brandwijk (1962) listed many other possibly edible species, occasionally to humans and mainly to stock. Horticulturally Euphorbia is significant. The Mexican Euphorbia pulcherrima (‘poinsettia’) is sold in enormous numbers around Christmas in the northern hemisphere and is said to be worth millions of dollars as the best-selling pot-­plant in North America. The Madagascan E. milii is also important as a pot-plant, especially in eastern Asia and many herbaceous species of subg. Esula are commercially important as ornamentals in Europe. More informally, in many parts of tropical Africa large species like E. abyssinica, E. ampliphylla, E. ingens and the ubiquitous E. tirucalli are widely propagated by cuttings and used as

Fig. 1.95.  Euphorbia tirucalli planted as cuttings and growing to form a dense hedge to ± 2 m tall, used to delimit properties and fields, near Yechila in northern Ethiopia (© PVB).

Fig. 1.96.  Euphorbia ingens planted as cuttings and forming almost impenetrable, wall-like hedge, delimiting field near Jos, Plateau State, Nigeria, 16 Sept. 2013 (© PVB).

1 Introduction

fences demarkating properties. Usually 0.5–1 m long when planted, these cuttings are placed so closely that they quickly form a living, impenetrable, fence-like hedge which becomes denser as they begin to grow and spread upwards and outwards. In tropical Africa a species of sect. Crepidaria (the former Pedilanthus) is often planted as an ornamental, though its use is unclear. In southern India fences are also made of E. antiquorum, occasionally E. nivulia, E. tirucalli and E. tortilis. Also horticulturally significant are the many succulent species from South Africa such as E. globosa, E. meloformis and E. obesa. These are very popular subjects in specialist collections around the world and are now widely propagated by seed. Watt & Breyer-Brandwijk (1962) listed many additional uses for sap from Euphorbia. These range from stunning fish in ponds, as an ingredient in ‘arrow-poisons’ (but perhaps only as a binding agent), as bird-lime, for brewing beer to treating skin-cancers and warts. Many more were listed by Hargreaves (1976), including as chewing gum and as a ‘personal lubricant’, sometimes with painful consequences. The drying and sawing into planks of the soft wood from trees of E. robecchii in Somalia to make packing cases for exporting bananas was recorded by G.W.  Reynolds near Kismayo in October 1953. ‘A fire is lit around the Euphorbia tree which burns off the bark and eliminates the latex. When dry, the tree is felled and sawn at ground level, and as near the lowest branch as possible. The logs are then transported to the sawmills, where they are sawn into slats for the making of crates.’ There have been occasional attempts to exploit some species such as E. ingens, E. tirucalli and E. triangularis commercially for rubber and small processing plants were even set up for this (White et al. 1941, for E. triangularis). This enterprise was hampered greatly by the low percentage of rubber in the latex compared to its high resin-content (almost 80%) and was abandoned. Leach (1973a), too, mentioned the export of dried sap of Euphorbia from Angola, presumably originating from E. carunculifera. This was also for the production of rubber, but was soon abandoned. The caustic or irritant nature of the sap of Euphorbia has been recorded many times (Phillips 1924; Watt and Breyer-­ Brandwijk 1962; Leach 1966: 176; Hargreaves 1976, 1984: 94 and many others) and experienced at first hand after pressing many plants of Euphorbia for herbarium specimens. This ability to cause pain and discomfort is not restricted to the fresh sap and dried remnants of the sap can also affect the eyes, soft tissues of the nose and near the mouth. Investigations of the contents of the sap were carried out for E. cooperi (Gschwendt and Hecker 1973) and E. triangularis (Gschwendt and Hecker 1974). Irritants were found to be several different diterpene esters, but they did not have a direct, carcinogenic effect. In this respect

1.10 Key to the subgenera of Euphorbia in Southern Africa

they were much less active than esters isolated from ‘Croton oil’, extracted from the seeds of Croton tiglium. This was at one time used in the pharmaceutical industry, but it has now largely been removed because of its possible carcinogenic effect.

1.9

Systematic account

Euphorbia L., Sp. Pl.: 450 (1753). Type (designated by Millspaugh 1909: 306): Euphorbia antiquorum L. Arthrothamnus Klotzsch & Garcke (see under subg. Euphorbia). Chamaesyce Gray (see under subg. Chamaesyce). Dactylanthes Haw. (see under subg. Athymalus). Esula (Pers.) Haw. (see under subg. Esula). Lyciopsis (Boiss.) Schweinf. (see under subg. Athymalus). Medusea Haw. (see under subg. Athymalus). Monadenium Pax (see under subg. Euphorbia). Tirucalia Raf. (see under subg. Euphorbia). Treisia Haw. (see under subg. Athymalus). Geophytes, prostrate to erect herbs, woody or succulent shrubs or trees, with or without spines, all parts containing copious milky sap. Leaves alternate to opposite, entire to denticulate (rarely lobed), often deciduous, often reduced to minute caducous rudiments, shortly petiolate to sessile; stipules often absent or present as two small spines (in sect. Euphorbia). Synflorescences solitary and terminating branches to axillary and many towards tips of branches, forming cymes, umbels or clusters of cyathia borne on peduncles and subtended by leaf-like to scale-like often

55

paired bracts (sometimes partly fused in sect. Monadenium). Cyathium bisexual or unisexual (then often with rudiments of other sex), consisting of cup-like involucre, with (1) 4–5 (−12) spreading contiguous to widely spaced stalked to nearly sessile glands along its upper margin alternating with 4–5 (−12) usually inflexed (rarely erect or spreading), entire to finely toothed and often pubescent reddish green to white lobes each enclosing a group of male florets and all surrounding single female floret; glands often brightly or differently coloured to rest of cyathium, often with differently coloured tooth-like to finger-like appendages on outer margins, usually exuding sticky sweet secretion on upper surface; male florets consisting of rapidly maturing and shed anthers, each without ‘calyx’ on short, often pubescent pedicel, surrounded and intermingled with glabrous to pubescent bracteoles; female floret consisting of sessile to pedicellate glabrous to pubescent spherical to ellipsoidal or roundangled ovary included within or exserted from cyathium, occasionally with variously shaped rudimentary ‘calyx’ around its base at top of pedicel, ovary with 4–5 (6) cells, each cell with 1 pendulous ovule; styles 3 (−5), free to base or united into column above ovary, with entire or bifid tips. Capsule 3- (rarely 4- to 6-) locular, locules usually separating explosively when ripe from central axis, splitting along inner face of middle of each locule to free seeds; seeds smooth to sculptured, usually 1–3 (−10) mm long, spherical to cylindrical or 4-angled, sometimes with caruncle at point of attachment to capsule. Euphorbia is a cosmopolitan genus with roughly 1840 species. Of these, 172 species are found in southern Africa and 128 of these are endemic. Four subgenera are recognized and all four are represented in southern Africa (Table 1.1).

1.10 K  ey to the subgenera of Euphorbia in Southern Africa 1. Cyathia in a terminal umbel with several rays..................................................................................................................2. 1. Cyathia axillary (often solitary at end of short axillary branches) or terminal but then not produced in umbel...............................................................................................................................................4. 2. Leaves estipulate, umbels with central ± sessile cyathium and others on rays surrounding it, stem and branches without tubercles, seeds with caruncle..................................................................................subg. Esula 2. Leaves often with glandular or filamentous stipules, umbel without central ± sessile cyathium, cyathia on branching rays, stem and branches sometimes with tubercles, seeds without caruncle..................................3. 3. Stipules glandular or absent, cyathial glands often 4, bilobed or entire, seeds wrinkled or tuberculate................................................................................................................subg. Chamaesyce (sect. Frondosae) 3. Stipules filamentous or absent, cyathial glands 5, crenulate or with many finger-like processes on outer margins, seeds smooth or wrinkled...........................................................................subg. Athymalus 4. Synflorescences axillary on short, non-tuberculate peduncles.................................................................subg. Euphorbia 4. Synflorescences terminal or borne on tips of axillary and finely to coarsely tuberculate short shoots.....................................................................................................................................................5.

56

1 Introduction

5. Annuals or non-succulent (non-geophytic) ephemerals often with aborting main axis, cyathia minute (1–2.2 mm diam.), glands often 4 and frequently with entire petaloid (white or pink) appendages.....................................................................subg. Chamaesyce (sect. Anisophyllum & sect. Tenellae) 5. Plant not ephemeral (sometimes geophytic with deciduous leaves), cyathia > 2.5 mm diam., glands 5 and without petaloid appendages.............................................................................................................6. 6. Woody shrub to small tree with shiny peeling bark (lacking green photosynthetic stem and branches), involucral glands entire, seeds smooth and with caruncle.....................................................................................................................subg. Chamaesyce (sect. Espinosae) 6. Shrub to dwarf succulent with fleshy green photosynthetic stem and branches without peeling bark or geophyte, involucral glands often crenulate or with finger-like processes, seeds usually finely to coarsely tuberculate and without caruncle.............................................................................................7. 7. Plant a geophyte with deciduous above-ground parts.......................................................................................................8. 7. Plant not geophytic, with perennial above-ground branches, usually succulent and sometimes with very reduced and ephemeral leaves..............................................................................................................10. 8. Stem and branches (if any) wholly subterranean and perennial with deciduous leaves and peduncles arising from their apex beneath soil...........................................................................................subg. Athymalus (sect. Rhizanthium) 8. Branches bearing leaves and cyathia above soil, usually deciduous.................................................................................9. 9. Cyathia bisexual, with conspicuous finger-like projections on outer margins of glands.....................................................................................subg. Athymalus (sect. Dactylanthes) 9. Cyathia unisexual, with finely crenulate or entire outer margins of glands....................................................................................................................... subg. Chamaesyce (sect. Gueinziae) 10. Stem and branches not tuberculate and not angled (frequently slender and cylindrical), plant not geophytic, leaves usually opposite and often with glandular stipules, synflorescences with few small generally unisexual cyathia, margins of glands entire..............................................................................subg. Chamaesyce (sect. Articulofruticosae) 10. Stem and branches mostly tuberculate and very fleshy with tubercles often fused or arranged into angles, leaves usually alternating or spirally arranged, usually estipulate, synflorescences with few often large bisexual cyathia (often 6–10 mm diam. or broader), margins of glands often crenulate or extended into finger-like processes.................................................................................................................................subg. Athymalus

Euphorbia subg. Athymalus

Euphorbia subg. Athymalus Neck. ex Reichb., Conspect. Reg. Veg. 1: 194 (1828). Athymalus (Neck. ex Reichb.) Raf., Fl. Tellur. 4: 112 (1838). Type: Euphorbia tridentata Lam. (= Euphorbia patula Mill.). Dactylanthes Haw. (see under subsect. Dactylanthes). Euphorbia subg. Rhizanthium (Boiss.) Wheeler (see under ser. Rhizanthium). Euphorbia subg. Trichadenia (Pax) S. Carter (see under subsect. Dactylanthes). Lyciopsis (Boiss.) Schweinf. (see under sect. Lyciopsis). Medusea Haw. (see under subsect. Medusea). Treisia Haw. (see under ser. Treisia).

2

Bisexual (occasionally unisexual), annual herbs, geophytes, shrubs or small trees, often succulents with cylindrical or angled stem and branches, sometimes with spines developing from tips of branches or from sterile short-shoots on branches. Leaves alternate or spirally arranged, obovate to lanceolate and often reduced to minute rudiments, mostly deciduous and often rapidly caducous, often arising on raised tubercles (these often arranged in rows or joined into angles); stipules occasionally present and glandular or filiform. Synflorescences in axils of tubercles towards apices of stem and branches or terminal on peduncle, peduncles usually simple, often with alternating or opposite bracts arising on

Fig. 2.1. Relationships between the southern African members of subg. Athymalus. The sections represented in southern Africa are red (those not represented are green). The length of the vertical side of each of the triangles is proportional to the number of species in that section (adapted from Peirson et al. 2013).

© The Author, under exclusive license to Springer Nature Switzerland AG 2022 P. V. Bruyns, Euphorbia in Southern Africa, https://doi.org/10.1007/978-3-030-49402-5_2

57

2  Euphorbia subg. Athymalus

58

small tubercles, bracts often differently shaped from leaves and sometimes larger just beneath cyathium, further peduncles with terminal cyathia occasionally developing from axils of uppermost bracts. Cyathia bisexual or unisexual (then males larger than females); glands 4–5, usually elliptic and flat with entire, crenulate, pectinate or deeply toothed outer margins, marginal teeth sometimes differently coloured, finger-like and sometimes branched, without petaloid appendages; male florets almost always with pubescent pedicels.

Capsule obtusely 3-lobed to spherical, smooth to verrucose, glabrous to pilose, 2.5–12  mm diam., sessile to exserted, dehiscing explosively. Seeds 3 per capsule, 4-angled to ellipsoidal or ± spherical, smooth to tuberculate, ecarunculate. Of the seven sections recognized in subg. Athymalus by Peirson et al. (2013), three are represented in southern Africa. These are sect. Anthacanthae (with most of the species), sect. Crotonoides and sect. Lyciopsis.

Key to the sections of subg. Athymalus in southern Africa 1. Annual or short-lived perennial herbs with prominent leaves with serrate margins, not succulent and without tubercles on stems and branches...............................................................................................................................2.2. sect. Crotonoides 1. Perennial leafy deciduous shrubs, trees, geophytes or succulents with tuberculate stem and branches with leaves often reduced to minute caducous rudiments...................................................................................2. 2. Pencil-stemmed succulent shrubs or medusoid, cactoid, meloform or shrubby succulents with tuberculate and often angled green stem and branches, often bearing spine-like sterile short-shoots...............................................2.1. sect. Anthacanthae 2. Woody or semi-succulent shrubs to trees or geophytes, leafy (mostly deciduous), with grey, yellow or brown, sometimes peeling bark (usually not green), stems and branches not tuberculate...............................................................3. 3. Woody herbs, shrubs to small trees, often with peeling yellow or brown papery bark............................2.3. sect. Lyciopsis 3. Semi-succulent geophytes with swollen rootstock...........................................................................2.1. sect. Anthacanthae

2.1

Sect. Anthacanthae

Euphorbia sect. Anthacanthae Lem., Ill. Hort. 2: misc. 69 (1855). Euphorbia subsect. Anthacanthae (Lem.) Pax & K.Hoffm. in Engler, Nat. Pflanzenfam., ed. 2, 19c: 216 (1931). Type (designated by Wheeler 1943): Euphorbia heptagona L. Bisexual or unisexual geophytes, dwarf succulents or succulent shrubs with herbaceous to succulent branches, stem and branches often tuberculate (sometimes plant consisting of solitary cylindrical tuberculate stem with tuft of deciduous leaves at apex), usually with green to grey-green photosynthetic bark, often armed with spines developing from branch-tips or from persistent hardened short-shoots arising in axils of tubercles. Leaves deciduous to ephemeral, entire, alternate, narrowly lanceolate or often reduced to minute rudiments, often sessile and often arising at apices of conical tubercles (these often arranged or fused vertically into 4–20 straight or spiralling rows or angles along stem and branches), mostly without stipules. Synflorescences solitary and terminating branch or solitary in axils of tubercles (often a single cyathium terminating a short to 10 cm long sometimes tuberculate and sometimes bracteate peduncle

which may branch among uppermost bracts to bear further cyathia and even become umbel-like), cyathia bisexual or unisexual 2.5–19  mm diam. (male cyathia often twice the diameter of female cyathia) surrounded by leaf-like to much reduced bracts, glands 4–5 with outer margins entire to crenulate or with 2 to many finger-like spreading and sometimes branched projections. Capsule obtusely 3-lobed, 4–11  mm diam., sessile to exserted on pedicel to 15 mm long. Seeds smooth to verrucose, 2–5  mm long, ovoid to top-shaped, sometimes slightly 4-angled, grey to black, without caruncle. This section contains 87 species, which only occur in subSaharan Africa. Of these 87 species, 73 are found in southern Africa and the other 14 species are widely but sparsely distributed over the rest of sub-Saharan Africa. Five subsections are recognized. Sect. Anthacanthae was the second of two sections named by Lemaire (1855), after five of those species from the Cape with ‘flower-bearing spines’ (Haworth used the name Florispinae to indicate the same thing) and Lemaire suggested that, if it were to be recognized as a genus, the name Anthacantha could be used. Here sect. Anthacanthae contains much more than just the flowerspined species that Lemaire was familiar with.

Crotonoides Lyciopsis

Section Anthacanthae

Medusea Platycephalae Pseudeuphorbium

Dactylanthes Florispinae

Sub-section

Hystrix Meleuphorbia Rhizanthium Treisia

Series

Namibia (endemic) 13 (1) 0 0 0 0 0 0 7 (1) 0 6 (0) 4 (1) 2 (0)

South Africa (endemic) 71 (55) 6 (5) 26 (25) 4 (4) 14 (13) 3 (3) 5 (5) 32 (24) 0 7 (1) 1 (0) 0

Southern Africa (endemic) 73 (68) 6 (5) 26 (26) 4 (4) 14 (14) 3 (3) 5 (5) 33 (33) 1 (0) 7 (4) 4 (2) 3 (0)

Elsewhere (total) 19 (87) 1 (6) 0 (26) 0 (4) 0 (14) 0 (3) 0 (5) 0 (33) 10 (10) 6 (10) 6 (8) 20 (20)

Table 2.1  The sections, subsections and series of Euphorbia subg. Athymalus occurring naturally in southern Africa. Annuals 0 0 0 0 0 0 0 0 0 0 4 0

Perennial, non-succulent herbs 0 0 0 0 0 0 0 0 0 0 0 1

Woody shrub to tree 0 0 0 0 0 0 0 0 0 0 0 2

Succulents 67 4 23 4 14 0 5 33 0 7 0 0

Geophytes 6 2 3 0 0 3 0 0 1 0 0 0

2.1 Sect. Anthacanthae 59

60

2  Euphorbia subg. Athymalus

Key to the subsections of sect. Anthacanthae 1. Geophytes, with rosettes of deciduous leaves on or near surface of ground.............................2.1.2. subsect. Florispinae 1. Shrubs or dwarf succulents, if geophytic then with annual branches above ground.........................................................2. 2. Geophytes, with subterranean tuber and annual branches above ground, leaves prominent and deciduous......................3. 2. Non-geophytic dwarf or shrubby succulents, with perennial branches above ground (branches sometimes small but not annual) with green or grey-green photosynthetic bark on stems and branches (or with shiny grey-brown non-photosynthetic bark in E. quadrata), leaves ephemeral and often minute, always borne on distinct tubercles (small but clearly present in E. quadrata), tubercles often arranged into straight or spiral rows or fused into angles along the stem and branches............4. 3. Cyathial glands with obscurely crenulate outer margins......................................................2.1.4. subsect. Platycephalae 3. Cyathial glands with many finger-like, often forked processes on outer margins..................2.1.1. subsect. Dactylanthes 4. Cyathia unisexual..............................................................................................................................................................5. 4. Cyathia bisexual................................................................................................................................................................6. 5. Stems and branches with ± prominent tubercles arranged in rows or with indistinct tubercles fused into 4–20 often prominent angles; branches usually with many sterile short-shoots each arising in the axil of a tubercle and often developing into spines...............................................................................................................2.1.2. subsect. Florispinae 5. Stems and branches with prominent tubercles randomly arranged or very roughly organized into 3 rows (occasionally tubercles indistinct but then not fused into rows)..........................................................2.1.5. subsect. Pseudeuphorbium 6. Some branches elongating and terminated by cyathium on fertile short-shoot, fertile short-shoots (at least near their bases) indistinguishable from other branches....................................................................................................................7. 6. Cyathia on fertile short-shoots arising in axils of tubercles on branches and not terminating branch, fertile short-shoots different in appearance to branches...................................................................................................................................8. 7. Plants highly rhizomatous or slender and scandent from a turnip-shaped tuber (E. bruynsii), with green branches and often prominent tubercles on branches bearing leaves or leaf-rudiments that persist for most of growing season (forming large and very dense mound of short bright green branches in E. polycephala, smaller in E. globosa), often with fingerlike processes on outer margins of cyathial glands corrugated above, inner margins of glands folded outwards over base of gland (absent in E. bruynsii)..................................................................................................2.1.1. subsect. Dactylanthes 7. Plants not rhizomatous (if with much swollen rootstock then with ± shiny grey-brown non-photosynthetic bark and small tubercles on branches in E. quadrata) often with pencil-shaped grey-green branches with indistinct tubercles bearing fleeting leaves or leaf-rudiments........................................................................................2.1.5. subsect. Pseudeuphorbium 8. Plants with thickened, coarsely tuberculate stem with more finely tuberculate and more slender branches arising near its apex (occasionally plant densely mound-like or shrubby with stem not visible as in E. braunsii, E. clavarioides)....................................................................................................................2.1.3.subsect. Medusea 8. Plants consisting of a single erect cylindrical, tuberculate stem with few to many branches, tubercles on stem and branches all of similar size, branches of similar thickness to stem...................................................................................9. 9. Cyathial glands with entire to finely crenulate outer margins; fertile short-shoots 0.1–6 (16) cm long.........2.1.2. subsect. Florispinae 9. Cyathial glands with deeply toothed outer margins; fertile short-shoots 3–20 cm long.................................................10. 10. Plants with prominent tuft of deciduous leaves at apex of stem (and branches, if present); fertile short-shoots with conspicuous deltate bracts 20–40 mm broad.................................................................2.1.5. subsect. Pseudeuphorbium 10. Plants with ephemeral and relatively inconspicuous leaves at apex of stem; fertile short-shoots with linear to ± spathulate inconspicuous bracts 1–4 mm broad...................................................................................................2.1.3. subsect. Medusea

2.1.1 Subsect. Dactylanthes Euphorbia subsect. Dactylanthes (Haw,) Pax & K.Hoffm. in Engl., Nat. Pflanzenfam., ed. 2, 19c: 216 (1931). Euphorbia sect. Dactylanthes (Haw.) A.Berger, Sukkul. Euphorb.: 10, 104 (1906). Dactylanthes Haw., Syn. Pl. Succ.: 132 (1812). Type (designated by Wheeler 1943): Dactylanthes anacantha (Aiton) Haw. (= Euphorbia patula ssp. anacantha (Aiton) Bruyns. Euphorbia sect. Trichadenia Pax in Engl., Veg. Erde 9, 3 (2): 152 (1921). Euphorbia subg. Trichadenia (Pax) S. Carter,

Kew Bull. 40: 816 (1985). Type (designated by Wheeler 1943): Euphorbia trichadenia Pax. Euphorbia sect. Euphorbium Boiss. in DC., Prodr. 15 (2): 10, 85 (1862). Type (designated by Wheeler 1943): Euphorbia anacantha Aiton (= Euphorbia patula Mill.). Bisexual geophytes with herbaceous stem and deciduous branches or small, rhizomatous to large mound-forming succulents with short tuberculate stem and branches arising from tubers or underground rhizomes. Leaves

2.1 Sect. Anthacanthae

deciduous, to 80 mm long on non-tuberculate branches or reduced to minute fleeting rudiments at apex of conical tubercles (tubercles arranged loosely into spiralling rows along branches), alternate, entire, shortly petiolate or sessile, with small rod-like stipules or estipulate. Synflorescences terminal on stem or branches, with solitary bisexual cyathium 4.5–19 mm diam. with small leaflike bracts (other cyathia sometimes arising in axils of uppermost bracts) with glands (4) 5 (6), usually reticulated-corrugated above, with rectangular-truncate inner margins recurved over basal part of gland, outer margins divided into several finger-like irregularly reticulated-corrugated processes or rarely only slightly indented or ± entire. Capsule obtusely 3-lobed, smooth and glabrous, 5–8 mm diam., ± sessile to exserted up to 15 mm. Seeds smooth to finely papillate, ± 3 mm long, without caruncle.

61

With the re-alignment of the names in E. patula (Bruyns 2018), this subsection contains six species. One of them (E. trichadenia) is widespread in Angola, Botswana, South Africa and Zimbabwe and the others are confined to South Africa (Fig.  2.2). Two species (E. pseudotuberosa and E. trichadenia) are non-succulent geophytes from the tropical to sub-tropical parts of southern Africa. The other four are rather similar-looking, often rhizomatous succulents with tuberculate branches and greatly reduced leaves. They inhabit the karroid parts of South Africa. Note: Wheeler (1943: 490) stated that Hässler (1931: 318) had chosen Euphorbia trichadenia as the type of Pax’s sect. Trichadenia. However, no trace of this selection could be found on page 318 or anywhere else in Hässler’s account. Therefore, this selection is considered to have been made by Wheeler in 1943.

Fig. 2.2.  Distribution of Euphorbia subsect. Dactylanthes in southern Africa, showing number of species per half-degree square, with succulents in bold, rest geophytes (© PVB).

62

2  Euphorbia subg. Athymalus

Key to the species of Euphorbia subsect. Dactylanthes 1. Geophytic herb, with subterranean tuber and non-succulent annual branches above ground..............................................2. 1. Non-geophytic dwarf succulent (though usually with tuberous rootstock), with succulent perennial branches above ground (sometimes small but not annual)............................................................................................................................3. 2. Cyathial glands with 2–6 prominent finger-like processes on outer margins with corrugated upper surfaces, ovary soon exserted from cyathium on finely pubescent pedicel 5–7 mm long, leaves with small rod-like stipules at bases................. ................................................................................................................................................................E. pseudotuberosa 2. Cyathial glands with (3) 6–12 finely and irregularly divided finger-like processes on outer margins with smooth upper surfaces, ovary included within cyathium and raised on glabrous pedicel 1.5–4 mm long, leaves estipulate...........................E. trichadenia 3. Plants with few slender and scandent branches from a turnip-shaped tuber, cyathial glands with entire outer margins, ovary of female floret not pushed out far beyond cyathial lobes, inner margins of glands not folded back over upper surface of gland...................................................................................................................................................E. bruynsii 3. Plant highly rhizomatous or forming very dense mound of short bright green branches, outer margins of cyathial glands with prominent finger-like processes heavily corrugated above, inner margin of each gland folded back over upper surface of gland...................................................................................................................................................4. 4. Plant forming dense, cushion-like flat mat to hemispherical mound, not spreading by rhizomes......................................5. 4. Plant forming small clusters of branches, often spreading underground with rhizomes to form further such clusters....................................................................................................................................................................E. patula 5. Branches spherical, leaf-rudiments 2–5 × 1–2 mm, lanceolate, processes on outer margins of cyathial gland about twice as long as rest of gland.........................................................................................................................................E. globosa 5. Branches conical to cylindrical and tapering towards tips, leaf-rudiments 3–10 × 2–6 mm, elliptical to nearly circular, processes on outer margins of cyathial glands about as long as rest of gland..............................................E. polycephala

Euphorbia bruynsii L.C.Leach, J. S. African Bot. 47: 103 (1981). Type: South Africa, Cape, Steytlerville, Bruyns 1814 (PRE, holo.; SRGH, iso.). Bisexual spineless and glabrous dwarf succulent 20–300  mm tall from spherical to turnip-shaped subterranean tuber (up to 150  mm beneath surface), with slender simple or sparingly branched stem, tapering beneath tuber into fibrous roots. Stem and branches erect to scandent in surrounding bushes, 20–300 × 3–8 mm, irregularly cylindrical, with scattered tubercles, finely and sparsely pubescent, grey-green with fine longitudinal red lines (appearing brownish green if exposed); tubercles conical, 2–4 mm long and merging into stem, spirally arranged, with fine grooves around upper edges; leaves on tips of new tubercles on growing stem and branches, 5–25 × 2–4 mm, ascending, deciduous, lanceolate to narrowly obovate, entire, acute, ± sessile. Synflorescences 1 per branch at apex, each of 1–4 bisexual cyathia, consisting of one terminal cyathium on slender peduncle 2–6 mm long with 2–3 obovate-lanceolate glabrous bracts 3–10 × 2–3 mm in axils of which further pair of bisexual cyathia arise on branches up to 60 mm long with bracts in several opposite pairs (in axils of pair of bracts beneath cyathium further cyathia may develop similarly); cyathia cupular, glabrous, 3–4 mm broad (3 mm long below insertion of glands), with 5 lobes with finely incised margins, grey-green; glands 5, transversely reniform or elliptic,

± 1.5 × 0.5 mm, contiguous, ascending to spreading, slightly convex and pitted above, green becoming reddish on edges, outer margins spreading and slightly indented otherwise entire; stamens with pubescent reddish pedicels with red anthers, bracteoles filiform and pubescent; ovary ellipsoidal, glabrous, green faintly suffused with red, just included in cyathium on glabrous pedicel 2–3 mm long; styles 2.5–3 mm long, branched and widely diverging above middle. Capsule 6  mm diam., obtusely 3-angled, dull green, glabrous, ± sessile. Distribution & Habitat Euphorbia bruynsii occurs on a low conglomerate and pebble-covered hillock and in the surrounding loamy flats just south-east of the little town of Steytlerville. Specimens are found inside other shrubs, climbing among their branches for support. These shrubs, which are species of Eriocephalus, Pteronia and Rhigozum obovatum, are mostly not more than 30 cm tall and the habitat is rich in succulents, with Faucaria, Glottiphyllum, Rhombophyllum dolabriforme of the Aizoaceae, Ceropegia caespitosa, C. geminata, Fockea comaru, F. edulis, and Raphionacme zeyheri of the Apocynaceae and several species of Euphorbia present. Diagnostic Features & Relationships In E. bruynsii each plant arises from a neat, considerably swollen, ellipsoidal tuber that is located some distance below

2.1 Sect. Anthacanthae

63

Fig. 2.3.  Distribution of Euphorbia bruynsii (© PVB).

the surface of the ground, with a slender, often unbranched stem rising from the tuber to the surface. The above-ground portion of the plant is particularly insignificant, with slender branches clambering around among the twigs of the surrounding shrub, much like the stem and branches of various species of Ceropegia. The branches are greyish green, becoming more brownish if exposed to direct sunlight. They are covered loosely with low tubercles and on the tip of each of these a conspicuous leaf arises, which persists for most of the season in which it was produced. The leaves are also grey-green, becoming more glaucous if exposed to direct sunlight. During periods of vigorous growth (over most of the summer months, at least in cultivation; in habitat only when rain is received) a terminal cyathium develops at the tip of each of the longer branches and is surrounded by up to three small, leaf-like bracts. Further cyathia may develop in the axils of these bracts on quite long and slender branchlets, whose length and thickness depends on the amount of shelter received. Here again each cyathium is subtended by (this time usually two) leaf-like bracts and further cyathia may develop from their axils. In this way an eventually quite extensive, more or less dichotomously branched structure may arise. The cyathia are small and unusually long relative to their breadth, with an inconspicuous, grey-green colour. This insignificant colour extends into the small, transversely elliptic, green glands with entire but sometimes reddish outer margins. The capsule develops on a short, erect pedicel which, however, does not push it out of the cyathium so that at maturity it remains close to the old cyathium. While the cyathium is comparatively small, the mature capsule is of a size typical for many of the species to which E. bruynsii is related.

Fig. 2.4.  Euphorbia bruynsii, PVB 1814, SE of Steytlerville, South Africa (© PVB).

64

2  Euphorbia subg. Athymalus

Fig. 2.5.  Euphorbia bruynsii. A, portion of stem with leaf-rudiments (scale 4 mm). B, C, cyathium from side (scale 2 mm, as for C). D, anther (scale 1 mm, as for E). E, side view of dissected cyathium (with bracteoles). Drawn from: PVB 1814, SE of Steytlerville, South Africa (© PVB).

With its slender stems bearing fairly prominent and rather scattered tubercles, this species resembles a very much etiolated form of E. patula, a species which occurs nearby. Leach (1981), however, considered that it was most likely to be related to E. gariepina, in view of the similar, irregularly tuberculate stems, the somewhat similar leaves, the small cyathia and the very similar seeds. Contrary to this, our analysis of molecular data (Bruyns et al. 2006) indicated that E. bruynsii is closely related to E. globosa and E. polycephala, both of which are in turn closely allied to E. patula. The swollen tuber beneath the ground, with the slender stem arising from this, is unlike anything else known among these species but the stems are not dissimilar to those of E. patula, though the leaves in these species are far smaller and the tubercles are usually considerably more prominent. However, florally this relationship is unsuspected. This is especially so since the cyathia of E. bruynsii are much smaller, their glands have entire margins, lacking the prominent, brightly coloured, corrugated processes on their outer margins as well as lacking the recurved inner margin. Furthermore, the ovary is not exserted nearly as far from the cyathium as is typical for E. globosa or these other species. Geographically, however, this relationship does make sense since both E. globosa and E. patula are found not far from E. bruynsii, while E. gariepina is known only from a long distance off in the north-western parts of South Africa.

History Euphorbia bruynsii was first found on the 8th of December 1978, on the same field-trip during which the type of E. albipollinifera was collected and Haworthia bruynsii was found. It was gathered again near the type locality in 1998, when a few plants were located only with great difficulty, for the area was very overgrazed and degraded, with many of the small shrubs in rather desperate straits. The drawings published here were made from these plants. No other records have been made of this very inconspicuous species. Euphorbia globosa (Haw.) Sims, Bot. Mag. 53: t. 2624 (1826). Dactylanthes globosa Haw., Philos. Mag. J. 62: 382 (1823). Medusea globosa (Haw.) Klotzsch & Garcke, Abh. Königl. Akad. Wiss. Berlin 1859: 61 (1860). Lectotype (Bruyns 2012): South Africa, Cape of Good Hope, received 1821, Bowie (painting number 808/15 by T. Duncanson at K). Euphorbia glomerata A.Berger, Sukkul. Euphorb.: 104 (1906). Type: South Africa, Cape (missing). Bisexual spineless glabrous dwarf succulent 20–100  mm tall and forming dense mat 50–300 mm in diameter with many short densely packed branches, arising from small stem 10–50 mm tall above surface and 8–30 mm thick, sunken into ground and tapering beneath into fibrous roots. Branches erect,

2.1 Sect. Anthacanthae

rebranching extensively (upper branches, if present, more slender than those at ground level), 20–60 × 5–30 mm, spherical to clavate and tapering into slender neck towards bases, covered with few tubercles, smooth, green to yellow-green and becoming covered with grey corky bark with age; tubercles broad and flattened to occasionally conical, 1–3  mm long with (sometimes elongated) ± hexagonal base, vertically arranged into 3–7 obscure spiralling rows, with fine grooves between them; leafrudiments on tips of new tubercles towards apices of branches and of stem, 2–5 × 1–2 mm, ascending to spreading, fleeting, lanceolate, entire (minutely toothed along margin), acuminate, sessile. Synflorescences 1–2 per branch, with branch tapering or abruptly changing into synflorescence, sometimes arising in axils of tubercles very near apex of branch, each of 1 terminal bisexual cyathium on fleshy (later becoming slender and often somewhat wiry) peduncle 2–60 × 1–2 mm, sometimes reaching 160 mm long by branching from axils of uppermost bracts and then producing succession of further 2–3 cyathia (cyathia on branches of peduncle sometimes all male), with 2–3 ovatelanceolate to narrowly oblanceolate glabrous bracts 2–7 × 1–2 mm subtending cyathium and further 2–4 shorter and narrower bracts along peduncle; cyathia conical, glabrous, 10–20  mm broad (2–4  mm long below insertion of glands), with 5 lobes with finely incised margins, dark grey-green to green; glands 5 (1 sometimes smaller than others), ± cuneate,

65

5–7 mm long, ascending, above reticulated with white to cream or greenish ridges (especially towards edges) on green to yellow background; inner margins white or cream, rectangulartruncate and incurved over gland (sometimes finely pubescent); outer margins divided into 2–5 ascending and slightly spreading finger-like processes with irregularly white- to green-margined pits in lower two thirds (pits disappearing towards tips); stamens few with sparsely pubescent slightly reddish pedicels and pinkish yellow anthers, bracteoles very few to absent, filiform and pubescent only near apices; ovary ellipsoidal, glabrous, grey-green suffused with red, initially erect and almost exserted from cyathium (on pedicel 1.5–5  mm long) then becoming pushed out of cyathium on glabrous spreading to later decurved reddish pedicel 4–7 mm long; styles 3–7 mm long, branched to below middle. Capsule 6–7  mm diam., obtusely 3-angled, glabrous, exserted on decurved and later erect pedicel 4–7 mm long. Distribution & Habitat Euphorbia globosa is known only in a small area between Port Elizabeth, Uitenhage and Addo. Much of its habitat no longer exists and was destroyed by the enormous and recent expansion of the city of Port Elizabeth, though several small populations have survived within the limits of the city and on its outskirts around Uitenhage.

Fig. 2.6.  Distribution of Euphorbia globosa (© PVB).

Euphorbia globosa occurs in flat spots and on gently sloping, low hills among alluvial gravel and pebbles, usually in somewhat open patches between dense bushes. Larger plants are often completely in the open but smaller specimens may be sheltered by low bushes or stones. These open patches between the denser bush harbour a wealth of

small succulents. Euphorbia globosa often grows together with such species as E. clava, E. meloformis, E, rhombifolia and E. stellata, as well as other succulents of the Aizoaceae (especially species of Corpuscularia, Delosperma echinatum and Glottiphyllum) and of the Crassulaceae.

66

2  Euphorbia subg. Athymalus

Fig. 2.7.  Euphorbia globosa, west of Uitenhage, South Africa, 20 May 2012 (© PVB).

Diagnostic Features & Relationships Euphorbia globosa is a most characteristic species, with its profusion of round branches that develop from the small, hidden and usually mostly subterranean stem. The lower branches close to the ground often root as well and are usually almost spherical, tapering abruptly towards their bases

into a thin neck joining them to the stem. These branches may be produced in very large numbers and are densely packed into a tight mat up to 30 cm in diameter. Although they are a dull green when young, in exposed plants many of the lower ones become covered with a grey, corky bark. Many of the nearly spherical, lower branches produce a

Fig. 2.8.  Euphorbia globosa, clump ± 30 cm diam., west of Uitenhage, South Africa, 20 May 2012 (© PVB).

s­ uccession of one to three (or even four or more) secondary branches from the axils of some of the tubercles near their apex. These secondary branches are progressively more slender and considerably longer (often somewhat more club-

shaped or even cylindrical) than the lower ones from which they arose. In habitat the secondary branches may be short, but they can become very long if sheltered. They are often grazed off.

2.1 Sect. Anthacanthae

Cyathia are produced in E. globosa over much of the summer, whenever water is received. They arise on a terminal peduncle (or sometimes in the axils of the uppermost tubercles, generally those right next to the apex of the branch). Some peduncles remain short but many of them persist and gradually may become anything between 5 and 16 cm long and exceedingly slender (often somewhat wiry, though not at all rigid), branching repeatedly successively from near their apex. The peduncle has a few small, greygreen bracts along its length and is terminated by a cyathium that is subtended by two to three slightly larger bracts. Further cyathia may develop on peduncles arising in the axils of these larger bracts and they usually only develop once the previously terminal cyathium has fallen off. Here the cyathia are comparatively large, but are mostly a dark, almost bluish grey-green suffused with red. They have especially long and slender processes on the outer margins of the glands with a much pitted upper surface. Each bears a female floret that is initially erect (closely enclosed by the quite large lobes), but it is soon pushed to one side and somewhat beyond the lobes on a spreading pedicel. Relatively few male florets develop and these usually bear pale, creamcoloured pollen. Related to E. polycephala and E. patula, E. globosa is most similar in its form of growth to E. patula. From this it differs by the lack of a rhizomatous habit, the much denser clustering of the branches (though they are not as densely packed as in E. polycephala), the almost spherical shape of the lower branches and the club-shaped form that they assume when they become more elongated (narrowing strongly to

Fig. 2.10.  Euphorbia globosa, PVB 12135, Despatch, near Uitenhage, South Africa, Jan. 2014 (© PVB).

67

Fig. 2.9.  Euphorbia globosa, PVB 6877, Soutkloof, north of Uitenhage, South Africa, 3 Dec. 2005 (© PVB).

68

2  Euphorbia subg. Athymalus

Fig. 2.11.  Euphorbia globosa. A, upper portion of branch with leaf-rudiments (scale 3 mm). B, cyathium from side (scale 2 mm). C, cyathial gland from above (scale 1 mm, as for D, E). D, E, female floret. Drawn from: PVB 12135, Despatch, near Uitenhage, South Africa (© PVB).

their bases) and also by their less prominent tubercles. Nevertheless, when sheltered, the branches become slender and may be difficult to separate from those of E. patula. The leaf-rudiments are also generally smaller in E. globosa. In E. globosa the processes on the outer margins of the cyathial glands are particularly slender and greenish and are most similar to those of E. patula subsp. patula. History Euphorbia globosa was first sent to England by James Bowie in 1821, after which it soon became popular and fairly widespread in cultivation. Bowie must have collected it late in 1820 or during January 1821 (Gunn and Codd 1981). Haworth saw it growing at Kew and described it in 1823 as Dactylanthes globosa. A figure of it was published in the Botanical Magazine in 1826 and Sims used this opportunity to move it to Euphorbia. This figure was drawn from material grown by Mr Hood, a surgeon from South Lambeth in London. He cultivated many rare and unusual succulents and the genus Hoodia (Apocynaceae, now included in Ceropegia) was named after him.

Fig. 2.12.  Euphorbia globosa, South Africa, Cape of Good Hope, received 1821, Bowie. Watercolour 808/15 by T. Duncanson (© Royal Botanic Gardens, Kew).

2.1 Sect. Anthacanthae

69

Euphorbia patula Mill., Dict., ed. 8: Euphorbia no. 11 (1768). Dactylanthes patula (Mill.) Haw., Syn. Pl. Succ.: 132 (1812). Medusea patula (Mill.) Klotzsch & Garcke, Abh. Königl. Akad. Wiss. Berlin 1859: 61 (1860). Neotype (Bruyns 2012): South Africa, Cape, sheet 328, specimen (one of two) labelled ‘Grimwood’s St’ by Haworth (OXF).

finely pubescent spreading to later decurved pale green to faintly reddish pedicel 4–6  mm long; styles 4–7  mm long, branched and widely spreading in upper third. Capsule 6–7 mm diam., obtusely 3-angled, glabrous, green, exserted from cyathium on decurved and later erect pedicel 4–6 mm long.

Bisexual spineless glabrous dwarf succulent 20–150 mm tall and forming dense to diffuse mat 50–300 mm in diameter with many short (sometimes densely packed) branches forming extensive underground network from rhizomes, arising from small stem 10–50 × 8–20  mm above surface, sunken into and swollen under the ground, with tap-root 20–50 mm thick tapering beneath into fibrous roots. Branches erect to spreading, simple to rebranching (sometimes extensively) if sheltered, 10–150 (400) × 5–15 (30) mm, globose to cylindrical and often tapering from thick base towards tips, covered with few tubercles, smooth, purplish to dark green to grey-green; tubercles conical, 3–8 mm long with elongated though sometimes very small hexagonal base, vertically arranged into 3–5 obscure spiralling rows, with fine grooves between them; leaf-rudiments on tips of new tubercles towards apices of branches and of stem, 2–6 × 1–2 (4) mm, ascending to spreading, fleeting, linear to subulate, channelled above, minutely toothed along sometimes reddish margin, acute, sessile but tapering towards base and merging into apex of tubercle. Synflorescences 1 per branch at apex with branch tapering into synflorescence, each of 1 terminal bisexual (unisexual) cyathium on fleshy peduncle 6–30 (100)  mm long, sometimes branching from axils of uppermost bracts and then producing succession of further 2–4 or more cyathia, with 2–5 ovate, ovate-lanceolate to narrowly oblanceolate glabrous bracts 2–7 × 1–2 mm subtending each cyathium and further 2–4 shorter and narrower bracts along peduncle; cyathia shallowly conical to cupular, glabrous, 10–17 mm broad (2–3 mm long below insertion of glands), with 5 pale green to cream lobes with deeply incised margins, pale green; glands 4–5 (6), ± cuneate, 3–9 mm long, above heavily reticulated with cream to yellow ridges on green background, with inner margins extended into white rectangular–truncate process recurved over darker basal part of gland, outer margins divided into 2–6 ascending to spreading finger-like processes 2–7 mm long with irregularly reticulated-corrugated and pitted cream to pale green, brownish green or yellow upper surface; stamens with sparsely pubescent pedicels suffused with red towards apices, bracteoles very few, filiform and sparsely pubescent only near apices; ovary ellipsoidal and obscurely 3-angled, glabrous, initially erect and almost exserted from cyathium (on pedicel ± 2 mm long) then becoming pushed out of cyathium on glabrous to

Euphorbia globosa and E. polycephala are similar to and closely related to E. patula and the differences between them are discussed under these other species. Euphorbia bruynsii is also closely related but is less similar and their differences are discussed under E. bruynsii. Euphorbia patula was discovered very early in the exploration of the Cape of Good Hope. The first figure known is one probably made by Hendrik Claudius in 1686 or 1687. What may be the original of this is Folio 188 among 343 paintings and studies of Cape Plants in the ‘Icones Plantarum et Animalium’ (or IPA, which is located in the ‘Africana Museum’ of the Johannesburg Municipal Library, South Africa), though all of these paintings are without headings, signatures or dates (Macnae and Davidson 1969). This is the most complete known collection of the illustrations of Cape wildlife that was commissioned by Nicolaas Witsen and is often known as the Codex witsenii. Many of these figures were made during the expedition of Simon van der Stel to the Copper Mountains in 1685 (Wilson et  al. 2002: 21). Copies of these paintings were supplied to various people and a mirror image of Claudius’ illustration of E. patula was reproduced by Johannes Burman in 1738. Another, quite different figure was published by D’Isnard (1720). These early figures are not detailed but they show the rather irregular growth of many branches with low tubercles, the very short peduncles and that of D’Isnard also shows some details of the cyathia, including the long, furrowed processes on the outer margins of the glands. Aiton (1789) cited the figures of Burman and D’Isnard under his name E. anacantha (which he took from D’Isnard’s phrase-name for the plant) but he referred to the ‘flowers’ as terminal, solitary and sessile, which is true of the plant depicted in the IPA, rather than that of D’Isnard, where one terminal cyathium is solitary while the other is surrounded by three further cyathia on obvious peduncles. In D’Isnard’s figure the branches are quite slender, they are also extensively rebranched well above the ground and they exhibit more pointed tubercles than in Claudius’ figure. Linneaus (1753) included E. patula (in the form of Burman’s and D’Isnard’s figures but at that stage without a name) as varieties under E. caput-medusae. He treated the figures of D’Isnard (var. β) and Burman (var. γ) separately.

2  Euphorbia subg. Athymalus

70

Fig. 2.13.  Distribution of Euphorbia patula (▼ = ssp. anacantha; ■ = ssp. brucebayeri; ● = ssp. patula; ▲ = ssp. wilmaniae) (© PVB).

Here four subspecies are recognized and they may be separated as follows: 1. Peduncles usually producing several cyathia and with time developing slender branches, rough upper surface of processes of glands cream, yellowish, brownish green or green........................................................................................2. 1. Cyathia usually solitary at tips of branches, rarely with others developing on short branches 5–20 mm long beneath terminal cyathium, rough upper surface of processes of glands mostly cream.....................................................3. 2. Plant with several branches, often rhizomatous from central tuber, outwardly folded flap of inner margin of glands ± square, upper surface of processes of glands cream, yellowish or green......................................subsp. patula 2. Plant with few branches rising directly to surface from tuber and not rhizomatous, outwardly folded flap of inner margin of glands ± rectangular and broader than long, upper surface of processes of glands brownish green............subsp. brucebayeri 3. Plant forming clump with occasional short rhizomes, cyathia solitary, 10–17 mm diam., outwardly folded flap of inner margin of glands ± square.........................................................................................................................subsp. anacantha 3. Plant diffusely spreading with many slender underground branches and rhizomes, terminal cyathium sometimes with additional cyathia developing on short branches 5–20 mm long, cyathia 6–10 mm broad, outwardly folded flap of inner margin of glands ± rectangular and broader than long..............................................................................subsp. wilmaniae

Euphorbia patula subsp. patula Euphorbia tridentata Lam., Encycl. 2 (2): 416 (1788). Medusea tridentata (Lam.) Klotzsch & Garcke, Abh. Königl. Akad. Wiss. Berlin 1859: 61 (1860). Type: South Africa, collector unknown (P-LAM, holo.; K, iso.). Euphorbia ornithopus Jacq., Fragm. Bot.: 76, t. 120, fig. 2 (1809). Euphorbia tridentata var. ornithopus (Jacq.) Van Veldh. & Lawant, Euphorbia World 10: 9 (2014). Lectotype (Bruyns 2012): Jacq., Fragm. Bot.: t. 120, fig. 2 (1809).

Euphorbia leachii Lawant & Van Veldh., Euphorbia World 10: 5 (2014). Type: South Africa, Eastern Cape, along N10 highway SE of Cradock, Becker & Möller 1326 (UNIN, holo.; PRE, iso.). Plant often forming small clusters of branches connected by rhizomes radiating from central tuber. Branches erect, 20–150 × 5–12 mm, globose to cylindrical and often tapering towards tips, greyish green; tubercles flattened-conical,

2.1 Sect. Anthacanthae

2–8 mm long. Synflorescences 1 to many per branch at apex with branch tapering shortly into synflorescence, each of 1 terminal bisexual cyathium on fleshy peduncle 10–100 mm long, often branching from axils of uppermost bracts and then producing succession of further 2–4 or more cyathia, with larger bracts 2–7 × 1–2 mm; cyathia 10–17 mm broad; glands 4–5, ± cuneate, 3–8 mm long, above heavily reticulated with cream or pale yellow to orange ridges on green background, with white ±  square inner margins recurved over darker basal part of gland, outer margins divided into 2–5 spreading often very slender finger-like processes 3–6 mm long. Distribution & Habitat Subsp. patula is found in several widely separated areas in slightly arid to semi-arid parts of the former Cape Province of South Africa. On the Klein Karoo it is known in a small area to the south of Calitzdorp. It is also of very sporadic, but fairly wide occurrence on the Steytlerville Karoo. On the Great Karoo it is known around Beaufort West, both a little south of the town and to the north near the foot of Molteno Pass. It is probably most common in the Eastern Cape

71

between Cradock and Somerset East along the Fish River Valley and eastwards to Grahamstown, Fort Beaufort and Kingwilliamstown. Plants found in the hills south of Patensie are also referable to subsp. patula. Euphorbia patula subsp. patula is usually found in flat areas or on gently rolling hills, where it occurs in hard, loamy ground. Sometimes this is of alluvial origin with many rounded pebbles, both in the ground and on its surface (as in the Fish River Valley). It is also often encountered between low outcrops of vertical shale strata, sometimes tightly wedged into the crevices between these strata, sometimes in apparently more open patches between these outcrops, where the strata are subterranean (and the rhizomes are again wedged among the strata). Plants may shelter beneath small bushes but are frequently fully in the open and usually occur with a wealth of other small succulents. In some spots it has been seen together with Euphorbia huttoniae, E. meloformis, E. rhombifolia and E. stellata. Other small succulents such as Ceropegia modesta, species of Bergeranthus, Chasmatophyllum, Hereroa, Mestoklema, Ruschia (of the Aizoaceae), and the ubiquitous Senecio radicans abound in these habitats.

Fig. 2.14.  Euphorbia patula ssp. patula, among short scrub, PVB 12127, NW of Grahamstown, South Africa, 3 Apr. 2012 (© PVB).

Fig. 2.15.  Euphorbia patula ssp. patula, among short scrub looking just like a small stapeliad, SW of Cradock, South Africa, 2 May 2008 (© PVB).

Diagnostic Features & Relationships In E. patula subsp. patula the plant consists of a central stem, whose below-ground parts may be considerably swollen into a tuber-like structure. From this, the branches radiate extensively underground by means of rhizomes (Fig. 1.11B) and so the central stem and tuber is mostly very hard to locate and may even die off while the plant continues to flourish from rooted branches. Some of the rhizomes may form

smaller, subsidiary root-systems, giving rise to new clumps of branches and further rhizomes. Usually only short portions of the branches appear above the surface, but their length depends greatly on the amount of shelter received. Consequently, when they are fully exposed, they may be as little as 20 mm long and up to 15 mm thick while, if sheltered, they may reach more than 10 cm long. The usually dull dark green branches are gathered into small clumps (occa-

72

2  Euphorbia subg. Athymalus

sionally larger mats up to 30 cm broad may develop). They are covered with rather randomly arranged, low to prominent tubercles (these are especially prominent and slender – up to 8 mm long – in plants from the Steytlerville Karoo and from near Beaufort West) surrounded by fine grooves. On newly

developed branches, each tubercle tapers into a small, narrow, leaf-rudiment. In habitat the plants may resemble one of the small, mat-forming stapeliads of Ceropegia sect. Duvalia or sect. Piaranthus. In subsp. patula flowering usually takes place in spring but may continue throughout the summer and into autumn if conditions are favourable. Generally only a few branches produce cyathia and do so by elongating into a short, stout peduncle, which is slightly more slender than the branch itself, with several leaf-like bracts along its length. There are between two and five slightly larger bracts in a rosette immediately below the terminal cyathium. Buds in the axils of these upper bracts are soon activated and more cyathia develop from them (occasionally even slender branches without cyathia may develop from them too). As a result the peduncle may elongate extensively by successive forking, even exceeding 10  cm long on occasion. The cyathia are comparatively large with prominent glands and a fairly small central part with the florets. When the female floret matures, it is wrapped up quite tightly in the overlapping, grey-green lobes (usually these overlap to the right when one looks towards the centre of the cyathium) and these gradually become pushed apart somewhat by the development of the relatively few male florets and a few filiform bracteoles. The male florets have reddish pedicels and red filaments, with red pores in the otherwise pale yellow anthers. Pollen is unusually variable in colour in subsp. patula: in some collections it is bright orange (as around Calitzdorp, PVB 8192, in plants from near Steytlerville and in those from near Beaufort West), while some have yellow pollen and yet others around Grahamstown produce cream or white pollen. In the earlier stages of the cyathium, the female floret is erect but soon its pedicel extends considerably and bends over to lift the ovary beyond the side of the cyathium. While the males mature, the ovary is held near the edge of the cyathium and the long styles are more or less outside the whole structure. Once the

Fig. 2.17.  Euphorbia patula ssp. patula, PVB 11052, SW of Cradock, South Africa, 18 Jan. 2014 (© PVB).

Fig. 2.18.  Euphorbia patula ssp. patula, PVB 11928, between Bedford and Grahamstown, South Africa, 2 Mar. 2013 (© PVB).

Fig. 2.16.  Euphorbia patula ssp. patula, with leaf-rudiments, mostly solitary cyathia and similar to ssp. anacantha, PVB 6860, south of Hankey, South Africa, 11 Feb. 2005 (© PVB).

2.1 Sect. Anthacanthae

capsule is mature, the pedicel gradually becomes erect again (though without elongating further). While the cyathium is mostly a dull grey-green, the glands may be more striking. They have broad, short bases and finger-like, spreading processes on the outer margins. These processes (most commonly three in number) are variable in length and become exceptionally slender, almost filiform in some plants from the Steytlerville district. Near their bases the glands are dark green above and the remainder of the

73

upper surface is heavily corrugated with a cream to pale green crust, which extends onto the long processes, where it also has numerous corrugations and round pits. Initially (and until the female floret begins to mature), the inner margins of the glands are pressed against the cyathial lobes. Later they gradually bend back over the gland to form an almost square, cream-coloured process that covers most of the base of the gland. When brightly coloured, these processes may somewhat resemble small petals (as in Fig. 2.19 below).

Fig. 2.19.  Euphorbia patula ssp. patula, with unusually dark orange pollen, PVB 11517, W of Steytlerville, South Africa, 10 Nov. 2012 (© PVB).

Fig. 2.20.  Euphorbia patula ssp. patula (A, C–E, G) and ssp. anacantha (B, F). A, B, cyathium from side (scale 2 mm, as for B–F). C–F, cyathial gland from above. G, female floret (scale 1 mm). Drawn from: A, C, D, PVB 5015, Hellspoort, near Grahamstown, South Africa. B, F, PVB 7641, Heidelberg, South Africa. E, PVB 11517, W of Steytlerville, South Africa. G, PVB 11052, SW of Cradock, South Africa (© PVB).

74

History The name Euphorbia patula Mill. has been a source of considerable confusion, as outlined in Bruyns (2018), where it was shown that this is the earliest valid name for what has usually been known as Euphorbia ornithopus or E. tridentata. In 2014 the new name Euphorbia leachii was described. However, all its distinctive features, as discussed by Lawant and Van Veldhuisen (2014a), show such an extent of overlap that there are no grounds for recognising it as a new species. Here Fig. 2.17 is from the area where this species was supposed to occur, but florally this is impossible to separate from Fig.  2.18, from near Grahamstown. Consequently, it was reduced to synonymy in Bruyns (2018) under subsp. patula. Plants that are included under subsp. patula from near Calitzdorp were first recorded there by P.A. van Breda in October 1957 (PRE records). The other outlying locality near Beaufort West was discovered by the herpetologist W.R. Branch around 1980. Euphorbia patula subsp. anacantha (Aiton) Bruyns, Haseltonia 25: 45 (2018). Euphorbia anacantha Aiton, Hort. Kew., ed. 1, 2: 136 (1789). Dactylanthes anacantha (Aiton) Haw., Syn. Pl. Succ.: 132 (1812). Lectotype (Bruyns 2012): J. Burm., Rar. Afric. Pl.: t. 7, fig. 2 (1738).

2  Euphorbia subg. Athymalus

mm; cyathia 10–22  mm broad; glands (4) 5, ±  cuneate, 3–8  mm long, above heavily reticulated with bright cream ridges on green background, with white ± square inner margins recurved over darker basal part of gland, outer margins divided into 3–5 spreading finger-like cream processes 3–6 mm long. Distribution & Habitat Subsp. anacantha is found only on the southern coastal plain of the former Cape Province, where it is recorded from Heidelberg sporadically eastwards via Riversdale to Mossel Bay. Subsp. anacantha usually grows on gently sloping areas, in hard, loamy ground of alluvial origin with many rounded pebbles both in the ground and on its surface, sometimes sheltering beneath small bushes. Where it occurs, there is mostly a wide selection of other small succulents belonging to the genera Anacampseros, Ceropegia, Crassula, Haworthia and several members of the Aizoaceae, such as Acrodon, Glottiphyllum, Ruschia and Trichodiadema.

Plant forming a dense cluster of branches with occasional short rhizomes. Branches erect, 20–150 × 5–12 mm, globose to cylindrical and often tapering towards tips, shiny bright green; tubercles flattened-conical, 1–3  mm long. Synflorescences 1 per branch at apex with branch tapering shortly into it, each of 1 terminal bisexual cyathium on fleshy peduncle 2–5 (20) mm long with larger bracts 2–7 × 1–2

Diagnostic Features & Relationships In subsp. anacantha the branches are shiny bright green, the tubercles rise very little out of the surface of the branches and do not taper into the bases of the leaf-rudiments, the peduncle is in most cases only 2–5 mm long and mostly bears only a solitary, terminal cyathium. This contrasts with subsp. patula, where the branches are dull, dark green (to purplish), the tubercles project considerably from the surface of the branches and taper into the bases of the leaf-rudiments, the peduncle exceeds 5 mm long and axils around the terminal cyathium give rise to other cyathia, so that some branched peduncles may be up to 10 cm long. In subsp. anacantha the cyathia emit a sweetish odour.

Fig. 2.21.  Euphorbia patula ssp. anacantha, large cyathium 20  mm diam., PVB 7641, Heidelberg, South Africa, 24 Dec. 2019 (© PVB).

Fig. 2.22.  Euphorbia patula ssp. anacantha, PVB 12140, NE of Albertinia, South Africa, 13 Mar. 2018 (© PVB).

2.1 Sect. Anthacanthae

Plants of subsp. anacantha are reputed to become very large. J. Dekenah found some covering an area of up to 4 × 6 feet in 1942 near Riversdale (PRE records), though such huge specimens have not been seen in recent years. With the

75

drastic reduction of the renosterveld on the southern coastal plain by agriculture, this subspecies is now quite rare and restricted to a few small patches of relatively undisturbed veld.

Fig. 2.23.  Euphorbia patula ssp. anacantha, an unusual peduncle with three cyathia (the first one already fallen off), cyathia at an early stage in their development and with only four glands, PVB 7641, Heidelberg, South Africa, 24 Dec. 2019 (© PVB).

History Although the name ‘anacantha’ is more recent than others in this complex, subsp. anacantha was the first member of this complex to be found and recorded. This is discussed in more detail above. This subspecies was illustrated under the name Euphorbia tridentata by Phillips (1925). Euphorbia patula subsp. brucebayeri Bruyns, Haseltonia 25: 46 (2018). Type: South Africa, Ladismith distr., southern foot of Touwsberg, 450 m, 28 July 1991, Bruyns 4732 (BOL, holo.; NBG, iso.). Plant with several short branches in small clump to 80 mm diam. arising from apex of irregularly shaped underground tuber to 100 × 50 mm. Branches erect, 10–30 (80) × 5–12 mm, cylindrical and often tapering towards tips, smooth, green; tubercles narrowly conical, ascending, 4–8 mm long; leaf-rudiments 2–4 × 1–2 mm, greyish green. Synflorescences 1 per branch at apex, each of 1–3 (5) bisexual cyathia, sessile or on fleshy peduncle 4–80 mm long, with larger bracts 2–5 × 3–4 mm; cyathia cupular, 6–15  mm broad; glands 4–5, 3–4 mm long, above finely reticulated with brownish green ridges on pale green background, with brownish white rectangular inner margins recurved over darker brown basal part of gland, outer margins divided into 3–6 spreading fingerlike brownish green processes 2–3 mm long.

Distribution & Habitat Subsp. brucebayeri is known from a single locality at the southern base of the Touwsberg on the Klein Karoo, where it grows tightly wedged into crevices in low outcrops of shale, among a scanty covering of low shrubs. Diagnostic Features & Relationships Unlike the other subspecies, subsp. brucebayeri does not spread by underground rhizomes and the branches rise directly to the surface of the ground from the tuber, where they form a small cluster. In habitat this cluster is not usually more than 50 mm in diameter and the branches rarely project from the ground for more than 20 mm. In cultivation they do not remain so short and may reach 10 cm long. The branches bear slender tubercles, longer than in the other subspecies of E. patula (except for plants of subsp. patula from the Steytlerville area, where they may be similar). Subsp. brucebayeri produces several cyathia per branch on gradually elongating, slender peduncles. Florally, subsp. brucebayeri differs from the other subspecies mainly by the much less corrugated upper surface of the glands and their processes and also in the fan-shaped rather than rectangular recurved base of the gland.

76

Fig. 2.24.  Euphorbia patula ssp. brucebayeri, PVB 4732, south of Touwsberg, Klein Karoo, South Africa, 24 Apr. 2014 (© PVB).

2  Euphorbia subg. Athymalus

Fig. 2.25.  Euphorbia patula ssp. brucebayeri, PVB 4732, south of Touwsberg, Klein Karoo, South Africa, 24 Apr. 2014 (© PVB).

Fig. 2.26.  Euphorbia patula ssp. brucebayeri. A, tip of branch with leaf-rudiments (scale 3 mm, as for B). B, cyathium from side. C, dissected cyathium from side (scale 2 mm, as for D). D, cyathial gland from above. Drawn from: PVB 4732, south of Touwsberg, Klein Karoo, South Africa (© PVB).

2.1 Sect. Anthacanthae

History This westernmost subspecies of E. patula was first observed by M. Bruce Bayer around 1990, at a farm where grazing experiments were conducted by the staff of the Department of Agriculture’s Worcester Veld Reserve. Euphorbia patula subsp. wilmaniae (Marloth) Bruyns, Haseltonia 25: 45 (2018). Euphorbia wilmaniae Marloth, S.  African Gard. 21: 133 (1931). Lectotype (Bruyns 2012): South Africa, Cape, Boetsap, Pagan sub Marloth 6125a (PRE). Euphorbia planiceps Marloth ex A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 963 (1941). Lectotype (Bruyns 2012): South Africa, Cape, farm near Griquatown, Sept. 1939, Venter (BOL).

77

ments 2–4 × 1–2 mm, greyish green. Synflorescences 1 per branch at apex, each of 1–3 (5) bisexual cyathia, sessile or on fleshy peduncle 4–20 mm long, with larger bracts 2–5 × 3–4 mm; cyathia cupular, 6–10 (15) mm broad; glands 5, 3–4 mm long, above heavily reticulated with cream ridges on green background, with white broadly rectangular inner margins recurved over dark purple-red basal part of gland, outer margins divided into 2–5 spreading finger-like cream processes 2–3 mm long.

Plant with many scattered rhizomatous branches forming clumps and occasionally forming densely packed cushionshaped mass up to 300 mm diam. Branches erect to spreading, 10–30 × 5–12 mm, ± cylindrical and tapering towards tips, dark green; tubercles conical, 2–6 mm long; leaf-rudi-

Distribution & Habitat Euphorbia patula subsp. wilmaniae occurs in Griqualand West, from north of Kuruman to close to Prieska and Douglas. It is not known outside this region. Subsp. wilmaniae always grows in hard loam on low, stony hills, where the branches are tightly wedged among stones and fill up crevices between them, spreading under them by means of rhizomes. In some areas it occurs on calcrete outcrops, while in others it thrives on hillsides covered with dark lumps of dolerite, metamorphised shales and chips of jasper.

Fig. 2.27.  Euphorbia patula ssp. wilmaniae, small clusters of branches in crevices in calcrete with Prepodesma orpenii, PVB 4480, Cotton End, Kuruman distr., South Africa, 25 Jan. 1991 (© PVB).

Fig. 2.28.  Euphorbia patula ssp. wilmaniae, larger mats of small branches in calcrete, PVB 4486, Postmasburg, South Africa, 16 Jan. 2009 (© PVB).

Diagnostic Features & Relationships Subsp. wilmaniae is separated from subsp. patula mainly by the more slender, more bluish green branches with slightly more prominent tubercles which have a narrower base. The cyathia of subsp. patula are often broader than those of subsp. wilmaniae, with longer p­ rojections on the

glands, but this feature is variable within subsp. patula and may not be reliable. Another difference is the purple-red basal part of the upper surface of the glands in subsp. wilmaniae, obscured by the recurved inner margins. This area is generally dark green in subsp. patula.

78

2  Euphorbia subg. Athymalus

Fig. 2.29.  Euphorbia patula ssp. wilmaniae, cyathia in early female stage, PVB 4486, Postmasburg, South Africa, 12 Nov. 2004 (© PVB).

Fig. 2.30.  Euphorbia patula ssp. wilmaniae, cyathia in male stage, PVB 4486, Postmasburg, South Africa, 13 Nov. 2015 (© PVB).

History ‘Euphorbia wilmaniae’ was first recorded by Maria Wilman in April 1921. According to Wilman (1946: 256), the type

specimen was collected by Mrs H.H.  Pagan and her sister Miss Keith on the farm Boetsap in the Barkly West district of Griqualand West in 1923.

Fig. 2.31.  Euphorbia patula ssp. wilmaniae, plant with both cyathia and capsules, between blocks of calcrete, PVB 4486, Postmasburg, South Africa, 16 Jan. 2009 (© PVB).

The name Euphorbia planiceps was given to plants forming a compact body of densely packed branches with relatively few spreading, underground rhizomes. In their peduncles and cyathia they ‘agree so closely...as to suggest that E. planiceps should be regarded as a variety of’ E. wilmaniae (White et  al. 1941: 523). Here E. planiceps is

treated as an ecologically induced, unusually dense form of subsp. wilmaniae and it is reduced to synonymy. These densely-growing forms (as seen here in Fig. 2.28) were discovered by Maria Wilman on the farm Banksfontein (in the Hay district of Griqualand West) in March 1921 (Wilman 1946).

2.1 Sect. Anthacanthae

79

Fig. 2.32.  Euphorbia patula ssp. wilmaniae. A, tip of branch with leaf-rudiments (scale 3 mm, as for B). B, cyathium from side. C, cyathial gland from above. D, anthers and bracteole (scale 1 mm, as for E). E, F, female floret. Drawn from: PVB 4480, Cotton End, Kuruman distr., South Africa (© PVB).

Euphorbia polycephala Marloth, S. African Gard. 21: 133 (1931). Lectotype (Bruyns 2012): South Africa, Cape, near Mortimer, Aug. 1913, Shoesmith sub Marloth 5295 (PRE). Bisexual spineless glabrous densely mat- or mound-forming succulent 20–500 mm × 0.2–1 m with very large numbers of tightly packed short branches developing from stem 30–200 mm thick usually wholly sunken into ground, stem often dividing below ground into several swollen roots ± 50–200 × 20–50 mm, tapering beneath into fibrous roots. Branches erect to spreading, simple to rebranching, 15–150 × (5) 10–20 mm, conical to cylindrical and tapering towards tips, tuberculate, smooth, green to yellow-green; tubercles conical, 2–6 mm long with elongated hexagonal base, vertically arranged into 3–5 very obscure spiralling rows, with fine grooves between them; leaf-rudiments on tips of new tubercles towards apices of branches, 3–10 × 2–6 mm, ascending to spreading, fleeting, elliptic to nearly circular, slightly concave above, minutely ciliate along margins, apiculate, sessile. Synflorescences 1 to several per branch at or near apex with branch tapering into terminal cyathium and

sometimes others arising later in axils of upper tubercles on branch, each of 1–4 bisexual cyathia on fleshy peduncle 2–30 mm long terminated by cyathium and with others arising in axils of bracts, with 2–6 ovate or elliptic glabrous or finely pubescent bracts 2–4 × 2–3 mm; cyathia shallowly conical to slightly cupular, glabrous, 8–19  mm broad (2–3 mm long below insertion of glands), with 5 red-green lobes with conspicuously white-ciliate margins, grey-green suffuse with red; glands 4–5, ± cuneate, 3–6 mm long, heavily reticulated above with cream ridges with grey-green in recesses, with white inner margin rectangular-truncate and recurved over basal part of gland, outer margins divided into 3–5 spreading finger-like irregularly reticulated-corrugated white processes; stamens with sparsely pubescent pedicels pale green below and suffused with red towards tips (with red anthers and red filaments), bracteoles very few, broadly linear and pubescent along apical margins; ovary ± spherical and obscurely 3-angled, red, glabrous to pubescent with white hairs near base, initially erect and almost exserted from cyathium (on pedicel ±  2  mm long) then becoming pushed out of cyathium on glabrous spreading to later decurved pale green pedicel 4–6  mm long; styles 4–6  mm

80

long, branched and slightly spreading in upper third, red near base and green suffused with red above. Capsule 6–7  mm diam., obtusely 3-angled, glabrous, shiny green with bright red down sutures and edges, exserted on decurved and later erect pedicel 4–6 mm long.

2  Euphorbia subg. Athymalus

Distribution & Habitat Euphorbia polycephala occurs in the mountainous and hilly country between Hofmeyr, Cradock and Pearston in the Eastern Cape.

Fig. 2.33.  Distribution of Euphorbia polycephala (© PVB).

Plants are usually found in stony ground on gentle slopes, often near the bases of larger hills and much more rarely in flat areas. While young specimens begin life under a bush or

Fig. 2.34.  Euphorbia polycephala, mat-like plants (the largest slightly over 1 m in diam.) among rocks and shrubs of Rhigozum obovatum, PVB 11083, west of Cradock, South Africa, 26 Oct. 2012 (© PVB).

in the protection of a few stones, with age they outgrow this protection and most larger plants are found fully in the open or filling up patches of ground between large stones.

2.1 Sect. Anthacanthae

Euphorbia polycephala grows in areas where Angora goats are kept for the production of mohair and it is possible that grazing by these animals has caused it to become very rare in some areas. However, even plants protected from being grazed by goats by fencing have the tips of many of the branches chewed off by small buck or rodents. If heavily grazed, the branches can begin to die, leaving an exposed patch which may burn in the sun and can lead to the death of the entire clump. Diagnostic Features & Relationships Euphorbia polycephala is a remarkable species, which may form impressive mounds up to half a metre tall and 1 m in diameter. Growth appears to be slow and the larger specimens must achieve an extraordinary age. Such large mounds

81

consist of hundreds to thousands of short, small branchlets that are exceptionally tightly packed (so tightly that they cannot be pried out without damaging them and many of the surrounding branches) into a nearly continuous surface, underneath which there is a further tight and complex network of older branches that have arisen from the stem. In very stony, sloping ground the mounds are convex to almost hemispherical and may be up to 0.5 m tall but in such situations completely flat, mat-like plants also develop. In flat spots they can be entirely sunken into the ground, forming a carpet or mat that is level with the surface of the soil. Such mats may also be up to 1  m in diameter. The plants have almost no tendency to be rhizomatous and branches do not spread underground for any significant distance from the main plant-body.

Fig. 2.35.  Euphorbia polycephala, large mound ± 1 m diam., among rocks (with another behind to the right), PVB 11083, west of Cradock, South Africa, 2 May 2008 (© PVB).

Fig. 2.36.  Euphorbia polycephala, flat mat ± 1 m diam., among rocks, PVB 11083, west of Cradock, South Africa, 2 May 2008 (© PVB).

The terminal branchlets are mostly short and conical (often scarcely longer than thick), tapering rapidly from their base to a rounded tip. However, if sheltered, they can etiolate to 150  mm long. Mostly they are a very uniform, bright green colour which does not tend to fade with age. Most of them are around 12  mm thick and each has few, comparatively stout, shortly conical tubercles. When they are young, each tubercle bears a prominent, often almost circular leaf at its apex, which persists for several weeks. On somewhat more etiolated branches the tubercles become obsolete and the branches may be nearly round in cross-section. Flowering is sporadic throughout the summer, though the major flowering time appears to be between October and November. Even in ‘good’ years, the plant does not become

covered with cyathia, though up to 50 may be present in various stages of development on a plant. Each flowering branch elongates into a short peduncle bearing a single, terminal cyathium. Up to three more cyathia may develop from the axils of bracts around this terminal cyathium. In subsequent seasons other, short peduncles may arise from the axils of some of the tubercles just below the apex of the branch, bearing further cyathia. The cyathium is shortly conical, with four or five conspicuous and spreading glands, which have at least three long, spreading processes on their outer margins. The cyathium expands greatly in breadth during its lifetime. When the female floret initially matures in an erect position, the cyathium is often only 8  mm across but it may reach 15 mm or more by the time the males are mature. From the time of maturity of the female the cyathium gives off an

82

2  Euphorbia subg. Athymalus

Fig. 2.37.  Euphorbia polycephala, cyathia in various stages of development, among rocks, PVB 11083, west of Cradock, South Africa, 26 Oct. 2012 (© PVB).

Fig. 2.38.  Euphorbia polycephala, early female stage, flowering in habitat, PVB 11647, SE of Cradock, South Africa, 30 Nov. 2016 (© PVB).

unusual odour somewhat like scented soap or in some cases like a water-based paint. The inner margin of the gland is folded outwards to cover most of the base of the gland. Almost the entire upper surface of each gland is heavily corrugated with a thick white crust which extends onto the long marginal processes. This gives the glands a white colour so that, together, they form a white ring around the red lobes and reddish florets in the centre of the cyathium. The few male florets are dark red on red-green pedicels and, similarly,

the female floret is mainly red. Initially the female floret is erect on a pedicel about 2 mm long, with the upper part of the ovary hidden by or partly protruding from the lobes. As the males mature, the pedicel of the female at least doubles in length and inclines to the side to push the ovary and styles beyond the cyathium, becoming erect again just before dehiscing. The mature capsule is an attractive shiny green with red lines down the sutures and along the edges.

Fig. 2.39.  Euphorbia polycephala, male stage, flowering in cultivation, PVB 11647, SE of Cradock, South Africa, 30 Feb. 2015 (© PVB).

Fig. 2.40.  Euphorbia polycephala, capsules in different stages of development, PVB 11054, west of Witmos, South Africa, 26 Oct. 2012 (© PVB).

Euphorbia polycephala is closely related to E. globosa and E. patula. Euphorbia patula is common along the Fish River Valley near where E. polycephala occurs, but the two have not been found growing together, with E. patula seem-

ing to prefer the lower, hotter, flat and stony areas near to the river and E. polycephala more prevalent on the hillsides further from the foot of the valley. They are most easily separated vegetatively. The main differences are the lack of

2.1 Sect. Anthacanthae

rhizomes and the very dense, mound- to mat-forming growth, the stouter, brighter green branches with fewer, broader tubercles that do not taper into the leaves and the considerably larger, broader and more rounded leaves in E. polyceph-

83

ala. Florally they are similar, though the glands and their processes are shorter and more brightly white in E. polycephala while those in nearby forms of E. patula are mostly greener.

Fig. 2.41.  Euphorbia polycephala. A, tip of branch with leaf-rudiments (scale 3 mm, as for B, C). B, cyathium from side. C, cyathium from above. D, cyathial gland from above (scale 2 mm, as for E). E, anthers and female floret in dissected cyathium. F, anthers, bracteole and female floret at early stage in dissected cyathium (scale 1 mm, as for G). G, female floret after males mature. Drawn from: A, PVB 11054, west of Witmos, South Africa. B–G, PVB 11083, west of Cradock, South Africa (© PVB).

History One of the last two species described by Marloth (the other being E. wilmaniae), E. polycephala was first recorded in June 1913 by Schönland (under his number 2000 at GRA and K). The type was collected near Mortimer, south of Cradock by H.W. Shoesmith in August 1913. Mr Shoesmith farmed in the area and used to chop out the plants to mix with Opuntia as fodder to sustain his stock during dry periods (Frick 1931). The small specimen made by Schönland in the Albany Museum and at Kew (Schönland 2000) is also annotated ‘good food for ostriches in times of drought’. Euphorbia polycephala appears to have been unknown to White et al. (1941) and Dyer made no collections of it. The species was first noted around Pearston by Sita Palmer in September 1961 (Palmer 1966) and the decline of this population was observed during the 1980’s by Court (1988). However,

other populations in the Cradock area seem to be fairly stable, especially where the plants have been fenced off to protect them from grazing by goats and sheep. Euphorbia pseudotuberosa Pax, Bull. Herb. Boiss., sér. 2, 8: 637 (1908). Type: South Africa, Transvaal, Pretoria, 1892, Fehr 43 (Z, holo.). Bisexual spineless glabrous dwarf geophyte 20–100 mm tall above ground, arising from one or more perennial often slightly woody stems below ground of variable length, stems arising from narrowed apex of turnip-shaped to very irregular tuber 50–200 × 20–100 mm and covered with brownish bark, tapering beneath into fibrous roots. Branches herbaceous and deciduous, erect to spreading, often rebranching extensively, 20–100 × 2–5 mm, without tubercles, glabrous to puberulous, pale green often suffused with red; leaves

84

present mainly towards tips of branches, alternating lower on branches and opposite below forks in branches and below flowering nodes, 15–60 × 1–6 mm (lower ones much shorter, sometimes scale-like and rapidly caducous), ascending to spreading, linear, glabrous, with entire sometimes finely ciliate and often reddish margins, usually longitudinally folded upwards, acute, sessile, sometimes with small rod-shaped stipules 0.5–1 mm long at bases. Synflorescences each of 1 terminal bisexual cyathium on peduncle 1–2 × 1–1.5 mm, often with further bisexual cyathia developing on branches 5–15 mm long from axils of bracts below terminal cyathium, with 2–3 linear glabrous bracts (barely distinguishable from leaves) 10–50 × 2–8 mm subtending cyathium; cyathia conical to cupular, glabrous to minutely puberulous, 6–10  mm broad (± 2 mm long below insertion of glands), with 5 reddish lobes with finely incised margins, grey-green often suffused with red; glands 4–5, ±  cuneate, not contiguous, 2.7–4 mm long, ascending, above reticulated with white to cream ridges (especially towards edges) on green background; inner margins white or cream, rectangular-truncate and incurved over gland; outer margins divided into 2–6

Fig. 2.42.  Distribution of Euphorbia pseudotuberosa (© PVB).

2  Euphorbia subg. Athymalus

spreading finger-like processes 1–3 mm long with irregularly white-margined pits above in lower two thirds (pits disappearing towards tips); stamens few with pubescent pedicels distinctly reddish towards their tips, with red filaments and reddish yellow anthers, bracteoles very few, filiform and pubescent only near apices; ovary ellipsoidal, glabrous, red, exserted on sparsely pubescent reddish spreading to decurved pedicel 5–7  mm long; styles 3–6  mm long, branched into spreading arms above middle. Capsule 6–8  mm diam., obtusely 3-angled, glabrous, exserted on gradually erect pedicel 4–8 mm long. Distribution & Habitat Euphorbia pseudotuberosa is known only in South Africa, where it is mainly found in the southern parts of the former Transvaal from Pretoria southwards to the Free State and westwards into the former Cape Province around Vryburg. Euphorbia pseudotuberosa usually grows in flat, stony ground with scattered trees, among tufts of grass and other small geophytes, often in places subject to fires during the dry season.

2.1 Sect. Anthacanthae

Fig. 2.43.  Euphorbia pseudotuberosa, excavated plants with elongated, irregular tubers, PVB 12437, south of Vryburg, South Africa, 7 Jan. 2013 (© PVB).

Fig. 2.44.  Euphorbia pseudotuberosa, flowering before the leaves are fully developed in burnt-off grass, Faerie Glen, Pretoria, South Africa, 8 Sep. 1979 (© PVB).

85

Diagnostic Features & Relationships Euphorbia pseudotuberosa is a geophyte, arising from a sometimes quite substantial subterranean tuber. The tuber is very variably shaped, often dividing and branched, with each portion tapering to the base into fine roots and the whole giving rise at the top to one or more usually erect, perennial, slightly woody branches that rise to just below the surface of the soil. From axillary buds along these perennial parts, the annual shoots rise out of the ground each season and they consist of several non-succulent, slender, cylindrical branches. Each branch bears several leaves, which range from small, scale-like structures around the base of the branch, to others up to 60 mm long towards the tip of the branch. Those towards the base of the branch are alternate while those just beneath the cyathia and at the forks in the branches are in opposite pairs. The leaves are generally remarkably grass-like, with slender bases and slender, finely acute tips. They are greyish green, sometimes suffused with red. Tiny, rod-like stipules may be present at their bases. Euphorbia pseudotuberosa flowers mainly between September and November. This is usually after the first rains have fallen and before the surrounding grasses have recovered from being burnt off. Then the cyathia are produced on very short branches close to the ground and are quite clearly visible. In such plants flowering is profuse and each branch may be terminated by a cyathium on a very short peduncle, with further cyathia developing at the tips of slightly longer peduncles that arise in the axils of the bracts underneath the terminal cyathia. The bracts are indistinguishable from the leaves and are always in opposite pairs. In E. pseudotuberosa the cyathia have a relatively small involucre sur-

Fig. 2.45.  Euphorbia pseudotuberosa, among burnt-off grass, cyathia with 4 or 5 glands, PVB 12212, east of Pretoria, South Africa, 20 Sep. 2012 (© PVB).

86

2  Euphorbia subg. Athymalus

rounded by four or five ornate glands. Each cyathium’s ovary is soon pushed out on a long, curved pedicel, after which the few male florets develop together with a few slender bracteoles. The glands are particularly ornate, with a portion of the inner margin folded outwards. Their outer margin is deeply divided into two or three widely spreading

processes whose upper surface is white and deeply pittedrugulose with green or red in the pits. The base of the gland is similarly coloured, with white ridges on a darker background and the same is true of the folded-back portion of the inner margin.

Fig. 2.46.  Euphorbia pseudotuberosa, flowering with fully developed leaves, PVB 12437, south of Vryburg, South Africa, 7 Jan. 2013 (© PVB).

Fig. 2.47.  Euphorbia pseudotuberosa, with cyathia and capsules, PVB 12212, east of Pretoria, South Africa, 20 Sep. 2012 (© PVB).

Molecular results place E. pseudotuberosa with the highly succulent E. globosa and E. patula. While it does not resemble these succulents vegetatively, florally they are similar. This can be seen in the relatively large cyathia with similar glands, usually heavily pitted on their upper surfaces, few male florets and bracteoles and a slender pedicel beneath the ovary, which pushes the ovary well out of the cyathium before the male florets mature. Euphorbia pseudotuberosa and E. trichadenia were said to differ by the lack of corrugation on the upper surface of the glands in E. trichadenia (Brown 1915; White et al. 1941). Brown deduced these differences from comparison with Pax’s type specimens. However, Pax (1908) distinguished E. pseudotuberosa from E. benguelensis (considered to be a synonym of E. trichadenia) by the simple appendages of the glands and the smaller cyathia. Nowhere in his descriptions of either E. trichadenia or E. pseudotuberosa did he mention any corrugations on the upper surfaces of the glands and it could be that ‘simple’ has been interpreted as meaning ‘without corrugations’.

Here, we take plants where the cyathial glands have two to six processes on their outer edges and are corrugated above as E. pseudotuberosa. The plants whose glands are smooth above and have many fine processes on their edges are treated as E. trichadenia. Vegetatively they differ mainly in that in E. trichadenia the tuber is often covered with a brown, somewhat rough bark which continues onto the slender neck that reaches to ground-level (the bark of the tuber is pale and smooth in E. pseudotuberosa, where the tubers are also more slender). Euphorbia trichadenia has fewer, longer branches with longer internodes and fewer, longer leaves, while E. pseudotuberosa is more densely branched close to the ground. In E. pseudotuberosa the cyathium is considerably smaller and much more slender towards its base, the female florets are held on a much longer pedicel and are also far more slender. The respective cyathial glands are clearly different in shape, size and ornamentation. The capsule is also much further exserted in E. pseudotuberosa, while it is often nearly sessile in E. trichadenia.

2.1 Sect. Anthacanthae

87

Fig. 2.48.  Euphorbia pseudotuberosa. A, side view of male cyathium (scale 3 mm, as for B–D). B, C, cyathium from above. D, cyathial gland. E, anthers and bracteoles (scale 1 mm, as for F). F, female floret. Drawn from: PVB 12212, east of Pretoria, South Africa (© PVB).

History Euphorbia pseudotuberosa was described from a specimen collected by Johann Wilhelm Fehr (1856–1914) in 1892 near Pretoria. Before this, Carl Zeyher had collected it (probably in November 1841) during his joint expedition with Joseph Burke to the Transvaal in 1841–2. Euphorbia trichadenia Pax, Bot. Jahrb. Syst. 19: 125 (1894). Type: Angola, Lunda, between Kimbundo and the Quango, Sept. 1876, Pogge 116 (missing). Neotype (Bruyns 2012): Angola, Huilla, near Lopollo towards Nene, Oct.-Nov. 1859, Welwitsch 282 (BM; G, K, iso.). Euphorbia benguelensis Pax, Bull. Herb. Boiss. 6: 741 (1898). Lectotype (designated here): Angola, Huilla, source of Luala, Antunes 362 (COI). Euphorbia subfalcata Hiern, Cat. Afr. Plants Welw. 1 (4): 948 (1900). Type: Angola, Huilla, near Lopollo towards Nene, Oct.-Nov. 1859, Welwitsch 282 (BM, holo.; G, K, iso.).

Euphorbia gossweileri Pax, Bot. Jahrb. Syst. 43: 88 (1909). Lectotype (Bruyns 2012): Angola, Malanje distr., left bank of River Quanza, Gossweiler 994 (K). Euphorbia trichadenia var. gibbsiae N.E.Br., Fl. Trop. Afr. 6 (1): 524 (1911). Lectotype (designated here): Zimbabwe, Matopos, near Isotye, 5000’, Feb. 1905, Gibbs 234 (BM; K, iso.). Bisexual spineless glabrous dwarf geophyte 20–100 mm tall above ground, arising from one or more perennial often slightly woody stems below ground of variable length, stems arising from narrowed apex of turnip-shaped to very irregular tuber 50–200 × 20–100  mm and covered with brownish bark, tapering beneath into fibrous roots. Branches herbaceous and deciduous, erect to spreading, sometimes rebranching extensively, 20–100 × 2–5 mm, without tubercles, glabrous to puberulous, pale green often suffused or speckled with red; leaves present mainly towards tips of branches, alternating lower on branches and

88

opposite below forks in branches and below flowering nodes, 15–60 × 1–6 mm (lower ones much shorter, sometimes scale-like and rapidly caducous), ascending to spreading, linear, glabrous, with entire sometimes finely ciliate and often reddish margins, usually longitudinally folded upwards, acute, sessile, estipulate. Synflorescences each of 1 terminal bisexual cyathium on peduncle 1–2 × 1–2 mm, often with further bisexual cyathia developing on branches 5–15 mm long from axils of bracts below terminal cyathium, with 2–3 linear glabrous bracts (barely distinguishable from leaves) 10–50 × 2–8  mm subtending cyathium; cyathia conical-cupular, glabrous to minutely pubescent, 8–10 mm broad (2–3 mm long below insertion of glands), with 5 reddish often pubescent lobes with finely incised margins, green often suffused with red; glands 5, ±  cuneate to fan-shaped, contiguous, 2.7–5  mm long, ascending, above pinkish cream in upper half and green in

Fig. 2.49.  Distribution of Euphorbia trichadenia (© PVB).

2  Euphorbia subg. Athymalus

basal undivided part; inner margins green above and slightly raised and incurved over gland; outer margins divided into (3) 6–12 ascending and slightly spreading finger-like pinkish cream smooth processes 1.5–3  mm long; stamens few with pubescent pedicels distinctly reddish towards their tips, red filaments and yellow anthers with red pore, bracteoles very few, filiform and pubescent only near apices, red; ovary ellipsoidal, pubescent, green, partly exserted on sparsely pubescent reddish erect pedicel 1–2 (4) mm long; styles 4–7.5 mm long, branched into spreading arms in upper third. Capsule 9–11 mm diam., obtusely 3-angled, usually glabrous, sessile to slightly exserted on erect pedicel 1 (–4) mm long. Distribution & Habitat Euphorbia trichadenia is known to be widely distributed across south-tropical Africa from Angola to Botswana and

2.1 Sect. Anthacanthae

89

Zimbabwe into South Africa. In South Africa it occurs mainly in the northern parts of the former Transvaal and eastwards to the Nelspruit district, as well as in parts of KwazuluNatal from the highlands to near the coast. Specimens of E. trichadenia grow in shallow soils overlaying rocks or in stony ground, usually in open areas between trees. It often occurs in relatively mesic areas where there are few other succulents. Diagnostic Features & Relationships A geophyte, Euphorbia trichadenia possesses a sometimes quite substantial subterranean tuber (Fig. 2.50). Its shape depends on pressures from rocks or other roots but it can be neatly turnip-like if growing unrestricted. At the base the tuber gives rise to fine roots and at the top there are perennial, slightly woody branches that rise to the surface of the soil. It is from these underground perennial branches that the annual, herbaceous branches emerge above the soil. The greyish green, often reddish branches are slender and cylindrical. Leaves vary from tiny, alternating sometimes scale-like structures around the base of the branch, to 60 mm long and opposite towards its tip. They are generally slender and grass-like and are usually a greyish green colour, sometimes suffused with red. No stipules have been observed at the bases of the leaves (unlike the case of E. pseudotuberosa). Euphorbia trichadenia flowers mainly between September and December at the tips of the first growth that emerges after the winter. Further cyathia may develop from the axils of the bracts beneath the terminal cyathium. As in

Fig. 2.51.  Euphorbia trichadenia, pale firm sand under trees, PVB 13359, near Menongue, Angola, 4 Jan. 2017 (© PVB).

Fig. 2.50.  Euphorbia trichadenia, plants with substantial tubers, that on the left ±  8  cm diam., PVB 12095, Lapalala, Waterberg, South Africa, 30 Dec. 2011 (© PVB).

90

2  Euphorbia subg. Athymalus

Fig. 2.52.  Euphorbia trichadenia, flowering before leaves fully developed, Waterberg near Vaalwater, South Africa, 5 Oct. 2008 (© R.G. Peckover).

Fig. 2.53.  Euphorbia trichadenia, with even more ornate appendages on cyathial glands, Waterberg near Vaalwater, South Africa, 5 Oct. 2008 (© R.G. Peckover).

E. pseudotuberosa, the bracts are almost indistinguishable from the leaves. In E. trichadenia the cyathia are comparatively large, with a stout involucre surrounded by four or five large, slightly stalked glands. Inside the cyathium, there are few, very slender bracteoles, relatively few male florets and a prominent female floret that is raised slightly on a stout pedicel. All the florets are strongly suffused with red, though in the anthers the red is mainly confined to the area

around their pores. The glands have a bright green basal area over which the inner margin is partly folded back. Their outer margin is finely divided into usually large numbers of irregularly finger-like processes which are pinkish cream on their upper surface and more or less smooth. Some of these details were shown for the first time by Pole Evans (1928b).

Fig. 2.54.  Euphorbia trichadenia, developing cyathia, PVB 12095, Lapalala, Waterberg, South Africa, 16 Oct. 2012 (© PVB).

2.1 Sect. Anthacanthae

Euphorbia trichadenia is closely related to E. pseudotuberosa and the differences between them are discussed under that species. History Euphorbia trichadenia was described from specimens collected by Paul Pogge in Angola in September 1876. Before

91

this, both Joseph Burke and Carl Zeyher had collected it (probably late in 1841, though the specimens at Kew are undated), during their joint expedition to the Transvaal in 1841–2. Welwitsch also gathered it in Angola in 1859. The specimen Zeyher 1539 was split by N.E. Brown into a part that he called E. trichadenia (identified on the basis of comparison with Pax’s type, which he saw in January 1911)

Fig. 2.55.  Euphorbia trichadenia. A, side view of cyathium (scale 3 mm). B, dissected cyathium with anthers, bracteoles and female floret (scale 2 mm). C, female floret (scale 1 mm). Drawn from: PVB 12095, Lapalala, Waterberg, South Africa (© PVB).

and a part that he named E. pseudotuberosa, which gave rise for the first time to the notion that there were two separate but similar-looking species sometimes growing close to one another in the Transvaal of South Africa.

2.1.2 Subsect. Florispinae Euphorbia subsect. Florispinae Haw., Philos. Mag. Ann. Chem. 1: 275 (1827). Type: Euphorbia stellispina Haw.

Unisexual or bisexual succulent shrubs sometimes consisting of a cylindrical tuberculate stem with a tuft of leaves at its apex, dwarf succulents or geophytes, usually with green photosynthetic bark, stem and branches mostly tuberculate, often armed with spines developing from persistent hardened sterile peduncles arising in axils of tubercles. Leaves spiral, mostly narrow and linear, occasionally ovate, entire, mostly sessile, deciduous and often reduced to minute ephemeral rudiments, arising at apex of conical tubercles (these mostly fused vertically into 4–20 straight or spiral

92

angles along stem and branches); stipules absent. Synflorescences usually solitary in axils of tubercles often reduced to a single cyathium terminating the short (rarely to 10  cm long) sometimes tuberculate peduncle (peduncle sometimes branching among uppermost bracts to bear further cyathia). Cyathia all similar and bisexual or unisexual with males usually larger than females, 2.5–10  mm diam.; glands 4–5, outer margins entire to finely crenulate. Capsule

2  Euphorbia subg. Athymalus

obtusely 3-angled, smooth, glabrous (pubescent), 5–8  mm diam., ±  sessile. Seeds faintly ridged to round, usually smooth (slightly pitted). This subsection includes 26 species, all of which are endemic to southern Africa. They are found in Lesotho, South Africa and Swaziland and are divided into four morphologically distinctive series.

Key to the series of subsect. Florispinae 1. Geophytes, with only deciduous leaves and inflorescences appearing above ground....................2.1.2.3. ser. Rhizanthium 1. Dwarf succulents or shrubs, with perennial stem and branches at least reaching the surface of the ground........................2. 2. Cyathia bisexual, only fertile short-shoots present, with sterile short-shoots absent................................2.1.2.4. ser. Treisia 2. Cyathia mainly unisexual (occasionally bisexual in E. polygona), both sterile and fertile short-shoots mostly present..........3. 3. Leaves conspicuous in growing season (then deciduous), tubercles arranged into obscure rows but not fused into angles along stems and branches, fertile short-shoots with 2–3 ± prominent bracts immediately beneath and exceeding ­cyathium…...................................................................................................................2.1.2.1. ser. Hystrix 3. Leaves much reduced to fleeting rudiments (though to 50 mm long in E. pulvinata), tubercles fused into clear angles along stem and branches, fertile short-shoots mostly with minute bracts (rarely with 2–3 slightly larger bracts immediately beneath cyathium and these rarely exceeding cyathium).......................................2.1.2.2. ser. Meleuphorbia

2.1.2.1 Ser. Hystrix Euphorbia ser. Hystrix Bruyns, Taxon 62: 1193 (2013). Type: Euphorbia hystrix Jacq. (= Euphorbia loricata Lam.). Unisexual shrubs (sometimes bisexual in E. bupleurifolia) with tuberculate stem and branches, tubercles arranged into obscure rows (but not fused into angles), usually with persistent, sterile short-shoots without prominent bracts near tips (spiny in E. loricata; firm but not spiny in E. multifolia;

weak and slender in E. oxystegia; absent in E. bupleurifolia). Leaves soft, conspicuous, deciduous, lasting for much of the growing season. Synflorescences solitary on peduncle with 2–3 fairly prominent bracts immediately subtending and usually considerably exceeding cyathium. Cyathia unisexual (bisexual sometimes in E. bupleurifolia). Four species belong to this series and are found in South Africa only. Of these, only E. bupleurifolia occurs outside the Greater Cape Floristic Region and is found in the Eastern Cape and KwaZulu-Natal.

Key to the species of ser. Hystrix 1. Plant a dwarf succulent often mostly submerged in ground with stem 25–70 mm thick (and sometimes with few similarly thick branches) with leaves in rosettes at tip, sterile short-shoots absent, rootstock without tubers................E. bupleurifolia 1. Plant a succulent shrub with many branches 6–30 mm thick, sterile short-shoots persistent as weak to stiff spikes, rootstock with several swollen tubers from which fibrous roots arise.................................................................................2. 2. Plant with many branches usually forming dense shrub, sterile short-shoots becoming hard and often sharp-tipped and persisting for several seasons as a protective armour around apices of branches................................................................3.

2.1 Sect. Anthacanthae

93

2. Plant with few branches forming diffuse shrub, sterile short-shoots remaining weak and pliable and persisting for one or two seasons around apices of branches..............................................................................................................E. oxystegia 3. Branches not thicker than 20 mm, sterile short-shoots usually sharp-tipped and spine-like (often white), leaves linear to oblanceolate and obtuse, not abruptly truncate....................................................................................................E. loricata 3. Branches often thicker than 20 mm, sterile short-shoots not sharp-tipped (yellowish brown), leaves linear and abruptly truncate..............................................................................................................................................................E. multifolia

Fig. 2.56.  Distribution of Euphorbia ser. Hystrix (showing number of species per half-degree square) (© PVB).

Euphorbia bupleurifolia Jacq., Pl. Hort. Schönbr. 1: 55, t. 106 (1797). Tithymalus bupleurifolius (Jacq.) Haw., Syn. Pl. Succ.: 138 (1812). Lectotype (Bruyns 2012): Jacq., Pl. Hort. Schönbr. 1: t. 106 (1797). Euphorbia proteifolia Boiss. in DC., Prodr. 15 (2): 92 (1862). Lectotype (designated here): South Africa, near Umtata, Drège 8196 (S; MEL, iso.). Bisexual spineless usually single-stemmed (rarely with few similar branches) glabrous dwarf succulent 40–200 mm tall, from cluster of fibrous roots. Stem erect, globose to cylindrical, 25–70 mm thick, covered with tubercles, brown; tubercles crowded and conical, 2–5 mm long with hexagonal base, spreading horizontally and often decurved near tips, brown, mostly with prominent paler scar at tip where leaf fell off, not joined into angles but arranged into 10–17 spiralling

rows, somewhat intermeshed towards bases and without clear grooves between rows; leaves in rosette or tuft around apex of stem and branches, each on tip of new tubercle, 50–230 × (5) 10–20 mm, ascending to spreading, slightly bluish grey-green, herbaceous, glabrous, present during growing season and shed during dry periods, linear-oblanceolate to slightly spathulate, longitudinally folded slightly upwards, with straight margins faintly reddish towards tips, acute to obtuse, tapering gradually into sparsely pubescent petiole 10–20  mm long and usually distinctly flattened or slightly channelled above. Synflorescences many per stem towards apex, each of one bisexual or unisexual cyathium in axil of tubercle, borne on slightly angled finely pubescent peduncle 5–50 × 2–2.5 mm, bearing 2 yellow-green bracts at apex forming cup around cyathium, bracts 7–12 × 9–22 mm, nearly circular and canaliculate along midrib, obtuse, apicu-

94

late, glabrous or slightly pubescent near base, closely surrounding cyathium, yellow-green to green becoming reddish; cyathia shallowly bowl-shaped to cupular, minutely puberulous outside, 6–8 mm broad and slightly shorter than bracts, with 5 lobes with deeply incised margins; glands 5, ascending, fan-shaped, slightly separated to contiguous, 1.5 × 2.5–3 mm, yellow-green, inner margins recurved towards rest of gland, outer margins entire to finely crenulate and slightly spreading; stamens with sparsely pubescent or glabrous pedicels, bracteoles very few and filiform or absent; ovary globose, glabrous or finely pubescent, raised on pedicel < 1 mm long; styles 3–5.5  mm long, branched only in upper third. Capsule 8–9  mm diam., obtusely 3-angled, glabrous, ± sessile.

2  Euphorbia subg. Athymalus

Distribution & Habitat Euphorbia bupleurifolia is rather poorly known in KwazuluNatal between Pietermaritzburg, Durban and Port Shepstone (where many of its former habitats appear to have been destroyed by agriculture and by urban expansion) and it is much more widely documented in the Eastern Cape, especially from the Kei River westwards to around Grahamstown, where it has been recorded at altitudes from near sea level to over 1100 m. Euphorbia bupleurifolia often grows in grasslands, though it appears to favour slightly rocky areas, where rounded outcrops of dolerite or sandstone protrude from among the tufts of grass and where the soil is slightly shallower. Occasionally it will be found between rocks among trees and shrubs.

Fig. 2.57.  Distribution of Euphorbia bupleurifolia (© PVB).

Diagnostic Features & Relationships Vegetatively, E. bupleurifolia is one of the more unusual members of the genus. The plant usually consists of a solitary, firm, stout, brown stem (only rarely is this branched), which may reach a length of 150 mm or more, but is usually much shorter. Below, the stem descends into the ground for

some distance (usually around 50 mm), after which it tapers off into slender roots. Often it is entirely hidden among grasses and other surrounding vegetation and leaf-litter, with only its apex protruding from the ground. In this case, only the leaves are visible and such specimens are essentially geophytes. The surface of the stem is covered com-

2.1 Sect. Anthacanthae

95

Fig. 2.58.  Euphorbia bupleurifolia, geophytic plant on edge of dense bush, PVB 10992, Bolo River, near Stutterheim, South Africa, 18 Jan. 2008 (© PVB).

Fig. 2.59.  Euphorbia bupleurifolia, flowering before leaves fully developed, PVB 10992, Bolo River, near Stutterheim, South Africa, 27 Sep. 2012 (© PVB).

pletely with brown tubercles. These are arranged into spiralling rows up the stem but their bases overlap considerably so that there is no flattened area or groove between the rows. Both the length and the breadth of the tubercle appears to increase substantially after the leaf has fallen off. With the leaf present, the tubercle is erect, has the same green colour as the petiole and is usually about 2  mm long and 3  mm broad. Once the leaf is gone, the tubercle soon becomes brown, its tip gradually descends and the base increases in breadth to 10 mm or more. Most tubercles have a prominent scar at their tip where the leaf was joined to it. However, some do not have this scar and these appear to be the tubercles that are produced at the beginning and the end of the growing season, when leaves do not develop properly. These tubercles are also the longest. Leaves are present during most of the summer months, each leaf usually lasting for most of the growing season. They are slender, tapering gradually towards the base into the petiole and are usually at least 10 cm long, though they may be up to 20 cm long on occasion. When very young they are green, sometimes suffused with red and they soon change to the distinctive bluish green colour, with a pale midrib and faint, pale side-veins, that is typical of this species.

Flowering usually takes place early in the growing season, between September and December. A rosette of cyathia develops at the apex of the stem and is soon pushed to the side by the rapidly developing tuft of leaves at the apex so that the cyathia usually mature a bit below this tuft of leaves. Each cyathium is solitary on a short, erect peduncle and is subtended by two prominent, yellow-green bracts which are up to twice as long as and much broader than the cyathium and hide it slightly. In material grown from along the Kei River the plants were unisexual. However, they may be bisexual and the cyathia seem to contain either none or very few small bracteoles. It has been established that E. bupleurifolia is related to the species in subsect. Florispinae of sect. Anthacanthae (Peirson et al. 2013), but its relationships within this subsection are unclear. Vegetatively, the tuberculate stem with prominent leaves suggests that it is a close ally of such species as E. clandestina and E. clava but the unisexual cyathia do not corroborate this. Its closest allies within this subsection are not known and it has been included provisionally in E. ser. Hystrix along with E. loricata. From the allied species it differs by the dense covering of narrow, brown tubercles on the stems and also by the very much larger leaves.

96

Fig. 2.60.  Euphorbia bupleurifolia, fruiting plant with stem ±  5  cm tall, Hogsback, Amatola Mountains, South Africa, 30  Dec. 1992 (© PVB).

2  Euphorbia subg. Athymalus

Fig. 2.61.  Euphorbia bupleurifolia, Hogsback, Amatola Mountains, South Africa, 30 Dec. 1992 (© PVB).

Fig. 2.62.  Euphorbia bupleurifolia. A, cyathium from side with bracts (scale 2 mm, as for B–D). B–D, cyathium from side with bracts removed. E, anthers and bracteole (scale 1 mm, as for F). F, female floret. Drawn from: A, B, PVB 4377, Queenstown, South Africa. C–F, PVB 10992, Bolo River, near Stutterheim, South Africa (© PVB).

2.1 Sect. Anthacanthae

History Euphorbia bupleurifolia was ostensibly brought back to Vienna by Franz Boos and Georg Scholl, the two gardeners from Schönbrunn, Vienna who gathered plants and other natural history material together at the Cape between May 1786 and February 1788. A painting of it was made by Robert J. Gordon, possibly from a plant in cultivation in his garden in Cape Town and it was also introduced to Kew in 1791 by Francis Masson. J.F.  Drège and his brother Carl observed it in at least two places during their travels along the eastern side of South Africa during 1832, one near the Katberg and the other near Umtata. These collections were placed by Boissier under the two names ‘bupleurifolia’ and ‘proteifolia’, which are today regarded as belonging to one species. Euphorbia loricata Lam., Encycl. 2 (2): 416 (1788). Lectotype (Bruyns 2012): Pluk., Phytographia 3: t. 230. f. 5 (1692). Euphorbia hystrix Jacq., Pl. Hort. Schönbr. 2: 43, t. 207 (1797). Treisia hystrix (Jacq.) Haw., Syn. Pl. Succ.: 131 (1812). Lectotype (Bruyns 2012): Jacq., Pl. Hort. Schönbr. 2: t. 207 (1797). Euphorbia armata Thunb., Prodr. Fl. Cap. 2: 86 (1800). Type: South Africa, Cape, Thunberg (UPS-THUNB, holo.). Euphorbia eustacei N.E.Br., Bull. Misc. Inform. 1913: 122 (1913). Type: South Africa, Cape, near Matjiesfontein, Oct. 1912, C.E. Pillans (K holo.; K, PRE, iso.). Unisexual spiny many-branched glabrous shrub to dense mound-forming succulent 0.1–1.5 × 0.15–1 m, branching from base and above on similar stem, stem tapering beneath ground then giving rise to several swollen fusiform roots from which fibrous roots arise. Branches spreading to erect, cylindrical, 8–20  mm thick (stem often much thicker near base), simple or rebranching above, covered densely with tubercles, smooth, pale green becoming brown then covered with a thin grey bark; tubercles conical, 2–4 mm long and vertically arranged (but not fused) into 6–12 obscure spiralling rows, with fine grooves around and separating them in early stages only, with solitary simple spines in their axils; spines developing from persistent sterile short-shoots, straight and ascending, rigid, 12–50 × 1–1.5 mm, with sev-

97

eral small alternating glabrous obtuse caducous bracts to 2–3 × 1–1.5 mm, initially pink to glaucous green, later drying brown and then grey or drying white and then later turning grey, glabrous; leaves on tips of new tubercles towards apices of branches and stem, 15–75 × 3–10 mm, spreading, gradually deciduous, linear to oblanceolate and lightly channelled above for most of length, obtuse to mucronate, tapering to narrow base, sessile, pale glaucous green, glabrous or minutely puberulous near apex and along margins. Synflorescences many per branch towards apex, each solitary in axil of tubercle on unbranched green to pink or red shortshoot 15–50 mm long (usually longer in males than females and stouter than sterile short-shoots) with 1 terminal unisexual cyathium, glabrous, with 2–4 broadly ovate glabrous green to red bracts 3–8 × 3–5 mm with fine marginal cilia just beneath and slightly overtopping cyathium, others below smaller (those around female cyathium often shorter and broader than around male); cyathia cupular, glabrous, 4.5– 6.5 mm broad (considerably broader in male than female), with 5 slightly pubescent often pinkish lobes with toothed margins, green; glands 5, elliptic, 2.5–3 mm broad, spreading and contiguous, bright green, slightly pitted on upper surface, outer margins entire and slightly decurved, inner margins not raised, stamens with pubescent or glabrous pedicels, bracteoles filiform and pubescent; ovary globose, slightly pubescent to glabrous, nearly sessile; styles 2–2.5 mm long, branched and slightly to strongly recurved in upper half. Capsule 6–7  mm diam., almost spherical, glabrous, grey-green, sessile. Distribution & Habitat Euphorbia loricata is common in the valley of the Olifants River, from near Citrusdal to Klawer and Vredendal. It is also widespread, though much less common and less known, in the drier parts of the Cedarberg, especially on its eastern flank overlooking the Tanqua Karoo. It is also found over a wide area of the escarpment, from Vanrhyn’s Pass to near Loeriesfontein and on the eastern side of the Tanqua Karoo from Gannaga Pass to Verlatekloof and southwards along the escarpment to Komsberg Pass. Here it mainly occurs on the upper slopes of the mountains. Slightly more isolated are the many plants found in the hills around Matjiesfontein in the Laingsburg district.

98

2  Euphorbia subg. Athymalus

Fig. 2.63.  Distribution of Euphorbia loricata (© PVB).

This species is mainly found on dry slopes in very rocky situations on scanty soils derived from sandstone, shale or tillite. In the west, between Citrusdal and Klawer, it is associated with low trees and large shrubs, especially of Montinia caryophyllacea and with many succulents, especially of the succulent Aizoaceae. On the eastern slopes of the Cedarberg it is a constituent of low, mainly succulent vegetation among low sandstone outcrops, while on the slopes of the escarpment it is often one of relatively few succulents in a fairly dense covering of low, non-succulent shrubs. Around Matjiesfontein it again occurs with many other succulents and many short shrubs of the Asteraceae, on gentle to steep, dry slopes among rocks and on soils derived from tillites of the Dwyka series, often with scattered boulders of Witteberg quartzite lying around. Diagnostic Features & Relationships Euphorbia loricata is a shrubby species in which the stem gradually develops a series of subterranean, swollen, sausage-like roots from which finer, fibrous roots arise. Plants are of very variable size. Along the Olifants River Valley (as for example from north of Citrusdal to around Clanwilliam), specimens often reach and exceed 1 m tall, they are somewhat taller than broad (with the stem the tallest part of the plant) and they are usually fairly laxly and slenderly branched (Fig. 2.64). In drier areas (such as the eastern Cedarberg, Botterkloof and the slopes of the escarpment at the southern and eastern edges of the Knersvlakte), plants of E. loricata are much more compact. Here they form more mound-like shrubs that are usually not more than 0.5 m tall and are often a great deal broader than tall,

Fig. 2.64.  Euphorbia loricata, laxly branched shrub ± 1 m tall, PVB 299, north of Citrusdal, Olifants River Valley, South Africa, 3 Jul. 2004 (© PVB).

2.1 Sect. Anthacanthae

in which the stem is difficult to distinguish (Figs. 2.65 and 2.66). The smallest and most densely branched plants occur along the escarpment from Gannaga Pass to Komsberg and around Matjiesfontein (Figs.  2.67 and 2.68). In such mounds the branches are densely packed and may be very hard to separate and there is no obvious stem. Along with changes in stature, one will also notice a change in the spines: along the Olifants River Valley the spines are longer and dry out to be brown and spiky, though they are not very sharp-tipped, while in the drier areas the spines are shorter,

Fig. 2.65.  Euphorbia loricata, ± 0.3 m tall, PVB 9293, Ramkraal, eastern Cedarberg, South Africa, 14 Sep. 2002 (© PVB).

Fig. 2.66.  Euphorbia loricata, mound ± 0.6 m diam., Blinkberg Pass, southern Cedarberg, South Africa, 10 Jul. 2006 (© PVB).

99

drying out pale brown to white. They are somewhat variable in this respect in the eastern Cedarberg, but dry out very strikingly white along the escarpment from Gannaga Pass to Komsberg and around Matjiesfontein. In these drier areas the spines are much sharper and form a relatively impenetrable protection for the plant. With the onset of rains (in areas where some summer rains may fall, E. loricata is usually stimulated most by the winter component of the rainfall), new growth occurs and the leaves appear at the tips of the new tubercles and change the appearance of the

100

2  Euphorbia subg. Athymalus

Fig. 2.67.  Euphorbia loricata, dense mound ± 0.3 m diam., PVB 2941, Komsberg Pass, east of Sutherland, South Africa, 30 May 2005 (© PVB).

Fig. 2.68.  Euphorbia loricata, plant without leaves and somewhat hedgehog-like, PVB 2941, Komsberg Pass, east of Sutherland, South Africa, 11 Jan. 1988 (© PVB).

plant from a dry (though often quite conspicuous) white to brown mound- to hedgehog-like shrub to a grey-green clump that may be surprisingly inconspicuous. The leaves, which persist for most of the moist season, are comparatively large (usually around 50 mm long) but are fairly slender and usually broadest near their tips (this obovate shape is more pronounced in drier areas, where they are often also broader), tapering gradually to the base and more abruptly

to the tip. They are not at all succulent and are covered with a fine, waxy layer which lends them their characteristically greyish colour. Flowering in E. loricata takes place between May and early July. Each plant is either male or female and there is no way of telling the sexes apart when they are not flowering other than by examining the remains at the tips of old fertile short-shoots. As soon as new growth begins in autumn, the

2.1 Sect. Anthacanthae

Fig. 2.69.  Euphorbia loricata, branch of male plant, PVB 299, north of Citrusdal, Olifants River Valley, South Africa, 3 Jul. 2004 (© PVB).

Fig. 2.70.  Euphorbia loricata, male cyathia, Aasvoëlbos, west of Matjiesfontein, South Africa, 30 Sep. 2008 (© PVB).

101

sterile and fertile short-shoots begin to develop in the axils of the tubercles. The short-shoots are usually 30–40  mm long, bearing small, leaf-like bracts that increase in size towards their tip, where several usually much broader bracts closely subtend the apical cyathium. Male cyathia are densely packed with florets and are considerably broader than the female ones, which contain the almost spherical ovary, several small sterile male florets around its base and some tiny bracteoles. The round and only very obscurely 3-angled capsules usually ripen between September and early November, by which time the leaves have also dropped off. In Bruyns et al. (2006), E. eustacei was maintained as distinct from E. loricata. However, this does not reflect the position as it is now known. Dense, low-growing and mound-forming plants with slightly broader, more obovate leaves and spines drying out white have always been taken as typical of E. eustacei and were assumed to be restricted to the Matjiesfontein area (White et al. 1941), while the more diffuse, taller plants with narrower leaves and spines drying out brown that are characteristic of the valley of the Olifants River between Citrusdal and Clanwiliam were traditionally taken as typical of E. loricata. Nevertheless, White et  al. (1941) hinted at a wider distribution for E. loricata and included some more densely branched plants (e.g. fig. 264) in their concept of this species. Now that the respective distributions have become better known it has been found that there is a gradation from the one into the other as one progresses eastwards from the valley of the Olifants River towards the escarpment (rather than two disjunct and distinct

Fig. 2.71.  Euphorbia loricata, capsules, Sandkraal, foot of Gifberg, South Africa, 14 Jul. 2009 (© PVB).

102

species each confined to particular areas) so that E. eustacei and E. loricata are ecotypes of one considerably more widespread species.

2  Euphorbia subg. Athymalus

Euphorbia loricata is similar to E. multifolia and also to E. oxystegia. In all three species the plant develops a series of swollen roots, the stem and branches are covered with small tubercles (that are not clearly fused into angles), each

Fig. 2.72.  Euphorbia loricata. A, side view of male cyathium (scale 2 mm, as for B, C). B, male cyathium from above. C, side view of female cyathium. D, anthers and bracteole (scale 1 mm, as for E). E, female floret with bracteoles. Drawn from: PVB 299, north of Citrusdal, Olifants River Valley, South Africa (© PVB).

of which bears a prominent non-succulent, deciduous leaf, all are armed with persistent twiggy or spine-like sterile and fertile short-shoots and the plants are unisexual. Euphorbia oxystegia differs by having relatively few, diffusely spreading branches with very much more widely separated tubercles and with persistent short-shoots that remain comparatively soft and do not become spiky. In E. oxystegia the leaves are also distinctly petiolate, while they are sessile in the other two. The cyathia in E. oxystegia are shallower and less cupular-urceolate than in the other two and are much more concealed by the bracts that surround them at the tips of the peduncles. As with the petioles and bracts, the female florets (and later the capsules) are more pubescent in E. oxystegia than in the others. The differences between E. loricata and E. multifolia are discussed under the latter.

History Euphorbia loricata was first recorded on the 15th of September 1685 by the artist of the expedition of Simon van der Stel to the Copper Mountains of Namaqualand, when the expedition came to and crossed the Olifants River a little south of the present-day town of Klawer (Wilson et al. 2002; also Folio 310 in the volume of the Codex witsenii known as Icones plantarum et animalium, Macnae and Davidson 1969). It is generally assumed that this artist was Heinrich Claudius (Gunn and Codd 1981; Wilson et  al. 2002). Euphorbia loricata is therefore one of the first species of Euphorbia to have been recorded from South Africa. When he described E. loricata, Lamarck cited figures by Plukenet (1692: t. 230, fig. 5) and Petiver (1709–11: t. 86, fig. 5). These figures all appear to be copies (sometimes slightly modified) of Claudius’ painting in the Codex

2.1 Sect. Anthacanthae

103

Fig. 2.73.  Euphorbia loricata from further east ‘eustacei’. A, side view of male cyathium (scale 2 mm, as for B, C). B, male cyathium from above. C, side view of female cyathium. D, anthers (scale 1 mm, as for E). E, female floret. Drawn from: PVB 2747, Pienaarsfontein se Berg, Klein Roggeveld, west of Sutherland, South Africa (© PVB).

Witsenii, which appears to have been unknown to Lamarck. There is no specimen of E. loricata in Lamarck’s herbarium in Paris and so Plukenet’s plate was selected as lectotype (Bruyns 2012). The plants named Euphorbia eustacei were sent to Kew by C.E. Pillans in October 1912 and survived there in cultivation until June 1914. It appears that they might have been known to Marloth around 1905 (though the specimen Marloth 5160 at PRE has no date of collection). Euphorbia multifolia A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 962 (1941). Lectotype (Bruyns 2012): South Africa, Cape, 30 miles from Laingsburg towards Ladismith, Aug. 1939, Herre (PRE). Unisexual subspiny many-branched glabrous usually dense mound-forming succulent 80–500 × 0.15–1 m, branching mainly from base of similar stem, stem tapering beneath ground then giving rise to several swollen fusiform roots from which fibrous roots arise. Branches spreading to erect,

cylindrical to slightly club-shaped, 15–30 mm thick, simple to more rarely rebranching above, covered densely with tubercles, smooth, pale green becoming brown then grey with age; tubercles conical, 2–4  mm long and vertically arranged (but not fused) into 9–12 obscure spiralling rows, with fine grooves around and separating them in early stages only, with solitary simple subspiny stalks in their axils; subspiny stalks developing from modified persistent sterile short-shoots, curved upwards, becoming woody, rigid, 20–55 × 1–1.5 mm, with several small alternating glabrous obtuse bracts to 2 mm long and 1.5 mm broad along length, initially pale glaucous green, later drying yellow-brown and then grey, glabrous; leaves on tips of new tubercles towards apices of branches and stem, 15–80 × 2–4.5 mm, spreading, gradually deciduous and persisting for most of season, linear-oblong and channelled above for most of length, subtruncate to obtuse, tapering to narrow base, sessile, pale glaucous green, glabrous or minutely puberulous near apex and along margins. Synflorescences many per branch towards apex, each solitary in axil of tubercle on unbranched green short-

104

2  Euphorbia subg. Athymalus

Fig. 2.74.  Euphorbia multifolia, on sandstones/ quartzites of the Witteberg series, steep, north-facing slopes with Aloe ferox and many other succulents, four mounds of Euphorbia are visible in the foreground, PVB 3716, Bosluiskloof, SE of Laingsburg, South Africa, 2 Sep. 2008 (© PVB).

shoot 15–55 mm long (usually longer in males than females) with 1 terminal unisexual cyathium, glabrous, with 2–4 broadly ovate glabrous green bracts 3–5 mm long with fine marginal cilia just beneath and slightly overtopping cyathium, others below smaller; cyathia cupular to slightly urceolate, glabrous, 4–5.5 mm broad (broader in male than female), with 5 slightly pubescent lobes with toothed margins, green; glands 5, transversely rectangular, 1.5–2.5 mm broad, spreading and contiguous, yellowish green, slightly pitted on upper surface, outer margins usually faintly toothed

Fig. 2.75.  Distribution of Euphorbia multifolia (© PVB).

and slightly decurved, inner margins slightly raised, stamens with pubescent pedicels, bracteoles filiform and pubescent; ovary globose, pubescent, nearly sessile; styles 2–2.5  mm long, branched and slightly recurved in upper half. Capsule 5–6  mm diam., almost spherical, pubescent, grey-green, sessile. Distribution & Habitat Euphorbia multifolia occurs in the dry mountains that flank the Swartberg range to the south and north, along or just

2.1 Sect. Anthacanthae

beyond the northern edge of the Klein Karoo. These mountains are mainly composed of quartzites of the Witteberg Series and run in a west-east direction from Touwsriver to a little north-east of Calitzdorp. Euphorbia multifolia is found primarily on substrates derived from sandstones (only very occasionally is it found on tillites of the Dwyka Series as, for example, north-east of Calitzdorp), where it usually grows in crevices between large rocks on north-facing slopes. The surrounding vegetation varies from arid fynbos to karroid scrub or bush, usually with succulents forming a major component.

Fig. 2.76.  Euphorbia multifolia, wedged tightly between quartzitic rocks, PVB 2420, SE of Konstabel, South Africa, 3 Jan. 1985 (© PVB).

Diagnostic Features & Relationships In E. multifolia the plants form dense, cushion-like mounds up to 1  m in diameter, consisting of hundreds of tightly packed branches which develop around the very similar stem

105

and fill up the crevices between rocks. The rootstock consists of several swollen roots from which finer, fibrous roots arise. Although these mounds may occasionally reach half a metre in height, they are generally much lower and rarely exceed 20 cm tall. The branches are finely tuberculate and, for most of their length, they bear persistent sterile and fertile shortshoots. These short-shoots form a difficult-to-penetrate protection around each branch as well as interlocking from neighbouring branches to hold the mound together, making it difficult to pry apart. After rain the growing tips of the branches soon become densely covered by the leaves, which completely hide the branches and the persistent short-shoots from view. The pale, grey-green leaves are always long and slender, widening gradually to their truncate or obtuse apex so that they are at their widest just below their tips. They seem to persist for up to a year. Flowering in E. multifolia takes place in August to September. Each cyathium is solitary at the tip of a peduncle with small bracts along its shaft and several larger bracts just beneath the cyathium. Male cyathia are broader than females and both have a somewhat cup-like shape, with conspicuous spreading glands around their margins. The finely pubescent capsules ripen between October and November. Euphorbia multifolia is similar to the compact, lowgrowing forms of E. loricata formerly known as E. eustacei and shares with these the mound-like habit, the manner in which the plant is covered with persistent sterile peduncles and the prominent leaves. The branches in E. multifolia are always thicker than in E. loricata. In E. multifolia the persistent short-shoots are much softer than the rigid thorns of ‘eustacei’ and are instead somewhat pliable like those of E. loricata as it occurs in the Olifants River Valley, but they do not become sharp-tipped as those of E. loricata always are. The leaves of E. multifolia are much narrower than in any of the forms of E. loricata and they persist longer in E. multifolia. Floral differences between the two are less easily observed: the cyathium is more urceolate in E. multifolia, the cyathial glands are more rectangular in E. multifolia (more rounded and elliptic in E. loricata) and their outer margins are slightly toothed in E. multifolia and smooth in E. loricata. Furthermore, the capsule is more densely pubescent in E. multifolia in its earlier stages. History Euphorbia multifolia appears to have been gathered by G.G. Smith and H. Herre at the same time in 1938 (though August 1939 is written on the existing specimen), possibly

106

2  Euphorbia subg. Athymalus

Fig. 2.77.  Euphorbia multifolia, large mound ± 1 m diam., PVB 3716, Bosluiskloof, SE of Laingsburg, South Africa, 2 Sep. 2008 (© PVB).

Fig. 2.78.  Euphorbia multifolia, with Crassula rupestris and C. arborescens, PVB 3716, Bosluiskloof, SE of Laingsburg, South Africa, 2 Sep. 2008 (© PVB).

Fig. 2.79.  Euphorbia multifolia, male cyathia, PVB 3716, Bosluiskloof, SE of Laingsburg, South Africa, 2 Sep. 2008 (© PVB).

Fig. 2.80.  Euphorbia multifolia, female cyathia, PVB 3716, Bosluiskloof, SE of Laingsburg, South Africa, 2 Sep. 2008 (with Crassula rupestris) (© PVB).

2.1 Sect. Anthacanthae

on a joint expedition into the Laingsburg district. It was on this material that White et al. (1941) based their description of this, at that stage, undescribed species. However, these were not the first records of it, since it was also known to Marloth, who had collected it in May 1907 in the Sand River Mountains near Prince Albert. This collection was listed under E. loricata by Brown (1915) and by White et al. (1941) and they figured another collection of it, made by Max Otzen

107

between Prince Albert and Laingsburg also under E. loricata, but in this case their identification was only tentative. At the time of publication of White et al. (1941), E. multifolia was poorly known, so that they were only able to describe it incompletely. Subsequently, with the increasing exploration of the mountainous borders of the Klein Karoo, it has become much better understood and the distribution can now be plotted with some confidence.

Fig. 2.81.  Euphorbia multifolia. A, side view of male cyathium (scale 2 mm, as for B, C). B, male cyathium from above. C, side view of female cyathium. D, anthers, bracteole and sterile female in dissected male cyathium (scale 1 mm, as for E). E, female floret with bracteoles in dissected female cyathium. Drawn from: PVB 3716, Bosluiskloof, SE of Laingsburg, South Africa (© PVB).

Euphorbia oxystegia Boiss., Cent. Euphorb.: 27 (1860). Lectotype (Bruyns 2012): South Africa, Cape, between Goedemanskraal and Kaus, Drège (S; BM, K, P, TCD, W, iso.). Unisexual spineless few-branched (rarely many-branched) minutely pubescent succulent shrub 50–300  mm tall and broad, irregularly branching from base and above on similar stem, stem with swollen partly subterranean base tapering into several swollen fusiform roots from which fibrous roots arise. Branches (and stem) spreading to erect, cylindrical, 6–20 mm thick, simple or rebranching above, sparsely tuberculate, smooth, pale green becoming brown then grey with age, spineless but with persistent grey remains of peduncles; tubercles conical, 2–4 mm long and scattered, with solitary simple persistent short-shoots in their axils, becoming obscure

with age; leaves on tips of new tubercles towards apices of branches and stem, 25–100 × 6–16 mm, spreading, gradually deciduous, lanceolate to oblanceolate and usually flat above becoming slightly folded towards petiole, obtuse to apiculate, tapering to slender finely puberulous petiole 10–30 mm long, pale glaucous green, glabrous. Synflorescences few to several per branch towards apex, each solitary in axil of tubercle on usually unbranched green to red short-shoot 5–100 mm long (usually longer in males than females) with 1 terminal unisexual cyathium and sometimes 1–4 more arising on branches near apex, finely pubescent, with 2–5 broadly ovate finely pubescent green to red bracts 5–11 × 4–7 mm just beneath and slightly overtopping cyathium, others below more lanceolate and longer (those around female cyathium often shorter and broader than around male); cyathia bowl-shaped, glabrous, 5–7 mm broad (broader in male than female), with

108

5 pubescent lobes with deeply toothed margins, green; glands 5, elliptic, 2–2.5 mm broad, spreading and contiguous, dark yellow-green, slightly pitted on upper surface, outer margins entire and spreading, inner margins not raised, stamens with pubescent or glabrous pedicels, bracteoles filiform and pubescent; ovary globose, pubescent, nearly sessile; styles ± 3 mm long, branched and spreading in upper third, often red-green. Capsule 7–8  mm diam., almost spherical, finely pubescent, grey-green, sessile.

2  Euphorbia subg. Athymalus

Distribution & Habitat Euphorbia oxystegia is found only in Namaqualand, where it has been recorded from Wildeperdehoek Pass (south-west of Springbok) to the Harras Mountains, which lie south-west of Anenous Pass and more or less west of Steinkopf. White et al. (1941) illustrated a specimen that ostensibly came from Stinkfontein (= Eksteenfontein), but there are no records of E. oxystegia from so far to the north and it is doubtful that it occurs there.

Fig. 2.82.  Distribution of Euphorbia oxystegia (© PVB).

Plants may be found on steep, west-facing slopes of the escarpment among large rocks and fairly dense vegetation composed mainly of succulent shrubs. They also occur on much more gentle slopes (even sometimes east-facing) and flat areas ontop of and slightly behind the escarpment among dense stands of Pteronia incana and renoster. Here they are usually tightly wedged among densely packed stones and often grow inside or around the bases of other bushes. Diagnostic Features & Relationships In E. oxystegia the stem arises from a distinctly swollen base, beneath which a small cluster of thickened roots radiates. The base and these thickened roots may be some distance below the ground and are generally well protected from predation by their depth in hard soil or by plenty of stones packed above and around them (Fig. 2.83–2.85). The stem is

usually solitary and somewhat irregularly erect to 30 cm tall but, if it is eaten off, one or more similar branches may replace it to form a loose cluster. Scattered, often quite widely along the stem, are many irregularly arranged tubercles which project a short distance on younger growth and gradually merge with the surface with age. The upper part of the stem (from at most one year previous to the present) is green and below that it becomes covered with a soft and thin, grey bark. Around the beginning of the rainy season (often as early as March or April) new tubercles develop at the tips of the stem and branches and each bears a leaf. The leaves are on petioles of very variable length and they are a glaucous green, so that the loose rosettes that they form round the tops of the stem and branches are quite conspicuous in habitat. When they are gone, the plants are much more difficult to locate.

2.1 Sect. Anthacanthae

109

Fig. 2.83.  Euphorbia oxystegia, heavily grazed, among rocks, PVB 2712, west of Bulletrap, north of Springbok, South Africa, 18 Oct. 2008 (© PVB).

Fig. 2.84.  Euphorbia oxystegia, ± 30 cm tall, among rocks and dense scrub, PVB 2712, west of Bulletrap, north of Springbok, South Africa, 18 Oct. 2008 (© PVB).

Flowering occurs between May and July. The cyathia are borne at the tips of long, slender, sometimes branching shortshoots. These arise in the axils of the tubercles a little below the most recent set of leaves, but generally ascend past them to project above the plant. Along its shaft, each short-shoot bears small bracts while, at its tip, the solitary cyathium nestles among a whorl of broad and short bracts that are similarly coloured to the leaves, though not nearly as long as they are. After flowering is over these short-shoots generally persist for a few seasons as dry stalks but eventually they wear off and disappear.

Euphorbia oxystegia is probably closely related to E. loricata and E. multifolia. Differences between it and E. loricata are discussed under the latter. History Euphorbia oxystegia was discovered by J.F. Drège and his brother Carl between August and September of 1830  in Namaqualand. The exact location of Kaus and Goedemanskraal, mentioned on the type specimen, is not known today. This material was annotated by the botanist Ernst Meyer as ‘Euphorbia bupleurifolia Jacq.’ but Boissier

Fig. 2.85.  Euphorbia oxystegia, short male plant with peduncles around 10 cm long, PVB 2674, west of Bulletrap, north of Springbok, South Africa, 5 Jul. 1987 (© PVB).

Fig. 2.86.  Euphorbia oxystegia, branch of female plant with capsules and much shorter peduncles than in previous figure, PVB 2712, west of Bulletrap, north of Springbok, South Africa, 12 Jul. 2006 (© PVB).

Fig. 2.87.  Euphorbia oxystegia. A, B, side view of female cyathium (scale 2 mm, as for B, C). C, female cyathium from above. D, female floret with bracteoles in dissected female cyathium. Drawn from: PVB 2712, west of Bulletrap, north of Springbok, South Africa (© PVB).

2.1 Sect. Anthacanthae

111

realised that it represented something quite different and gave it the new name Euphorbia oxystegia. It was first recollected in September 1883 by Harry Bolus and again in July 1926 by Gottlieb Meyer in the hills around Komaggas. Even today, it is not known from many collections.

drying into often formidable persistent spine. Leaves minute and caducous (larger in E. pulvinata). Cyathia surrounded by small bracts at tip of peduncle, unisexual, with the male cyathia usually ±  twice the diameter of the females. Seeds smooth, sticky.

2.1.2.2 Ser. Meleuphorbia

This series contains 14 species that are restricted to Lesotho, South Africa and Swaziland. They mainly occur in South Africa and most of them are found in the southern part of the Greater Cape Floristic Region (with their westernmost occurrence on the southern Tanqua Karoo), on the Klein Karoo, the Worcester-Robertson Karoo and the southern coastal plain. From here they are found eastward beyond the margins of the Greater Cape Floristic Region into the Great Karoo and Eastern Cape. Only Euphorbia pulvinata is more widespread, extending into the tropical parts of South Africa and into the small, mountainous countries of Lesotho and Swaziland. Of the species here, E. polygona may form large, cactuslike shrubs to 2 m tall. Others form shrubs of intermediate size, as in E. heptagona. Yet others, like E. susannae and some forms of E. pseudoglobosa, have a nearly geophytic habit, where the stem and branches are almost entirely subterranean and only their apices protrude above the ground. In E. meloformis and E. obesa the plant consists of an

Euphorbia ser. Meleuphorbia (A.Berger) Bruyns, Taxon 62: 1194 (2013). Euphorbia sect. Meleuphorbia A.Berger, Sukkul. Euphorb.: 10, 101 (1906). Euphorbia subsect. Meleuphorbia (A.Berger) Pax & K.Hoffm. in Engl., Nat. Pflanzenfam. 19c: 216 (1931). Type (designated by Pax 1921): Euphorbia meloformis Aiton. Unisexual (rarely bisexual in E. polygona) succulent shrub or dwarf succulent, sometimes consisting of short unbranched stem only, with mostly indistinguishable stem and branches along which tubercles joined into distinct angles, mostly armed with stout spines (except E. obesa, E. pseudoglobosa, E. tubiglans) derived from slender, non-succulent usually sterile short-shoots in the axils of the tubercles (one per axil except in E. polygona), short-shoots initially soft and with several minute caducous leaf-rudiments, later

Fig. 2.88.  Distribution of Euphorbia ser. Meleuphorbia (showing number of species per half-degree square) (© PVB).

112

unbranched, often quite short but relatively stout stem. Quite formidable spines, developing from the persistent sterile peduncles, are characteristic of many of these species. However, their presence is not diagnostic for the series: they are intermittently present in E. polygona (there are areas where all the plants of E. polygona are spineless) and spines

2  Euphorbia subg. Athymalus

do not develop at all in E. obesa, E. susannae and E. tubiglans and in some forms of E. pseudoglobosa. In E. meloformis, where sterile peduncles are absent, the fertile peduncles are persistent, but they do not become spiny. In all species the seeds are somewhat sticky, readily adhering to ones fingers when handled.

Key to the species of ser. Meleuphorbia 1. Stem partly to wholly subterranean and mostly hidden from sight with branches arising around its apex and often from below ground............................................................................................................................................2. 1. Stem mostly to wholly above ground, branches (if any) arising near base of stem...........................................................5. 2. Stem and branches ± flush with surface and scarcely rising above it, rarely protruding more than 20 mm.......................3. 2. Branches rising usually at least 50 mm above surface but often 150 mm or more long above ground..............................4. 3. Stem and branches with 12 to 16 angles, tubercles along angles prominent and tapering to fine (often recurved) point................................................................................................................................E. susannae 3. Stem and branches with 5 to 10 angles, tubercles along angles low and obtuse....................................E. pseudoglobosa 4. Cyathial glands almost flat on upper surface, styles < 2 mm long, bracts 2–3 mm long, branches often with persistent dried peduncles......................................................................................................E. pseudoglobosa 4. Cyathial glands with lateral margins curved inwards to form tube, styles > 2 mm long, bracts 3–6 mm long and often exceeding length of cyathium, branches without persistent dried peduncles..................................E. tubiglans 5. Sterile short-shoots absent, plant usually consisting of single subglobose stem only (branches rarely present and then also subglobose).....................................................................................................................................6. 5. Sterile short-shoots mostly present, plant usually consisting of cylindrical stem and many similar branches.....................7. 6. Fertile short-shoots all deciduous, stem dull grey-green to brown with many dull purple horizontal bands, angles low and obtuse, leaf-rudiments not exceeding 2 mm long..........................................................................E. obesa 6. Fertile short-shoots usually persistent, stem bright shiny green usually with several darker green or purple horizontal bands, angles prominent and acute (especially towards apex of stem), leaf-rudiments often reaching 5 mm long......................................................................................................................................E. meloformis 7. Sterile and/or fertile short-shoots branched (sometimes only fertile ones branched)......................................................14. 7. Neither sterile nor fertile short-shoots branched (rarely sterile ones absent and fertile ones non-persistent)..................8. 8. Sterile short-shoots absent, fertile short-shoots soon deciduous and plant spineless......................................E. polygona 8. Sterile short-shoots usually present and spinescent, fertile short-shoots often persistent.................................................9. 9. Short-shoots (forming spines) in groups of 3–5 (7) in axils of tubercles….................................…................E. polygona 9. Short-shoots (forming spines) solitary in axils of tubercles............................................................................................10. 10. Outer branches of clump rhizomatous and spreading away from stem beneath ground.................................E. cumulata 10. Branches not rhizomatous, usually densely clump-forming around stem, rarely creeping or scandent.........................11. 11. Shaft of male fertile short-shoot hidden by bracts............................................................................................................12. 11. Shaft of male fertile short-shoot clearly visible between bracts.......................................................................................13. 12. Plants forming densely and tightly packed mounds, angles on branches acute, cyathia often purple............E. pulvinata 12. Plants forming dense clumps but not tightly packed into mounds, angles on branches obtuse, cyathia green or yellowish green.........................................................................................................................................E. ferox 13. Tubercles on stems and branches separated by impressed horizontal lines, leaf-rudiments 2–6 mm long, bracts on fertile short-shoots spreading.....................................................................................................E. mammillaris 13. Tubercles on stems and branches not clearly separated by impressed horizontal lines, leaf-rudiments 1–2 mm long, bracts on fertile short-shoots pressed to shaft and to cyathium..............................................E. heptagona 14. Sterile and fertile short-shoots becoming branched as they develop, each consisting of blunt-tipped central shaft (sometimes initially bearing a cyathium) with whorl of sharp sterile spines surrounding it..........................E. stellispina 14. Sterile short-shoots unbranched, fertile short-shoots branched.......................................................................................15. 15. Mature plants 0.6–3 m tall, young branches bright green, leaf-rudiments (2) 3–8 mm long, bracts just beneath cyathia 3–5 mm long, pedicels of male florets pubescent................................................................E. pentagona 15. Mature plants 0.2–0.4 m tall, young branches dull blue-green, leaf-rudiments 1–2 mm long, bracts just beneath cyathia 1–2 mm long, pedicels of male florets glabrous.......................................................................E. pillansii

2.1 Sect. Anthacanthae

Euphorbia cumulata R.A.Dyer, Rec. Albany Mus. 4: 92 (1931). Type: South Africa, Cape, Botha’s Ridge, 10 miles from Grahamstown of Queen’s Road, Dyer 669 (GRA, holo.; K, iso.). Unisexual spiny rhizomatous many-branched glabrous succulent 80–300  mm × 0.2–1  m, branching from base of similar stem and spreading by short rhizomes to eventually form clump, with fibrous roots. Branches (and stem) ascending to erect from slender underground horizontally spreading base, cylindrical to very slightly club-shaped and 7- to 10-angled, 20–30  mm thick, simple, partly covered with tubercles, smooth, green becoming grey with age; tubercles obtusely conical, 2–4 mm long with transversely rectangular base (if bases distinguished), vertically fused into 7–10 very low obtuse angles, with broad triangular to nearly flat vertical grooves between angles, with solitary simple spines in their axils; spines developing from modified persistent sterile short-shoots, straight, rigid, 10–40 × 1–1.5 (3) mm, with several minute alternating glabrous spathulate bracts to 2 × 1 mm, initially green or red, drying to pale grey-brown, minutely pubescent; leaf-rudiments on tips of new tubercles towards apices of branches and stem, 2–4 × 1–1.5 mm, erect to spreading, fleeting, lanceolate, acute, sessile, minutely pubescent. Synflorescences many per branch towards apex, solitary in axil of tubercle on unbranched red to green

Fig. 2.89.  Distribution of Euphorbia cumulata (© PVB).

113

peduncle 2–10  mm long (usually longer in males than females) with 1 terminal unisexual cyathium, minutely pubescent, with 5–6 lanceolate minutely pubescent green bracts 1–1.5 mm long beneath cyathium; cyathia cupular to conical, minutely pubescent, 4–6  mm broad in male and 3 mm broad in female, with 5 deeply incised lobes, green to suffused with red; glands 5, transversely elliptic, 1–2  mm broad, ascending, widely separated, dark green, swollen and deeply wrinkled or pitted above, outer margins entire, stamens with pubescent pedicels, bracteoles filiform and slightly flattened and densely pubescent along margins especially towards apex; ovary globose, finely pubescent, borne on short pedicel ±  0.2  mm long; styles 3–4  mm long, branched and recurved in upper third to half. Capsule 6–8  mm diam., almost spherical, pubescent to becoming glabrous later, pale grey-green, sessile. Distribution & Habitat Euphorbia cumulata is only known in the dry parts of the Fish River Valley north of Grahamstown, where it is found from around Carlisle Bridge and Hellspoort in the west to Committees and Hunt’s Drift in the east. In these areas, E. cumulata is always associated with karroid, succulent scrub and grows in flat or gently sloping areas on shale substrates. Here the vegetation may be dominated by species of Euphorbia such as E. caerules-

114

cens, E. huttoniae and E. pentagona. Euphorbia cumulata is one of the smaller members of Euphorbia in this scrub, with the plants usually growing around the bases of other shrubs.

2  Euphorbia subg. Athymalus

Diagnostic Features & Relationships Euphorbia cumulata forms diffuse to dense clumps of more or less erect branches that split off from the stem and previous branches near their bases and spread away from them

Fig. 2.90.  Euphorbia cumulata, large plant ± 0.5 m broad among branches of old plant of Rhigozum obovatum, PVB 13533, north of Riebeeck East, South Africa, 24 Oct. 2018 (© PVB).

slightly by means of short, slender, underground rhizomes. The branches are green above and covered with a greyish corky bark lower down. They are ornamented with usually very shallow grooves between the low and quite inconspicu-

Fig. 2.91.  Euphorbia cumulata, showing prominent purplish spines, PVB 13533, north of Riebeeck East, South Africa, 24 Oct. 2018 (© PVB).

ous angles, with sharp spines along the crests of these angles. These spines vary greatly in length and may be longer and far more rigid in plants from the western areas around Carlisle Bridge than they are in plants from the eastern areas,

2.1 Sect. Anthacanthae

around Fort Brown, for example. When young, the spines are finely pubescent and usually reddish, bearing a few tiny bracts especially towards their tips, but these soon fall off to leave a sharp spike. These sterile short-shoots are slightly longer than the fertile short-shoots.

115

In E. cumulata, cyathia develop and mature any time between September and March. The cyathia of female plants are considerably smaller than those in male plants. In the males there are many bracteoles with pilose apices and these give the top of the cyathium between the cyathial lobes a

Fig. 2.92.  Euphorbia cumulata, with two small rhizomatous branches well to the right of the rest of the plant, PVB 13533, north of Riebeeck East, South Africa, 24 Oct. 2018 (© PVB).

furry appearance. However, probably the most notable features of the otherwise inconspicuous cyathia are the distinctly swollen (though small) glands with their pitted and convoluted upper surfaces and the quite long styles which are divided above into three spreading lobules. The female cyathia are borne on particularly short peduncles and the tops

of branches of female plants often become hidden by the cluster of finely furry capsules that may be produced after flowering is over. Once flowering and fruiting is finished, E. cumulata is another of those species in which the male and female plants can usually be distinguished by the differing remains at the tips of the persistent peduncles.

Fig. 2.93.  Euphorbia cumulata, male cyathia, PVB 10988, Fort Brown, South Africa, Nov. 2008 (© PVB).

Fig. 2.94.  Euphorbia cumulata, female cyathia, PVB 10988, Fort Brown, South Africa, Nov. 2008 (© PVB).

116

2  Euphorbia subg. Athymalus

Euphorbia cumulata is similar in appearance to E. mammillaris, in which the plant also branches mainly from the base, though E. mammillaris is never rhizomatous in the manner of E. cumulata. The branches of E. cumulata rarely rebranch above, whereas those of E. mammillaris do so occasionally. In addition, the tubercles forming the angles are far less conspicuous in E. cumulata. In E. cumulata the male and female cyathia are of very different sizes, while in E. mammillaris they are similar in size. They are also differently shaped, with minute bracts at their bases, contrasting with the larger, considerably more spreading bracts in E. mammillaris. The styles are also much more deeply divided in E. cumulata than in E. mammillaris. Finally, also (and unusually for this series), the ovaries and capsules are pubescent in E. cumulata, while they are glabrous in E. mammillaris. Superficially E. cumulata also resembles both E. hepta­ gona and E. pentagona. However, in both of these, all branching off the stem takes place above the ground and the branches repeatedly rebranch above. In both E. heptagona and E. pentagona the glands are not as noticeably plump, nor do they have the pitted upper surfaces of those of E. cumulata and the styles are not as deeply divided either. Euphorbia cumulata shares with E. heptagona and E. pentagona the very small bracts along the sterile short-shoots and it shares the very small bracts just beneath the cyathia on the fertile short-shoots with E. heptagona. These are noticeably different from the much larger bracts subtending the cyathia in E. pentagona. In both E. heptagona and E. pentagona the capsules are glabrous, while in E. cumulata they are finely pubescent.

Fig. 2.95.  Euphorbia cumulata, branch with capsules, PVB 10988, Fort Brown, South Africa, Oct. 2012 (© PVB).

History Euphorbia cumulata was first recorded by R.A.  Dyer, who collected it on Botha’s Ridge, 9–10 miles north of Grahamstown in November 1926 and published it on 9 January 1931 (Dyer 1931). In subsequent years he

pressed several specimens of it from different parts of the Albany District, but otherwise very few collections have been made and the species remains relatively little-known.

2.1 Sect. Anthacanthae

117

Fig. 2.96.  Euphorbia cumulata. A, male cyathium from side (scale 2 mm). B, female cyathium from side (scale 2 mm). C, anthers and bracteoles (scale 1 mm, as for D). D, female floret. Drawn from: PVB 10988, Fort Brown, South Africa (© PVB).

Euphorbia ferox Marloth, Trans. Roy. Soc. South Africa 3: 122 (1913). Type: South Africa, Cape, Klipplaat, 1905, Marloth 5147 (PRE, holo.; BOL, iso.). Unisexual spiny many-branched glabrous low shrub to mound-forming succulent 0.07–0.3 × 0.2–1 m, branching mainly from base and rarely above from similar stem, with fibrous roots on stem and on bases of many branches. Branches spreading to erect, cylindrical to slightly clubshaped and (8) 9- to 16-angled, (10) 20–45 (75) mm thick, mostly simple, partly covered with tubercles, smooth, dull dark green becoming grey with age; tubercles obtusely conical, 3–5  mm long and vertically fused into (8) 9–16 low obtuse angles, with broad triangular vertical grooves between angles and usually without horizontal lines separating tubercles, with solitary simple spines in their axils; spines from persistent sterile short-shoots, straight, woody, rigid, 8–50 × 1–2 mm, with several small alternating glabrous deltate bracts to 1.5 × 1 mm, initially glaucous pink to red, drying to brown then grey, glabrous; leaf-rudiments on tips of new tubercles towards apices of branches and stem, 2–3 × ±  1 mm, erect, fleeting, lanceolate-oblong and slightly channelled above, acute, sessile. Synflorescences many per branch towards apex, each solitary in axil of tubercle on unbranched green shoot 2–6  mm long (usually longer in males than

females) with 1 terminal unisexual cyathium, glabrous, with 2–4 broadly ovate finely and sparsely ciliate reddish green bracts 3–4  mm long just beneath cyathium hiding shaft of shoot, others below smaller (those around female cyathium shorter and broader than around male); cyathia cupular to slightly urceolate, glabrous, 2–4 mm broad (broader and longer in male than female, 1.5–3 mm long below insertion of glands), with 5 lobes with toothed margins, green; glands 5, transversely elliptic to almost circular, 1–1.5  mm broad, spreading, widely separated to contiguous, dark to bright green, slightly pitted on upper surface, outer margins entire to faintly toothed, stamens with pubescent pedicels, bracteoles filiform and sparsely pubescent; ovary globose, glabrous, raised on short pedicel < 0.5 mm long; styles 1.5–3 mm long, branched and recurved only in upper third. Capsule 5–6 mm diam., almost spherical, glabrous, grey-green, sessile. Euphorbia ferox is widespread on the Great Karoo and on the eastern part of the Klein Karoo between Calitzdorp and Dysselsdorp. It is similar in many respects to E. heptagona and E. mammillaris. The branches in E. ferox are generally comparatively short and thick, forming a much denser and lower clump than in either E. heptagona or E. mammillaris and they are armed with a denser cover of sharper and thicker spines. In E. ferox the cyathia are borne on particularly short

2  Euphorbia subg. Athymalus

118

Fig. 2.97.  Distribution of Euphorbia ferox (● = ssp. ferox; ▲ = ssp. calitzdorpensis) (© PVB).

shoots, with the bracts covering up the short shaft of the shoot and surrounding the base of the cyathium. In E. heptagona and E. mammillaris the cyathia are borne on longer peduncles, with small bracts so that the shaft is clearly visible below the cyathium, especially in the males. The cyathia of E. ferox are generally somewhat urceolate, a shape which is not found in either of the other two. In discussing the differences between E. ferox and E. mammillaris, White et al. (1941) made something of the horizontal darker lines separating individual tubercles along the angles. These are most prominent in E. mammillaris. In some plants of E. ferox they are present on some branches and not on others, while on other plants these horizontal lines are Fig. 2.98.  Euphorbia ferox ssp. ferox, plant with unusually long, weeping spines, west of Steytlerville, South Africa, 22 Oct. 2009 (© PVB).

absent, as White et al. (1941) maintained that they should be for this species. These horizontal lines are generally present in the plants from the eastern Klein Karoo that are now called E. ferox subsp. calitzdorpensis. Their presence is perhaps why these plants were taken to represent E. mammillaris in White et al. (1941: 596–602). Although this feature can often be used to recognise E. mammillaris, it is more useful to separate E. mammillaris from E. heptagona than E. mammillaris from E. ferox. Other features of these plants from the eastern Klein Karoo (the stoutness of their spines, their short peduncles hidden by the bracts, the urceolate shape of the cyathia) suggest that they belong rather under E. ferox. The two subspecies may be separated as follows:

2.1 Sect. Anthacanthae

119

1. Spines densely and evenly distributed along branches, cyathia in small numbers right at apices of branches............subsp. ferox 1. Spines in patches with spineless areas between them along branches, cyathia in large clusters at apices of branches, in males covering upper 1–2 cm of branches................................................................subsp. calitzdorpensis

Euphorbia ferox subsp. ferox Euphorbia captiosa N.E.Br., Fl. Cap. 5 (2): 345 (1915). Type: South Africa, Cape, near Aberdeen, flow. Sept. 1904, Schönland 1661 (K, holo.; GRA, iso.). Plants forming densely grouped clumps to 1  m diam., branches 9- to 16-angled, with spines densely and evenly arranged over surface.

Fig. 2.99.  Euphorbia ferox ssp. ferox, NE of Willowmore, South Africa, 1 Oct. 2006 (© PVB).

Fig. 2.100.  Euphorbia ferox ssp. ferox, plant with densely packed branches, PVB 11946, south of Steytlerville, South Africa, 7 Jun. 2011 (© PVB).

Distribution & Habitat Euphorbia ferox subsp. ferox is widespread in the vast plains of the Great Karoo known as Die Vlakte, lying between Beaufort West, Willowmore, Steytlerville and Graaff-Reinet. It occurs even more widely, though, as for example in the valley of the Fish River around Cradock. Similar plants (usually referred to as E. aggregata) occur in the mountains between Cradock and Fauresmith and in many respects these are intermediate between E. ferox and E. pulvinata.

120

Plants of subsp. ferox generally grow in flat areas, among low bushes, where they are often locally a significant to dominant feature of the vegetation. They usually occur in hard, loamy ground that may be devoid of rocks, but they may inhabit low, stony ridges as well. Diagnostic Features & Relationships The dense and dangerously spiky clumps of Euphorbia ferox subsp. ferox, usually around half a metre in diameter (and generally not more than 20 cm tall), are a familiar feature of many of the areas where it is found and have to be treated with care and respect when walking among them.

2  Euphorbia subg. Athymalus

These clumps consist of many erect branches, often rooting at their bases and quite loosely joined together there. For most of their length, the branches are armed with very sharptipped, spreading, sterile short-shoots. These are initially soft, pliable and reddish green (with several tiny ‘bracts’ scattered along their length), but rapidly dry out to become woody, very rigid and brown to grey. It is from these formidable spikes that the popular name voetangel is derived. Although the clumps of subsp. ferox never become very tall, they vary greatly in denseness and in any colony there are usually several densely branched, cushion-shaped plants (usually with shorter branches) and others that are more dif-

Fig. 2.101.  Euphorbia ferox ssp. ferox, plant with much more densely packed branches and intermediate with E. pulvinata, PVB 11648, Heuningvlei, SE of Cradock, South Africa, 5 Aug. 2011 (© PVB).

fuse, consisting of fewer, considerably longer branches. These diffuse plants often have stronger and longer spines (often more spreading or even decurved) than the more densely clustered specimens. The angles along the branches are usually somewhat rounded and the upper parts of the branches have a grey-green colour. Very small leaf-rudiments are present briefly on the youngest tubercles of the branches. Flowering is often a comparatively inconspicuous process in subsp. ferox and usually takes place between May and early September (sometimes late flowers appear on some plants where other plants nearby are in fruit). A dense cluster of cyathia develops at the apices of most of the branches and these interrupt the production of spines for a short time. The

peduncles may have a slightly reddish hue from the reddish green of the bracts, while the glands are a dull dark green and secrete plenty of nectar on their faintly pitted, upper surface. Cyathia of both sexes are produced on short peduncles. The peduncles of the males are slightly longer than those in the females, so that in the females the bracts along the peduncle hide most of the cyathium while in the males the cyathium protrudes somewhat more from among the bracts. The male cyathia are slightly broader and longer than the females. Cyathia in specimens from the Great Karoo emit odours that vary from faintly foetid (of long-unwashed socks) to faintly sweet. Capsules usually remain on the plant until around November.

Fig. 2.102.  Euphorbia ferox ssp. ferox, male cyathia, PVB 3148, Prutkraal, Rietbron, South Africa, 5 Jul. 2009 (© PVB).

Fig. 2.103.  Euphorbia ferox ssp. ferox, male cyathia (intermediate with E. pulvinata), PVB 11635, Venstersvlei, NW of Colesberg, South Africa, 3 Aug. 2011 (© PVB).

Fig. 2.104.  Euphorbia ferox ssp. ferox, female cyathia with fairly short peduncles, PVB 13220, Shirlands, Graaff-Reinet, South Africa, 1 Sep. 2016 (© PVB).

Fig. 2.105.  Euphorbia ferox ssp. ferox, capsules, PVB 13220, Shirlands, Graaff-Reinet, South Africa, 1 Sep. 2016 (© PVB).

122

There are problems distinguishing subsp. ferox from some of the western forms of E. pulvinata. Some of these plants have been referred to as E. aggregata, especially by White et al. (1941), but they appear more to represent inter-

2  Euphorbia subg. Athymalus

mediates between the two species. They are referred to in more detail under E. pulvinata and examples are shown here in Fig. 2.101 and 2.103.

Fig. 2.106.  Euphorbia ferox ssp. ferox. A, side view of male cyathium (scale 1 mm, as for rest). B, male cyathium from above. C, male cyathium with bracts removed. D, side view of female cyathium. E. female cyathium with bracts removed. F, anther with rudimentary female and bracteole. G, female floret. Drawn from: PVB 3148, Prutkraal, Rietbron, South Africa (© PVB).

History Subsp. ferox was described by Marloth from material that he collected in 1905 but it was known earlier to Schönland. Schönland’s collection flowered in cultivation in 1904 and he supplied material to N.E. Brown, who then described this as Euphorbia captiosa. This species was distinguished by the straight as opposed to ‘frequently or mostly curved’ spines but White et al. (1941) had no hesitation in reducing it to a synonym of E. ferox. In a letter attached to the part of the type at Kew, Schӧnland mentioned that it was ‘about a foot high and sparingly branched from the base’, but this would

easily correspond to many plants of subsp. ferox, which forms clumps of variable shape and density. Euphorbia ferox subsp. calitzdorpensis Bruyns, Haseltonia 25: 38 (2018). Type: South Africa, 12 miles east of Oudtshoorn, Aug. 1939, Dyer 4056 (PRE, holo.; K, iso.). Plants forming laxly grouped clumps to 0.5  m diam., branches 10- to 12-angled, with spines unevenly scattered over surface.

2.1 Sect. Anthacanthae

123

Fig. 2.107.  Euphorbia ferox ssp. calitzdorpensis, large clumps among various Aizoaceae and Pteronia paniculata, Highgate, Oudtshoorn, South Africa, 2 Dec. 2006 (© PVB).

Distribution & Habitat Euphorbia ferox subsp. calitzdorpensis is found on the eastern side of the Klein Karoo, between Calitzdorp and

Dysselsdorp (east of Oudtshoorn), well south of subsp. ferox. Here, it mostly occurs in flat to gently sloping loamy ground, among low, karroid scrub, with very few stones.

Fig. 2.108.  Euphorbia ferox ssp. calitzdorpensis, Highgate, Oudtshoorn, South Africa, 2 Dec. 2006 (© PVB).

Diagnostic Features & Relationships Plants of subsp. calitzdorpensis have often considerably fewer branches than those of subsp. ferox. They also have more prominent tubercles that are more clearly separated

from each other along the angles and there are far fewer, stouter spines that are often grouped into whorls between which there are distinct, spineless patches. The upper parts of the branches are a fairly bright green.

124

2  Euphorbia subg. Athymalus

Fig. 2.109.  Euphorbia ferox ssp. calitzdorpensis, with Aloe microstigma and Cynanchum viminale, PVB 10572, north of Oudtshoorn, South Africa, 1 Oct. 2006 (© PVB).

In subsp. calitzdorpensis flowering usually takes place in July and August. The glands on the cyathia are brighter green and so the cyathia are more conspicuous than in subsp. ferox. They may develop in remarkably dense clusters and often completely cover the tops of the branches.

Here, also, the relatively few and shorter spines do not hide the cyathia as much as the larger, thicker, very much more densely clustered spines do in the plants from the Great Karoo. The cyathia emit a slightly sweet odour.

Fig. 2.110.  Euphorbia ferox ssp. calitzdorpensis, male cyathia, Highgate, Oudtshoorn, South Africa, 14 Jul. 2011 (© PVB).

Fig. 2.111.  Euphorbia ferox ssp. calitzdorpensis, female cyathia, Highgate, Oudtshoorn, South Africa, 14 Jul. 2011 (© PVB).

History Euphorbia ferox subsp. calitzdorpensis was first collected in 1924 by Edith L. Stephens (sub NBG 246/24 at BOL). White et al. (1941) considered these plants to belong to E. mammil-

laris, which they recorded as coming from the Riversdale and Oudtshoorn districts. Most of the photographs that they showed of E. mammillaris came from the Oudtshoorn area and belong to subsp. calitzdorpensis.

2.1 Sect. Anthacanthae

125

Fig. 2.112.  Euphorbia ferox ssp. calitzdorpensis, capsules, PVB 12584, Dysselsdorp, east of Oudtshoorn, South Africa, 1 Sep. 2018 (© PVB).

Euphorbia heptagona L., Sp. Pl. 1: 450 (1753). Lectotype (Bruyns 2012): Boerh., Ind. Alter. Hort. Lugd.-Bat. 1: fig. opposite p. 258 (1720). Anthacantha desmetiana Lem., Ill. Hort. 5: misc. 64 (1858), nom. superfl. Type: origin unknown, probably from the Cape, cult. Louis Desmet (missing). Euphorbia enopla Boiss., Cent. Euphorb.: 27 (1860). Lectotype (designated here): South Africa, Cape, Witpoortsberg, 2000–3000’, Aug., Drège 8207 (S; BM, K, MEL, MO, P, TCD, W, iso.). Euphorbia heptagona var. fulvispina A.Berger, Monatsschr. Kakteenk. 12: 109 (1902). Type: none cited. Euphorbia morinii A.Berger, Sukkul. Euphorb.: 98 (1906). Type: South Africa, Cape, cultivated material sold by Co. Haage & Schmidt-Erfurt (missing). Euphorbia heptagona var. dentata (A.Berger) N.E.Br., Fl. Cap. 5 (2): 351 (1915). Euphorbia enopla var. dentata A.Berger, Sukkul. Euphorb.: 95 (1906). Lectotype (Bruyns 2012): South Africa, Cape, Witpoortsberge, Drège (P). Euphorbia atrispina N.E.Br., Fl. Cap. 5 (2): 342 (1915). Type: South Africa, Cape, near Prince Albert, received 1912, Pearson (K, holo.). Euphorbia heptagona var. ramosa A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 964 (1941). Type: South Africa, Cape, 17 miles north of Oudtshoorn, Aug. 1939, Dyer 4049 (PRE, holo.; GRA, K, iso.). Euphorbia heptagona var. subsessilis A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 964 (1941). Type: South Africa, Cape, 17 miles east of Ladismith (15 miles west of Calitzdorp), Aug. 1939, Dyer 4067 (PRE, holo.; K, iso.). Euphorbia heptagona var. viridis A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 964 (1941). Type: South

Africa, Cape, 11 miles west of Calitzdorp in Huis River Pass, Aug. 1939, Dyer 4065 (PRE, holo.; K, iso.). Euphorbia atrispina var. viridis A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 964 (1941). Type: South Africa, Cape, 12–15 miles from Montagu near Ouberg Pass, Aug. 1939, Dyer 4094 (PRE, holo.). Euphorbia enopla var. viridis A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 964 (1941). Type: South Africa, Cape, 17 miles north of Jansenville towards Graaff-Reinet, Aug. 1939, Dyer 4008 (PRE, holo.). Unisexual spiny many-branched glabrous succulent shrub 0.15–1 (2) × 0.3–1  m, branching from base and above on thicker eventually nearly cylindrical stem with fibrous roots. Branches erect, cylindrical to slightly club-shaped and (5-) 7- to 10-angled, (10) 20–40 mm thick, simple to rebranching, partly covered with tubercles, smooth, green to dark green or grey-green becoming grey with age; tubercles obtusely conical, 4–8 mm long with transversely rectangular to hexagonal base, vertically fused into (5–6) 7–10 low obtuse angles, with broad triangular vertical grooves between angles and fine horizontal lines separating tubercles, with solitary simple spines in their axils; spines from persistent sterile short-shoots, straight, rigid, 6–40 × 1–2 mm, with several small alternating glabrous spathulate bracts to 1–2 × 1–2 mm along length, initially black to green or red, drying to brown then grey, glabrous to minutely pubescent; leafrudiments on tips of new tubercles towards apices of branches and stem, 1–2 × 1 mm, spreading, fleeting, lanceolate, acute, sessile. Synflorescences many per branch towards apex, solitary in axil of tubercle on unbranched red to green peduncle (6) 10–25 mm long (usually longer in males than females) with 1 terminal unisexual cyathium, minutely puberulous to glabrous, with 2–3 rectangular-lanceolate glabrous reddish

126

green bracts 1–3 mm long beneath cyathium; cyathia cupular, glabrous, 2.5–6 mm broad (usually males nearly twice as broad as and taller than females), with 5 lobes with finely toothed margins, green to red-green; glands 5, transversely elliptic, 1.5–2.5  mm broad, widely separated, erect or ascending, green, slightly depressed in centre, outer margins entire, stamens with pubescent pedicels, bracteoles filiform and slightly flattened and pubescent along margins; ovary nearly spherical, glabrous, ± sessile; styles 2.2–3 mm long, branched and recurved or spreading near apex or in upper half. Capsule 4–6  mm diam., almost spherical, glabrous, green to red, sessile.

Fig. 2.113.  Distribution of Euphorbia heptagona (© PVB).

Fig. 2.114.  Euphorbia heptagona, Klein Waterval, NE of Laingsburg, South Africa, 19 Jun. 1989 (© PVB).

2  Euphorbia subg. Athymalus

Distribution & Habitat Widely distributed over the arid parts of southern South Africa, E. heptagona is known from the southern coastal plain, on the Klein Karoo and on the southern portion of the Great Karoo. On the coastal plain it has been recorded from near Riversdale to Mossel Bay. On the Klein Karoo it is of wide occurrence between Montagu and Willowmore. On the Great Karoo it is known from near Merweville to GraaffReinet, Somerset East and Cradock and it is especially regularly encountered along the northern slopes of the mountains from Matjiesfontein to Willowmore and Klipplaat. East of Somerset East and Cradock it is replaced by E. pentagona.

2.1 Sect. Anthacanthae

127

As in species such as E. polygona, E. heptagona is tolerant of a wide variety of soils and is found on the lower, more arid slopes of mountains on ground derived from sandstones, shales and tillites. When on shales and tillites, there is often a sprinkling of white quartzitic rocks of the Witteberg series among the plants. On hot, stony, north-facing slopes it is found with a multitude of other succulents usually of equal height, but plants are also found (though less frequently) in flat areas. Within the mountains of the Cape Fold series (as in the Kouga Mountains, the Kamanassie Mountains, the Baviaanskloof, Groot Winterhoek etc.) it occurs in locally arid, steep, usually north-facing slopes in gorges and river-valleys, also with a wide variety of other succulents, most of which are of a similar size. It usually occurs in extensive and quite dense colonies and may be locally dominant. Diagnostic Features & Relationships Euphorbia heptagona is one of the most characteristic succulent species of Euphorbia of the southern Karoo, with its many-angled, but comparatively slender stems forming dense and very spiky clumps. Plants vary greatly in size, in some areas (such as around Montagu on the western Klein Karoo) they rarely exceed 30 cm in height, while in others (especially on steep slopes in the sandstone mountains) they regularly reach 1  m tall. Along with these differences in height, one will also notice that the taller plants are more diffusely and laxly branched, while the shorter ones are often dense, hemispherical and practically impenetrable. The young shoots are green, with seven to nine low, rounded angles and they bear plenty of strong spines in the axils of

Fig. 2.116.  Euphorbia heptagona, Gamka Poort, west of Prince Albert, South Africa, 27 Dec. 1987 (© PVB).

Fig. 2.115.  Euphorbia heptagona, Ouberg Pass, NE of Montagu, South Africa (© PVB).

128

the low and indistinct tubercles. These spines, which are generally longer in the male plants than the females, start off as a soft, pliable and somewhat shiny short-shoot, bearing a few, very small bracts, mainly towards their apices. The bracts rapidly drop off and the whole structure soon dries out to become rigid and hard. When young, the spines may be strikingly nearly black, deep red, green or yellow and lend their hue to the whole bush, somewhat obscuring the colour of the branches. With time most of them fade to grey but their variability in colour when young has given rise to named, green-spined varieties of all but one of the specific epithets that have now been reduced to synonymy under E. heptag-

2  Euphorbia subg. Athymalus

ona. Many populations have plants with a wide variety of colours of spines and sometimes the colour may even vary from bright red to pale red or green on a single shrub. Those spikes that are derived from sterile short-shoots become sharp-tipped and hard. These are interspersed with rosettes of persistent fertile short-shoots which, while eventually equally hard, are shorter and are not sharp-tipped. They are slightly different on the males and females, allowing one to distinguish the sexes in mature plants even without flowers: in the females the slender 3-tipped core of the capsule persists at the tip of the peduncle for a long time after the capsule has dehisced, while in the males this structure is absent.

Fig. 2.117.  Euphorbia heptagona, male cyathia, Ouberg Pass, NE of Montagu, South Africa (© PVB).

In habitat, E. heptagona will often be found in flower in the spring (August to November), when the tips of the branches begin to grow, giving rise to many new peduncles. Since the peduncles are fairly long, the tips of the branches are only rather diffusely adorned with cyathia (unlike in E. ferox and E. mammillaris, where they are quite dense at the tips of the branches). They emit a sweet-rotten odour that suggests over-ripe fruit and is similar to that emitted by E. polygona, perhaps explaining why these two species hybridise as regularly as they do. In E. heptagona the male cyathia are much larger than the females, though both give off the same odour quite strongly. In both sexes the cyathia are green and the glands are usually green as well, though the whole structure often quickly becomes suffused with red. The tips of the branches in female plants may bear many capsules after flowering and the seeds are usually released in November and December.

Closely related to such species as E. pseudoglobosa and E. susannae (Bruyns et  al. 2006), E. heptagona is also closely allied to E. cumulata and especially to E. pentagona. Euphorbia heptagona and E. pentagona look superficially similar but there are many differences between them, most of which are discussed under E. pentagona. Their cyathia differ in that those of E. heptagona are mostly much smaller (they are especially small often in the females) and the glands are also smaller. The differences between E. heptagona and the somewhat similar E. pillansii are discussed under the latter. Euphorbia heptagona is also rather similar to the very local E. pseudoglobosa subsp. nesemannii. These two differ mainly in the very much shorter and more sparsely distributed spines on the stems and by the presence in subsp. nesemannii of a generally stouter, at least partially subterranean central stem from which the branches arise. The differences

2.1 Sect. Anthacanthae

Fig. 2.118.  Euphorbia heptagona, female cyathia, Ouberg Pass, NE of Montagu, South Africa (© PVB).

129

Fig. 2.119.  Euphorbia heptagona, female plant with red peduncles, PVB 11230, north of Kamanassie Dam, east of Oudtshoorn, South Africa, 4 Oct. 2008 (© PVB).

between E. heptagona and E. cumulata, E. ferox and E. mammillaris are discussed under each of these species. Hybrids between E. heptagona and E. polygona are not infrequent and are discussed under E. polygona and in Sect. 6.1.2 in volume 2.

Fig. 2.120.  Euphorbia heptagona, branch with capsules, PVB 6356, Klipdrift, south of Riversdale, South Africa, 23 Oct. 2012 (© PVB).

History Linnaeus described Euphorbia heptagona from a figure of Boerhaave, which appeared in 1720, and he even used the name that Boerhaave had applied to these plants. Their habitat was at that time unknown (Linnaeus 1753). This species was, therefore, one of the earliest to become known in Europe and must have been encountered by early travellers on the coastal plain east of Swellendam or possibly on the Klein Karoo since, at this time, the interior further north was very little-known. Lemaire (1858) gave the species the new name Anthacantha desmetiana, which was therefore superfluous. White et al. (1941) mentioned repeatedly that the names atrispina, enopla and heptagona had been applied to plants that were hard to distinguish (‘little distinction...beyond the general habit of growth of each’: p. 632; ‘series of characters...None of which is reliable’: p. 648, etc.). They recognised the two extremes in the size of the plants and their shapes as the two ‘species’ E. heptagona (0.6–1.3 m tall) and E. atrispina (‘forming small cushion-shaped plants’ 0.75– 0.2 m tall), with a third, E. enopla (0.3–1 m tall), also recognised but hard to distinguish from the other two. As they admitted, there are many intermediates between these ‘species’ and many plants can only be assigned to one of these

130

2  Euphorbia subg. Athymalus

Fig. 2.121.  Euphorbia heptagona. A, B, side view of male cyathium (scale 2 mm, as for B–D). C, D, side view of female cyathium. E, anthers and bracteoles (scale 1 mm, as for F, G). F, G, female floret. Drawn from: A, C, F, PVB 6356, Klipdrift, south of Riversdale, South Africa. B, D, E, G, PVB 10891, Klipfontein, south of Riversdale, South Africa (© PVB).

names if it is certain where it came from. The considerable confusion present in the herbaria among these names bears out their contention that all should probably be placed under a single name. As White et  al. (1941: 632) pointed out as well, for each of the three ‘species’ a ‘green variety’ had been described, providing another indication of how similar they all were. Field-experience has shown that one is not dealing here with three distinct entities but with only one. Consequently, all the names apart from E. heptagona, are placed in synonymy. The species was illustrated in Phillips (1930). Euphorbia mammillaris L., Sp. Pl. 1: 451 (1753). Lectotype (Wijnands 1983): C. Commelijn, Praeludia Bot.: 59, t. 9 (1703). Euphorbia fimbriata Scopoli, Delic. Fl. Faun. Insubr. 3: 8 (1788). Lectotype (Bruyns 2012): Delic. Fl. Faun. Insubr. 3: 8, t. 4 (1788). Euphorbia enneagona Haw., Misc. Nat.: 184 (1803). Type: none cited. Euphorbia erosa Willd., Enum. Pl., Suppl.: 27 (1814). Treisia erosa (Willd.) Haw., Suppl. Pl. Succ. 66 (1819). Type: none cited. Euphorbia odontophylla Willd., Enum. Pl., Suppl.: 28 (1814). Type: none cited. Euphorbia scopoliana Steud., Nomencl. Bot., ed. 2, 1: 615 (1841), nom. superfl.

Euphorbia cereiformis var. echinata (Salm-Dyck) Boiss. in DC. Prodr. 15 (2): 88 (1862). Euphorbia echinata Salm Dyck, Hort. Dyck.: 342 (1834). Type: none cited. Euphorbia cereiformis K.Schum., Monatsschr. Kakteenk. 8: 55 (1898), nom. illegit., non L. (1753). Euphorbia mammillaris var. spinosior A.Berger, Monatsschr. Kakteenk. 12: 109 (1902). Type: South Africa, Cape, probably ex hort. F. Ledien (missing). Euphorbia submammillaris A.Berger, Monatsschr. Kakteenk. 12: 125 (1902). Lectotype (Bruyns 2012): South Africa, Cape, cultivated plant from Berlin Botanic Garden (K). Euphorbia platymammillaris Croizat, Cact. Succ. J. (US) 4: 333 (1933). Lectotype (designated here): Croizat (1932), Fig. 4. Unisexual spiny many-branched glabrous succulent shrub 0.15– 0.3  m tall, branching from near base (and sometimes above) of similar but usually shorter stem with cluster of fibrous roots at its base. Branches erect to creeping in bushes, clavate, simple to occasionally rebranched, 0.08—0.3 (1) m × 12–40 mm, covered with tubercles, bright green towards tips and brown below; tubercles low and conical, 2–4 mm long with transversely rectangular to hexagonal base, vertically fused into 7–10 (15) low obtuse angles, with broad shallow grooves between angles and fine horizontal lines separating tubercles, with solitary simple spines in their axils, with fine horizontal ridges running up

2.1 Sect. Anthacanthae

to base of leaf-rudiments; spines from persistent sterile short-shoots, usually clustered in whorl-like bands at intervals along branches, slightly ascending to straight or descending, rigid, 4–20 × 1–2 mm, with several minute alternating leaf-like bracts to 1 × 1  mm mainly near tips, initially green to pink, drying to brown to grey, glabrous; leaf-rudiments on tips of new tubercles towards apices of branches and stem, 2–6 × 1–2 mm, ascending, fleeting, subulate to ovate, acute, sessile. Synflorescences many per branch towards apex, each in axil of tubercle on short-shoot 3–5 mm long in females and 8–10 mm long in males, with 1 terminal unisexual cyathium, glabrous, with 2–5 broadly obovate spreading finely and sparsely ciliate pale green bracts (with purple-red toothed margins) 1.5–3 mm long in whorl just beneath cyathium with some shorter bracts below these; cyathia cupular, glabrous, 3.5–5 mm broad (± 3 mm long below insertion of glands), sometimes males twice as broad as and taller than female, with 5 white-pubescent red

131

lobes with deeply divided margins, pale green; glands 5, transversely elliptic-cuneate to almost circular in males, 1.3–1.5  mm broad, contiguous to slightly separate, deep green to yellow-green, convex and slightly pitted to smooth and concave above, outer margins entire but slightly crenulate; stamens with pedicels with few to many hairs near apex, bracteoles filiform and pubescent in upper half; ovary ellipsoidal, glabrous, raised on very short pedicel < 0.3 mm long; styles ± 3 mm long, branching near apex only. Capsule 4–7  mm diam., almost spherical, glabrous, grey-green to red, sessile. Distribution & Habitat Euphorbia mammillaris is found mainly on the southern coastal plain of South Africa. Here it is scantily recorded from Heidelberg to Albertinia, becoming much more plentiful in the deep valley of the Gouritz River and on stony, dry slopes to around Mossel Bay. Further east it is well-known in

Fig. 2.122.  Distribution of Euphorbia mammillaris (© PVB).

the area between the Gamtoos River and the Sundays River valley, from where it continues to between Grahamstown and Bathurst. Euphorbia mammillaris usually occurs on small rocky outcrops in patches of dense scrub, where the bush is a little lower and less dense than in the surrounding areas. However, in the thickets of the so-called ‘Addo Bush’ plants will sometimes be found even in the denser patches of bush and here the branches may creep on the ground or clamber in some of the surrounding shrubs.

Diagnostic Features & Relationships Euphorbia mammillaris forms clumps of branches developing from a short, central stem that is hard to distinguish from the branches other than by its somewhat lesser length. These clumps may reach 30 cm in diameter if dense, but when the stems sprawl and become elongated, the plants are often much larger, though more diffuse. In plants growing fully exposed the branches are often quite short, but, even then, they are rarely more than 25 mm thick. Towards their tips they are a bright green colour but lower down they soon

132

2  Euphorbia subg. Athymalus

Fig. 2.123.  Euphorbia mammillaris, forming untidy shrub on stony slopes with many succulents, PVB 471, Mossel Bay, South Africa, 27 May 2005 (© PVB).

Fig. 2.124.  Euphorbia mammillaris, with Portulacaria afra, Hankey, South Africa, 7 Aug. 2011 (© PVB).

become covered with a grey-brown, corky bark. Slender short-shoots which soon become sharp, but slender spines abound towards the tips of the branches and are produced in ring-like patches interspersed with patches where they are absent. When they are young, the short-shoots are faintly suffused with red and some very tiny bracts are present near their tips for a short while. After drying out, the spines persist but gradually wear off and only vestiges of them remain in the lower halves of the branches. In some plants the spines are relatively few and weak. Cyathia are produced between May and October, in clusters on very short peduncles at the tips of the branches. Each

peduncle bears six to eight bracts which increase in size towards the tip, with the largest two or three spreading and quite conspicuous beneath the cyathium. The fairly long and narrow cyathia are larger in the males than the females (sometimes nearly twice as broad) and the males may have almost circular, slightly separated glands while the females have more wedge-shaped, contiguous glands. A closer examination reveals that the cyathial lobes are red, with a coarse white pubescence. The cyathia emit a slightly fruity odour. Some remains of the peduncles that bore fruit may persist so that one can often tell which plants are female and which are male long after flowering and fruiting is over.

2.1 Sect. Anthacanthae

133

Fig. 2.125.  Euphorbia mammillaris, dense shrub about 0.3 m tall with many other succulents, Hankey, South Africa, 7 Aug. 2011 (© PVB).

Fig. 2.126.  Euphorbia mammillaris, male cyathia, PVB 471, Mossel Bay, South Africa, 20 Aug. 2015 (© PVB).

Fig. 2.127.  Euphorbia mammillaris, male cyathia, Hankey, South Africa, 10 Aug. 2011 (© PVB).

Euphorbia mammillaris is similar to species such as E. ferox and E. heptagona and in all of these (and several others) the spines are derived from persistent sterile shortshoots, while the fertile short-shoots are much shorter and not at all spiny and most of them are shed after flowering and fruiting. The differences between E. mammillaris and E. ferox are discussed under E. ferox. Plants of E. mammillaris always have fairly slender branches and bear a considerable resemblance to E. heptagona. They differ from E. heptagona by the manner in which the tubercles are separated from one another by a fine horizontal line (entirely absent in E. heptagona) and the fine ridge running horizontally away from

each leaf-rudiment along the tubercle, by their often longer and broader leaf-rudiments and by their paler, more slender, fewer and shorter spines, which are usually very brightly coloured in E. heptagona when they are young and are far more robust. The bracts on the peduncles are smaller and less prominent in E. heptagona, as they are generally pressed to the surface of the shaft or to the cyathium, whereas they spread outwards from the peduncle in E. mammillaris and are quite conspicuous just beneath it. The cyathia are generally smaller in E. heptagona than in E. mammillaris and male cyathia in the latter often have almost circular glands, which are never found in E. heptagona.

134

2  Euphorbia subg. Athymalus

Fig. 2.128.  Euphorbia mammillaris, female cyathia, PVB 471, Mossel Bay, South Africa, 20 Aug. 2015 (© PVB).

Fig. 2.129.  Euphorbia mammillaris, capsules, Hankey, South Africa, 10 Aug. 2011 (© PVB).

History Euphorbia mammillaris was in cultivation in Amsterdam between 1686 and 1702, when it was illustrated by Jan Moninckx. This plate was included as the 47th figure of the Moninckx Atlas Volume 5 (Wijnands 1983) and was published by Caspar Commelijn in 1703. Although it does not show a plant in flower, the distinctively delineated tubercles and scanty armature of spines are characteristic of what we know today as E. mammillaris. The source of this material is unknown. Euphorbia mammillaris does not appear to have been common in cultivation then and it seems that Linnaeus knew it only from this

illustration. He believed that it originated in ‘Aethiopia’ (Linnaeus 1753). N.E. Brown (1915) and White et al. (1941) devoted some space to an account of E. cereiformis L., under which they included material closely resembling the present concept of E. mammillaris. This followed the concept of Boissier (1862). However, this interpretation was contested by Croizat (1934), who showed that E. cereiformis L. was synonymous with the Moroccan E. officinarum of subg. Euphorbia. With the typification of E. cereiformis by Wijnands (1983), this name became a synonym of E. officinarum L., a species that was known in Europe before 1570.

2.1 Sect. Anthacanthae

135

Fig. 2.130.  Euphorbia mammillaris. A, B, side view of male cyathium (scale 2 mm, as for B, C). C, D, side view of female cyathium. E, F, anthers and bracteoles (scale 1 mm, as for D, F, G). G, female floret. Drawn from: A, C, E, PVB 471, Mossel Bay, South Africa. B, F, Hankey, South Africa. D, G, PVB 6702, lower Gouritz River, South Africa (© PVB).

White et  al. (1941) associated plants from the eastern Klein Karoo (following Berger 1906: 91) with the name E. mammillaris. In this respect they deviated from previous treatments, especially from that by Brown (1915: 347), where the name E. mammillaris was associated with plants from the area around Port-Elizabeth and Uitenhage. Here I follow N.E. Brown’s and Dyer’s (1931: 91) concept for E. mammillaris and treat the plants from the eastern Klein Karoo as a subspecies of E. ferox, to which it is florally most similar. Euphorbia meloformis Aiton, Hort. Kew., ed. 1, 2: 135 (1789). Lectotype (Bruyns 2012): South Africa, Cape, Zwartkops, Masson, introduced 1774 (coloured plate by F. Masson, BM). Euphorbia pomiformis Thunb., Prodr. Fl. Cap. 2: 86 (1800). Euphorbia meloformis var. pomiformis (Thunb.) Marloth, S.  African Gard. 18: 45 (1928). Type: South Africa, Zwartkops, Thunberg (missing). Euphorbia falsa N.E.Br., Fl. Cap. 5 (2): 586 (1925). Euphorbia meloformis subsp. meloformis f. falsa (N.E.Br.) J.G.Marx, Euphorbiaceae Study Group Bull. 12: 32 (1999). Euphorbia infausta N.E.Br., Fl. Cap. 5 (2): 358 (1915), nom. illegit., non N.E.Br. (1912). Lectotype (Bruyns 2012): South Africa, Cape, sheet 332, specimen annotated ‘dead plant-split-1810’ by Haworth (OXF).

Euphorbia pyriformis N.E.Br., Fl. Cap. 5 (2): 359 (1915). Type: cultivated plant at Kew of unknown origin, pressed by N.E. Brown 14 Jan. 1913 (K, holo.). Euphorbia valida N.E.Br., Fl. Cap. 5 (2): 356 (1915). Euphorbia meloformis subsp. valida (N.E.Br.) G.D.Rowley, Euphorbiaceae Study Group Bull. 11: 97 (1998). Type: South Africa, Cape, Jansenville div., near Waterford, I.L. Drège (K, holo.). Euphorbia meloformis var. prolifera Frick, Cact. Succ. J. (US) 6: 74 (1934). Type: Cultivated material from seed imported from South Africa, A.C.S. 5-112-006 (missing). Euphorbia meloformis subsp. meloformis f. magna R.A.Dyer ex J.G.Marx, Euphorbiaceae Study Group Bull. 12: 13 (1999). Type: South Africa, Cape, Kwa Ncwane, near Peddie, 18 Mar. 1999, Marx 550 (GRA, holo.). Unisexual (sometimes bisexual) spineless glabrous dwarf succulent consisting of single swollen stem (20) 30–300 (400) × 25–140 (190) mm (rarely branched from base or around chewed-off apex) with somewhat depressed apex, rapidly tapering at base into slender tap-root from which fibrous roots spread out, often with simple or branched remains of persistent woody peduncles towards apex of stem. Stem erect, hemispherical when young becoming cylindrical with age (tapering to base and apex), 7- to

136

12-angled, partly covered with tubercles, smooth, green (sometimes shiny) usually with many darker green or purple horizontal bands; tubercles low and conical, 3–5 mm long and vertically fused into 7–10 prominent obtuse to acute angles and becoming more prominent and more acute towards apex, with broad and deep triangular vertical grooves between angles and often with horizontal lines separating tubercles; leaf-rudiments on tips of new tubercles towards apex of stem, 1–5 × 0.5–1 mm, erect, fleeting, linear and slightly channelled above, acute, sessile. Synflorescences many towards apex of stem, usually solitary in axil of each tubercle, each on unbranched to branched finely pubescent grey-green peduncle 2–30 (–80) × 2–5 mm (usually longer in males than females), with 1 terminal unisexual cyathium and others developing on 2 (–4) branches arising in axils of bracts just below it, puberulous, with 2–6 ovate finely and sparsely ciliate reddish green glaucous bracts 1–3 mm long (longest just beneath cyathium); cyathia cupular, puberulous, 2.5–6 mm broad (± twice as broad in male as in female), with 5 puberulous lobes with toothed margins, grey-green sometimes faintly suffused with red; glands 5, transversely elliptic to almost semicircular, 1–1.5  mm broad, ascending and slightly folded above, slightly separated, pale green to suffused or dotted with red along margins, slightly pitted on upper surface, stamens with pubescent pedicels, bracteoles filiform and sparsely pubescent; ovary ellipsoidal, glabrous, raised less than 0.25  mm on very short pubescent pedicel; styles ±  1  mm

Fig. 2.131.  Distribution of Euphorbia meloformis (© PVB).

2  Euphorbia subg. Athymalus

long, branched to below middle and spreading. Capsule 6–7  mm diam., obtusely 3-angled, glabrous to minutely puberulous, grey-green to purple sometimes with pale mark down sutures, sessile. Distribution & Habitat Euphorbia meloformis occurs in the Eastern Cape, from near Waterford and near Pearston eastwards to Adelaide and as far as about 15 km north-east of Peddie. Along the valley of the Fish River, it has been recorded as far north as Halesowen, near Cradock, though it is more common further to the south, especially near Grahamstown. It was also common in the dry flats between Port Elizabeth and Uitenhage, though the creation of new industrial areas and townships there has greatly reduced these populations. Euphorbia meloformis is usually found in flat or gently sloping areas in stony ground derived from shales or alluvium with scanty to quite deep, hard, loamy soil. In many places it grows among short grasses and other small succulents, but also occasionally among karroid bushes such as Pteronia incana and Ruschia spinosa. Large, older individuals may stand in the open, but most of the younger plants grow where they are sheltered under bushes or small clumps of grass. Diagnostic Features & Relationships Most plants of E. meloformis consist of a single, thick, usually hemispherical stem of very variable thickness

2.1 Sect. Anthacanthae

137

Fig. 2.132.  Euphorbia meloformis, three plants close together among short dry grasses) largest a female), all without persistent peduncles, PVB 10623, Coega, Port Elizabeth, South Africa, 4 Dec. 2006 (© PVB).

(commonly between 3 and 10  cm thick), which almost always remains closely pressed to the ground or even slightly sunken into it. Beneath, emanating from its abruptly narrowed base and anchoring it firmly into the soil, is a slender tap-root from which a small system of

fibrous roots develops. With age, the upper part of the stem elongates slowly so that the whole plant may eventually become almost cylindrical, though this is quite rare in most areas (though commoner in the driest places where it occurs, as around Waterford and on the Dikkop Flats along

Fig. 2.133.  Euphorbia meloformis, female plant, PVB 10623, Coega, Port Elizabeth, South Africa, 4 Dec. 2006 (© PVB).

Fig. 2.134.  Euphorbia meloformis, male plant with short peduncles, with a small plant of Ceropegia zeyheri behind it, PVB 10623, Coega, Port Elizabeth, South Africa, 4 Dec. 2006 (© PVB).

the Fish River). Running up the stem are seven or eight (sometimes more) prominent, acute angles that are straight or may be slightly spiralled towards the apex of the plant. In young plants the whole of the above-ground stem is green with an attractive banding of darker green or purple (plain rather shiny green north-east of Peddie), with the bands running from the tubercles horizontally into the grooves between the angles. On older plants the green surface is gradually replaced by a grey bark, so that the lower half of the stem often lacks the attractive mottling of the

younger part. In most plants the apex of the stem is slightly depressed and so the slender leaf-rudiments (which are only present briefly around the immediate apex) remain partly hidden. As the tubercles are pushed away from the centre, the peduncles begin to develop in their axils. When the first cyathium matures, the peduncle may be no more than 5 mm long. It branches from the axils of bracts just beneath the terminal cyathium, with usually two or three, rarely four such branches appearing from beneath a given cyathium.

138

2  Euphorbia subg. Athymalus

Fig. 2.135.  Euphorbia meloformis, female plant among stones and short grasses (small Asparagus on the left), PVB 10966, NE of Carlisle Bridge, South Africa, 11 Jan. 2008.

Each branch terminates in a cyathium and this process may be repeated many times to form, in some cases a very extensively and mostly dichotomously branched structure of very variable length, where a cyathium had been present in each fork. After bearing the cyathia, the peduncles usually dry out and persist as hard and woody, but not at all spiny outgrowths along the stem and gradually wear off towards the base of the plant. They may be somewhat twisted together over the top of the plant (especially if conditions are dry) and this may help to protect and hide the apex of the plant by breaking up its outline and obscuring its bright colour. Plants in the coastal areas near Port Elizabeth and in the moister grasslands north-east of Peddie often lack these persistent peduncles but further inland they are present on most specimens. The longest and stoutest peduncles are found among plants

in the drier areas amongst karroid scrub, especially south of Somerset East. The cyathia are produced mainly in spring or early summer (September to November, but sometimes as late as April). On a given specimen they are mostly unisexual (so that individuals are either ‘male’ or ‘female’), though occasionally the cyathia on side-branches in a male plant develop styles to become functionally female too. The male cyathia are about twice as broad as the females. Generally, the slightly spreading and faintly pitted cyathial glands are inconspicuous and the anthers have very short filaments. The female florets are unusual in having a short but pubescent pedicel and the hairs towards the top of the pedicel form a ring around the base of the ovary. A slightly musty-fruity odour is given off by the cyathia.

Fig. 2.136.  Euphorbia meloformis, three plants (largest male) among stones and short bushes, plants partly hidden by persistent peduncles, PVB 11658, SW of Somerset East, South Africa, 9 Jan. 2010 (© PVB).

Fig. 2.137.  Euphorbia meloformis, larger male plant ± 12 cm tall with much longer persistent peduncles and with seedling next to it, PVB 11658, SW of Somerset East, South Africa, 9 Jan. 2010 (© PVB).

2.1 Sect. Anthacanthae

With its popularity in cultivation, it is not surprising that several names exist for this species and, of these, E. valida is most often encountered. Marloth (1928) investigated the stated differences between E. meloformis and E. valida and found that plants from near Somerset East exhibited a particularly wide range in the length and thickness of the stem and also in the extent to which the peduncles branched, the main characters on which these two species had been separated. He concluded that he could find both ‘species’ among these plants (from near Somerset East) and also many intermediates, so he could not partition the individuals into two separate species. He also mentioned an unusually tall plant found in E. obesa too and believed that these were ‘sports’ that occurred occasionally in both species. Dyer (1931, 1935b) and White et  al. (1941: 578) commented that their respective distributions and the distinctions between them ‘seem to justify their continued maintenance as separate species’. Nevertheless, they mentioned the existence of plants

139

north of Grahamstown which were intermediate between the two ‘species’ and so contradicted their recognition of two species. Marx (1988, 1992, 1999a) discussed the two ‘species’ and mentioned (largely following White et al. 1941: 81) that they differed in that E. valida formed larger plants, had regular and robust, persistent peduncles and possessed a slightly shallower depression in the apex of the stem. Nevertheless, he admitted that the younger plants were more or less impossible to place under the two ‘species’ unless it was known where they came from. Much the same was stated by Phillips (1934), who declared that the difference in size and shape is almost entirely due to age. In my experience, plants of E. meloformis are very variable in habitat and the evidence produced by Marx (1988, 1992, 1993, 1999a) for separating the larger growth-forms as E. valida is unconvincing. I consider that regarding them as a single species reflects their great variability in some areas, as concluded by Marloth (1928).

Fig. 2.138.  Euphorbia meloformis, female plant ± 10 cm diam., sunken into mosses among bushes and stones, PVB 12255, NE of Peddie, South Africa, 5 Dec. 2012 (© PVB).

It is interesting that in E. meloformis the largest and oldest plants are found in the driest areas where it occurs, probably reflecting the fact that these plants grow the slowest and are then the most resilient. It may also reflect a greater prevalence of pests and diseases in the wetter areas to which the species is not fully resistant and where such venerable plants more rarely manage to survive. History Euphorbia meloformis was first introduced to England in 1774 by Francis Masson. It is most likely that these plants were collected by Masson and Thunberg around the middle of December 1773, when they reached the lower part of the

Zwartkops River, near the present-day Uitenhage. This was the furthest east that they went during their joint expedition to the southern and eastern parts of the Cape Colony. Material from this area was given the name Euphorbia pomiformis by Thunberg. An illustration of E. meloformis was published by François le Vaillant (1795, Vol. 2: t. 11), but Dyer (1949: 14) pointed out that this was a completion of an unfinished figure of Robert Gordon’s among Paterson’s collection of paintings (as in the cases of E. stellata and ‘E. cucumerina’). Euphorbia meloformis was also known to Ecklon & Zeyher, who collected it in the same area, probably on or before 1831 and there is little doubt that it would have been encountered by James Bowie as well, though this does not seem to be

140

2  Euphorbia subg. Athymalus

Fig. 2.139.  Euphorbia meloformis. A, side view of male cyathium (scale 2 mm, as for B). B, side view of female cyathium. C, anthers and bracteoles (scale 1 mm, as for D). D, female floret. Drawn from: PVB 10966, NE of Carlisle Bridge, South Africa (© PVB).

recorded. It seems to have fascinated botanists and horticulturalists ever since its discovery and Lemaire (1855) even created a special group ‘Echincactoides’ for it, because of its striking resemblance to some members of Echinocactus. Euphorbia obesa Hook.f., Bot. Mag. 129: t. 7888 (1903). Type: South Africa, near Graaff-Reinet, Kendrew, 2000’, Mar. 1897, MacOwan 3183 (K, holo.). Unisexual spineless glabrous dwarf succulent consisting of single swollen stem 30–200 × 40–100 (–250) mm often with somewhat depressed apex, rapidly tapering below ground into slender tap-root from which fibrous roots spread out. Stem erect, hemispherical when young becoming cylindrical with age, 7- to 10-angled, partly covered with tubercles, smooth, grey-green to brown with many dull purple horizontal bands; tubercles very low and obtusely conical, 3–5  mm long and vertically fused into 7–10 low obtuse angles becoming more prominent towards apex of stem, with broad and very shallow triangular vertical grooves between angles and without horizontal lines separating tubercles; leaf-rudiments on tips of new tubercles towards

apex of stem, 1–2 × ±  1 mm, erect, fleeting, oblong and slightly channelled above, acute, sessile. Synflorescences many towards apex of stem, 1–5  in axil of each tubercle, each on unbranched finely pubescent grey-green peduncle 2–12 (rarely to 75) mm long (usually longer in males than females and elongating slightly with maturing capsules) with 1 terminal unisexual cyathium, with 2–6 ovate to ovatelinear finely and sparsely ciliate glaucous to reddish green bracts 1–3  mm long (longest just beneath cyathium and these sometimes with 2–3 further shorter peduncles arising in their axils and also terminated by a solitary cyathium); cyathia cupular, puberulous, 2.5–6  mm broad (broader in male than female), with 5 puberulous greyish lobes with toothed margins, grey-green sometimes suffused with red; glands 5, transversely elliptic to almost circular, 1–1.5 mm broad, ascending and slightly folded above, widely separated, dark green to red, slightly pitted on upper surface, stamens with pubescent pedicels, bracteole s filiform and sparsely pubescent; ovary globose, glabrous, sessile; styles ±  2  mm long, branched to well below middle and widely spreading. Capsule 6–7 mm diam., very obtusely 3-angled and almost spherical, glabrous, grey-green, sessile.

2.1 Sect. Anthacanthae

141

Fig. 2.140.  Distribution of Euphorbia obesa (▲ = ssp. obesa; ● = ssp. symmetrica) (© PVB).

Two subspecies are recognised and may be separated as follows: 1. Cyathia in groups of up to five (in males) and three (in females) on each peduncle….....................................subsp. obesa 1. Cyathia solitary on each peduncle……..................................................................................................subsp. symmetrica

Euphorbia obesa subsp. obesa Peduncles solitary in axil of each tubercle. Male and female cyathia often more than one on each peduncle (1–5 for males and 1–3 for females)

Fig. 2.141.  Euphorbia obesa ssp. obesa, male plant next to small Monsonia, near Kendrew, south of GraaffReinet, South Africa, 3 Apr. 2012 (© PVB).

Distribution & Habitat Euphorbia obesa subsp. obesa occurs on the Great Karoo, a little to the south-east of Graaff-Reinet, near the banks of the Sundays River between Kendrew and Charlwood.

142

Generally, Euphorbia obesa occurs among stones and low bushes in firm loam on low ridges or in stony flat areas around the bases of these ridges. Plants of E. obesa generally grow at the base of or inside a small bush or next to a stone, where they are slightly protected from the sun and from being trampled by stock. Along the Sundays River, the low hills where subsp. obesa grows appear to be of alluvial origin

2  Euphorbia subg. Athymalus

and the stones lying around are all distinctly rounded and pebble-like. Here the small bushes are mainly Monsonia camdeboensis and occasional larger shrubs of Rhigozum obovatum, with scattered clumps of Aloe claviflora, Astraloba and Euphorbia ferox and plenty of Senecio radicans beneath the bushlets.

Fig. 2.142.  Euphorbia obesa ssp. obesa, female plant (with unbranched peduncles), near Kendrew, south of GraaffReinet, South Africa, 3 Apr. 2012 (© PVB).

Diagnostic Features & Relationships Euphorbia obesa subsp. obesa is one of the most distinctive and remarkable-looking species of succulent Euphorbia. It has been much collected and widely cultivated for its unusual shape. Fig. 2.143.  Euphorbia obesa ssp. obesa, near Kendrew, south of Graaff-Reinet, South Africa, 3 Apr. 2012 (© PVB).

The plant consists of a squat, often almost hemispherical, solitary stem pressed close to the surface of the ground and mostly between 30 and 80 mm tall and 50 to 70 mm broad, though very large plants to 250 mm in diameter have been recorded (Marx 1999a). Just beneath the surface of

2.1 Sect. Anthacanthae

the ground the plant-body tapers off rapidly into a slender tap-root which bears fibrous roots. Although the stem is initially hemispherical, it may become cylindrical with age and in habitat these longer plants usually lean to one side. Running along the stem are between seven and ten broad, often flattened or low angles separated by shallow grooves that usually become more prominent towards the plant’s apex. While the background colour of the stem is grey-

143

green to brown, the angles are also dully marked with parallel, horizontal, purple bands which are often slightly and broadly V-shaped. The stems are especially neat and attractive because of their complete lack of persistent sterile or fertile short-shoots. Minute leaf-rudiments are produced in the apex of the stem during growing periods, but these generally shrivel up and fall off quickly.

Fig. 2.144.  Euphorbia obesa ssp. obesa, three young plants, near Kendrew, south of Graaff-Reinet, South Africa, 3 Apr. 2012 (© PVB).

Flowering in subsp. obesa usually takes place between October and January, but flushes of cymes may arise as late as April. The inconspicuous synflorescences arise in a small, dense cluster close to the apex of the stem, each in the axil of a former leaf-rudiment. Often, they are apparently scentless, but the males sometimes give off a slight fruity odour and as supplementary attractants they secrete significant amounts of nectar from the glands. Each peduncle is generally very short, though it is longer in the males than in the females and the male cyathia are broader than the female. In the males the peduncle may also branch in the axils of the bracts around the terminal cyathium. The glands are insignificant and usually ascending, pressed to the sides of the cyathium and the whole structure mostly matches the colour of the stem. When the peduncles fall off (particularly the females) they leave a small depression and these shallow holes form neat lines along the middle of the angles. Once the capsule begins to develop, the peduncle may elongate from 2–5  mm long to around 10  mm long. However, the capsule remains firmly pressed to the top of the cyathium and the pedicel subtending it does not elongate at all.

Euphorbia obesa shares several features with the closely allied E. meloformis, especially with respect to the plant being reduced to a single swollen stem. Euphorbia meloformis generally has at least some persistent short-shoots around the apex of the stem (often a veritable ‘forest’ of them is present on mature plants), while such persistent structures are absent in E. obesa. The stem is also much more boldly mottled in E. meloformis than in E. obesa, with more striking, often shiny, deep green or purple lines on a paler background and, furthermore, the angles are generally far more prominent in E. meloformis. While these two species are not known to occur together, E. obesa usually grows together with the closely related E. ferox and E. polygona and occasional hybrids between E. ferox and E. obesa may be encountered. Apart from its great attractiveness to growers, Euphorbia obesa is fascinating as an example of a greatly reduced growth-form, found in other families and even in unrelated examples within Euphorbia. In the Cactaceae, species such as Astrophytum asterias and some forms of Lophophora williamsii resemble E. obesa very closely. Within Euphorbia, E.

144

2  Euphorbia subg. Athymalus

Fig. 2.145.  Euphorbia obesa ssp. obesa, male plant with branched peduncle, PVB 12243, near Kendrew, south of Graaff-Reinet, South Africa, 15 Nov. 2017 (© PVB).

Fig. 2.146.  Euphorbia obesa ssp. obesa, female plant with branched peduncle, PVB 12243, near Kendrew, south of Graaff-Reinet, South Africa, 15 Nov. 2017 (© PVB).

turbiniformis, from Somalia, has a very similar, highly reduced growth-form. The stem in E. turbiniformis is much smaller than in E. obesa but the whole plant is also reduced to a single, almost hemispherical body. However, E.

turbiniformis is not closely allied to E. obesa but is rather a much modified and greatly reduced member of subg. Euphorbia.

Fig. 2.147.  Euphorbia obesa ssp. obesa. A, side view of male cyathium (scale 2 mm). B, side view of female cyathium (scale 1 mm, as for C, D). C, anthers, bracteoles and sterile female in dissected male cyathium. D, female floret. Drawn from: PVB, near Kendrew, south of Graaff-Reinet, South Africa (© PVB).

2.1 Sect. Anthacanthae

History Subsp. obesa was first collected ‘in karroid plains’ at Kendrew near Graaff-Reinet by Peter MacOwan, during a short collecting trip to the Eastern Cape in February and March in 1897. Although Gunn and Codd (1981) mentioned that this trip was mainly to revisit old collecting sites from the period 1862–81 (when MacOwan worked in Grahamstown and in Somerset East), this species does not appear to have been collected before. In March 1897, MacOwan sent a sheet of pressed plants from the Herbarium MacOwanianum (under his number 3183) and three live female plants (according to a photograph on the type sheet at Kew) to the Royal Botanic Gardens at Kew under the name Euphorbia meloformis. In a letter dated 31 March 1897 to the Director of the Royal Botanic Gardens, Kew, he mentioned ‘I am sending you two Euphorbiae...The other, an old friend, is E. meloformis and was got at Kendrew, 20 miles S of Graaf Reinet (sic), alt. 2015.’ At least one of these plants flowered for the first time in July, 1899 and it was pollinated from a specimen of E. meloformis. At the time, a painting of it was made and shortly thereafter a description was drawn up. Soon after this, the plants disappeared from the collection (Hooker 1903) and may have been stolen (as was the fate later of the type-plant of Ceropegia furta as well, Bally (1963)). The type collection was cited in Bruyns (2012) as ‘3153’, following the information given on the JSTOR website, but

Fig. 2.148.  Euphorbia obesa ssp. symmetrica, male plant, PVB 3131a, near Rietbron, South Africa, 24 Oct. 2009 (© PVB).

145

N.E. Brown (1915) gave it as ‘3183’ and, although the ‘8’ is difficult to make out on the type sheet, the number that Brown gave is correct. MacOwan’s own collecting books (at PRE) confirm this since 3153 was recorded as ‘Aloe Greenii’. However, there is remarkably little information in his collecting books about MacOwan 3183 other than ‘In planis carroideis pr. Graaf Reinet Mart. 97’ (and ‘Graaf Reinet’ is even crossed out and replaced with ‘Clanwilliam’). Clearly, MacOwan did not recognise what an exciting discovery he had made. Because of its unusual and fascinating shape, subsp. obesa has been collected in appallingly large numbers since its discovery (as remarked on already by Marloth (1928)). With improved methods of propagation in Europe, South Africa and America, the pressure from collectors has gradually eased and it is possible to record that it still occurs in reasonable numbers near Kendrew so that, if it remains protected as at present, its survival in habitat is assured. Euphorbia obesa subsp. symmetrica (A.C.White, R.A.Dyer & B.Sloane) G.D.Rowley, Euphorbiaceae Study Group Bull. 11: 97 (1998). Euphorbia symmetrica A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 964 (1941). Lectotype (Bruyns 2012): South Africa, Cape, 19 miles NW of Willowmore on road to Rietbron, Aug. 1939, Dyer 4038 (PRE; K, iso.).

146

Peduncles solitary or up to 5 in axil of each tubercle. Male and female cyathia usually solitary on each peduncle. Distribution & Habitat Euphorbia obesa subsp. symmetrica grows in the dry country between Rietbron and Willowmore. Here it occurs on metamorphised shales in short vegetation that is typical of

2  Euphorbia subg. Athymalus

this region and dominated by scattered shrubs of Rhigozum obovatum. Plants may be found sheltering inside small bushes or, more rarely, in the open among stones. Euphorbia is well represented here, with E. decepta, E. ferox, E. juttae, E. mauritanica, E. polygona and E. stellispina all occurring close by.

Fig. 2.149.  Euphorbia obesa ssp. symmetrica, female plant with two small seedlings next to her, PVB 3131a, near Rietbron, South Africa, 24 Oct. 2009 (© PVB).

Diagnostic Features & Relationships The two subspecies of Euphorbia obesa are very similar indeed. Although various differences have been given between them (White et  al. 1941: 563), subsp. obesa and

Fig. 2.150.  Euphorbia obesa ssp. symmetrica, male plant, PVB 3131a, near Rietbron, South Africa, 24 Oct. 2009 (© PVB).

subsp. symmetrica differ mainly in that older plants of subsp. symmetrica may produce three to five peduncles from the axil of each tubercle, while those of subsp. obesa only produce a single peduncle from the axil of each tuber-

2.1 Sect. Anthacanthae

cle. It has also been found that several cyathia may develop on each peduncle in subsp. obesa while they are mostly solitary on each peduncle in subsp. symmetrica. History The first record of what was described as Euphorbia symmetrica was made by R.A.  Dyer in August 1939. Apart from the rather circuitous statements ‘The writer was intro-

147

duced to the plant in 1939’ and ‘the few collectors who knew of their occurrence in the area’ (Dyer 1940b), it has not been told how Dyer came to hear of these plants and make this first record. It is very likely that he heard of them through collectors such as Herre, the Lückhoff family and Marloth. Such distinctive plants would certainly also have been well-known to the owners of the farms on which they occurred.

Fig. 2.151.  Euphorbia obesa ssp. symmetrica, female plant, PVB 3131a, near Rietbron, South Africa, 24 Oct. 2009 (© PVB).

Fig. 2.152.  Euphorbia obesa ssp. symmetrica, female plant in flower and with capsules (one peduncle branched), PVB 3131a, near Rietbron, South Africa, 15 Dec. 2011 (© PVB).

Euphorbia pentagona Haw., Philos. Mag. Ann. Chem. 3: 187 (1828). Type: South Africa, Cape of Good Hope, introduced before 1825, Bowie (missing). Neotype

(Bruyns 2012): South Africa, Cape, Kei River Mouth, Flanagan 2344 (BOL; GRA, PRE, iso.).

148

Unisexual spiny many-branched glabrous succulent shrub 1–3 × 1 (5) m, branching from base and above on thicker eventually nearly cylindrical stem (reaching 150  mm thick around base) with fibrous roots. Branches erect, cylindrical and (4) 5- to 6 (7)-angled, 0.2–1 m × (10) 20–40 mm, simple to occasionally rebranched, partly covered with tubercles, smooth, pale to deep green becoming grey with age; tubercles acutely conical, 4–8 mm long with transversely rectangular to hexagonal base, vertically fused into (4) 5–6 (7) low acute angles, with broad shallow vertical grooves (with impressed line in centre) between angles and fine horizontal lines separating tubercles, with solitary simple spines in their axils; spines from persistent sterile short-shoots, straight to slightly ascending, rigid, 6–20 × 1–2 mm, with several small alternating glabrous spathulate bracts to 2 × 2 mm near apex, initially green to red and glabrous, drying to brown then grey; leaf-rudiments on tips of new tubercles towards apices of branches and stem, (2) 3–10 × 1–1.5 mm, ascending to spreading, fleeting, linear to linear-lanceolate, channelled above, acute, sessile. Synflorescences many per branch towards apex, each in axil of tubercle on solitary usually branched green peduncle 15–30 mm long (usually longer in males than females) with 1 terminal unisexual cyathium and

2  Euphorbia subg. Athymalus

2–3 unisexual cyathia on branches arising in axils of bracts just below tip of peduncle, minutely puberulous to glabrous, with 2–3 alternate bracts 1–2 mm long lower down and 2–3 bracts 3–5 mm long beneath each cyathium all rectangularlanceolate; cyathia cupular, minutely pubescent to glabrous, 3–4.5 mm broad, with 5 tomentose lobes with toothed margins, green to dull purple; glands 5, transversely elliptic, 2 mm broad, ascending-erect and nearly contiguous, green, slightly depressed in centre, outer margins entire or slightly crenulate, stamens with pubescent pedicels, bracteoles filiform and pubescent; ovary globose, glabrous, raised on very short pedicel < 0.2 mm long; styles 2.5–4 mm long, branched and recurved only near apex. Capsule 4–6  mm diam., obtusely 3-angled, glabrous, grey-green, sessile. Distribution & Habitat Euphorbia pentagona is known from the Eastern Cape between Uitenhage (with its westernmost occurrence near Kleinspoort, south-east of Steytlerville), Grahamstown, Fort Beaufort and the Kei River Mouth. There are further isolated populations along the Bashee River and also much further north along the Umtamvuna River, the boundary between what used to be Natal and the former Transkei.

Fig. 2.153.  Distribution of Euphorbia pentagona (© PVB).

Plants are usually found on flat to gently sloping or steep, rocky areas, usually on soils originating from shales. Euphorbia pentagona is especially associated with the dense, often spiny bush and scrub with many succulents that

develops along the warm and relatively dry sides of river valleys. Populations often flourish in the shallow soils rich in leaf-litter that develop in crevices on large outcrops of rock or on ledges on cliffs above rivers.

2.1 Sect. Anthacanthae

149

Fig. 2.154.  Euphorbia pentagona, shrub ± 1 m tall, Hellspoort, Grahamstown, South Africa, 3 Apr. 2012 (© PVB).

Diagnostic Features & Relationships Euphorbia pentagona is, as White et al. (1941) pointed out, among the largest of the species of ser. Meleuphorbia. Plants of E. pentagona may, on occasion, reach 3 m tall and achieve an enormous girth (Figs.  2.155 and 2.156). Still, they never reach the sheer bulk of large plants of E. polygona. In E. pentagona, the young branches are a bright and

Fig. 2.155.  Euphorbia pentagona, shrub ± 2.5 m tall (with Pavel Hanáček and Mario Neto), Hellspoort, Grahamstown, South Africa, 3 Apr. 2012 (© PVB).

shiny pale to deep green, mostly with five or six clear angles. This attractive colour does not last and is rapidly covered with a finely corky, grey-brown bark. Consequently, older plants are often very woody and grey in appearance and, in some areas, may also become heavily festooned with lichens. The stem (which may reach more than 10 cm thick just above the ground) and some of the first branches

150

2  Euphorbia subg. Athymalus

Fig. 2.156.  Euphorbia pentagona, shrub ±  3  m tall with lower branches grazed off, Hellspoort, Grahamstown, South Africa, 3 Apr. 2012 (© PVB).

g­ radually lose their spines and become almost cylindrical with age, with the angles reduced to lines of small tubercles. Leaf-rudiments, often well over 5 mm long, are present near the tips of the branches during the growing season and they are of a similar colour to the branches themselves. They soon fall off. Stout, simple, spines develop from the axils of most of the leaf-rudiments. These spines are initially pale green and pliable and bear several tiny, pale bracts near their apices. They very soon harden and become rigid. Euphorbia pentagona bears flowers in November– December (and later as well) and the capsules ripen in January–February. The slightly sweetly scented cyathia are borne on peduncles that bear much longer bracts than are found on the young spines. Each peduncle is terminated by a cyathium, while further cyathia develop on branches arising in the axils of bracts around the base of the first one. The cyathia are usually yellow-green and are subtended by two or three fairly conspicuous, spreading green bracts. Some of these peduncles persist as weak, branched structures among the tougher, simple spines, but many seem to disintegrate more rapidly than the spines. Euphorbia pentagona and E. heptagona are similar and the differences between them are subtle. The branches in E. heptagona generally have more angles on them (seven to nine being the most common counts), while in E. pentagona there are mostly only five or six angles. Consequently, the angles are a little more widely separated around the diameter of the branch in E. pentagona than they are in E. heptagona and they also have wider and more gently sloping grooves between them in E. pentagona. The bright and shiny green hue of young branches of E. pentagona is never seen in E. heptagona, where the surface is always slightly dull. The

Fig. 2.157.  Euphorbia pentagona, male cyathia, PVB 6892, Fort Beaufort, South Africa, 20 Nov. 2007 (© PVB).

Fig. 2.158.  Euphorbia pentagona, female cyathia, PVB 6892, Fort Beaufort, South Africa, 20 Nov. 2007 (© PVB).

2.1 Sect. Anthacanthae

151

leaf-rudiments in E. pentagona are considerably longer than those of E. heptagona and they have the same pale green colour as the branches, a colour rarely seen in the leaf-rudiments of E. heptagona. Spines are produced in both species but are much harder and thicker in E. heptagona, where they show also much greater variation in colour. They are often sparser along the stems in E. pentagona than in E. heptagona. The flowering structures of the two species differ in that the peduncles of E. pentagona are longer and invariably branch gradually (only rarely remaining unbranched), while those of E. heptagona are shorter and each always bears only one cyathium, remaining strictly unbranched. The bracts on the peduncles are considerably larger in E. pentagona, more spread out along the peduncle and those just beneath the cyathium are more conspicuous because of their spreading habit: in E. heptagona the minute bracts are found only just under the cyathium and remain pressed to it. The respective cyathia are similar, though they are generally larger with glands nearly twice as broad in E. pentagona than in E. heptagona. History The plants that Haworth described as E. pentagona were brought to England from the Cape by James Bowie in 1823. However, there is little (except ‘Spinae adhuc trilinearis ultraque’) in Haworth’s description to indicate that this name is applicable to the present concept of E. pentagona. The painting number 296/926 at Kew by G. Bond, said to represent this species, is of a very weak and imperfectly developed branch without spines (as noted by N.E. Brown on the painting) and it is doubtful whether this

Fig. 2.160.  Euphorbia pentagona, capsules with some male cyathia among them and many peduncles unbranched, PVB 12170, Perdepoort, near Jansenville, South Africa, 25 Oct. 2012 (© PVB).

Fig. 2.159.  Euphorbia pentagona, capsules on typically branched peduncles with some branched peduncles visible below too, PVB 13535, north of Riebeeck East, South Africa, 31 Oct. 2018 (© PVB).

152

even belongs to the present concept of this species. Brown (1915) doubted whether it was the specimen described by Haworth, since Haworth mentioned spines and these are absent from this drawing. A neotype was therefore selected.

2  Euphorbia subg. Athymalus

After that of Bowie, another early collection of E. pentagona was made by Johan F. and Carl Drège in March 1833 at 1500–2000’ on stony heights and precipitous banks of the Sundays River (Drège 8213 in S). This specimen was cited by Boissier (1862) and N.E. Brown (1915) as E. stellispina.

Fig. 2.161.  Euphorbia pentagona. A, side view of male cyathium (scale 2 mm, as for B). B, side view of female cyathium. C, anthers and bracteole (scale 1 mm, as for D). D, female floret. Drawn from: PVB 6892, Fort Beaufort, South Africa (© PVB).

However, the locality is too far to the east for E. stellispina. The specimen has numerous branched ‘spines’: the tips of the branches of the ‘spines’ are not at all sharp, they are slightly separated along the main shaft of the spine and many of them bear fruit. These features all suggest E. pentagona rather than E. stellispina where, in particular, the sharptipped branches of the spine are sterile and (in female plants) a single fruit is borne on the central shaft of the spine. The remarkable, disjunct population along the Umtamvuna River in southern Kwazulu-Natal was discovered by Anthony Abbott in October 1989. Euphorbia pillansii N.E.Br., Bull. Misc. Inform. 1913: 122 (1913). Type: South Africa, Cape, near Doornkloof River, between Muiskraal and Ladismith, Aug. 1907, N.S. Pillans sub BOL 12543 (BOL, holo.; K, iso.). Euphorbia pillansii var. albovirens A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 965 (1941). Type: South Africa, Cape, Paardekop near Spes Bona, 650 m, 3 Oct. 1925, Marloth 12543 (PRE, holo.). Euphorbia pillansii var. ramosissima A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 965 (1941). Type: South Africa, Cape, low hill between Montagu and Touws River, Aug. 1939, Dyer 4100 (K, holo.).

Unisexual spiny many-branched glabrous succulent shrub to 0.4 × 0.4 m, branching from base (and sparingly above) of similar stem tapering towards the base and continuing below ground into fibrous roots. Branches erect, clavate, 80–200 (400) × 30–60 mm, simple to occasionally rebranched, partly covered with tubercles, smooth, bluish green; tubercles obtusely conical, 3–7 mm long with transversely rectangular to hexagonal base, vertically fused into (4) 6–9 low angles, with broad shallow grooves between angles (sometimes flat between angles) and sometimes with faint fine horizontal lines separating tubercles, with solitary simple or branched spines in their axils; spines developing from modified persistent fertile or sterile short-shoots, straight to slightly ascending, rigid, 8–40 × 1.5–3 mm, with several small alternating glabrous (except for cilia along margins) spathulate bracts to 1–2 × 1 (2) mm along length, usually branching in axils of upper bracts into 3–7 ascending rigid spinelets 4–12  mm long, initially green to pink or purple and glabrous, drying to brown or grey; leaf-rudiments on tips of new tubercles towards apices of branches and stem, 1–2 × 1–2 mm, spreading, fleeting, deltate, acute, sessile. Synflorescences many per branch towards apex, each in axil of tubercle on solitary usually branched peduncle bearing 1–6 unisexual cyathia; cyathia cupular, glabrous, 2.5–5 mm broad (± twice as broad

2.1 Sect. Anthacanthae

in male as in female, 1.5–2  mm long below insertion of glands), with 5 lobes with crenulate margins, green; glands 5, transversely elliptic, 1.2–1.5  mm broad, erect, green, slightly depressed in centre, outer margins entire, stamens with glabrous pedicels, bracteoles filiform to hand-like, ciliate; ovary slightly ellipsoidal, glabrous, raised on very short pedicel < 0.2 mm long; styles 2–3 mm long, branched above middle and slightly spreading. Capsule 4–6 mm diam., very obtusely 3-angled and almost spherical, glabrous, greygreen, sessile.

153

Distribution & Habitat Euphorbia pillansii is fairly widely distributed in the semiarid parts of the south-western Cape. It is recorded on the south-western portion of the Tanqua Karoo from about 20 km north of Karoo Poort to north-west of Matjiesfontein. It is also widespread on the western portion of the Klein Karoo, from north-west of Montagu, around the southern foot of the Touwsberg, along the northern and southern foot of the Warmwaterberg and between Adamskraal, Muiskraal and Lemoenshoek.

Fig. 2.162.  Distribution of Euphorbia pillansii (© PVB).

Plants usually occur on gently sloping, stony hillsides and are often associated with the shale-like tillites of the Dwyka series. Here they grow together with shrubs of Rhigozum obovatum, usually with a host of other succulents and may be among the largest shrubs in these habitats. In other places, such as near Brandrivier, east of Lemoenshoek, they occur on shales overlain with a conspicuous cover of white, quartz pebbles, while near Adamskraal they occur on patches of broken-up Witteberg quartzite among shale ridges. In both habitats they grow together with various species of Gibbaeum: G. album and G. petrense around Brandrivier; G. heathii and G. pachypodium near Adamskraal.

Diagnostic Features & Relationships Euphorbia pillansii mostly forms compact and fairly dense shrubs that are usually a little broader than tall. Occasional old specimens consist of only a few rather long branches, but this is atypical. Although branches in most mature specimens are 6- to 9-angled, the stems are initially 4-angled and this persists in young specimens until they are some 40–50  mm tall. The relatively stout branches (frequently well over 40 mm in diameter) are a rather dull, bluish green colour that changes gradually to brown towards their bases. The colour of the branches is also somewhat obscured by the frequently dense armature of old fertile and sterile shortshoots. These spikes are very rigid and remain on the plant

154

Fig. 2.163.  Euphorbia pillansii, dense clump ±  30  cm diam., with Crassula rupestris, PVB 9971, west of Matjiesfontein, South Africa, 25 May 2005 (© PVB).

Fig. 2.164.  Euphorbia pillansii, laxly branched specimen, PVB 13157, NE of Montagu, South Africa, 12 Jun. 2016 (© PVB).

2  Euphorbia subg. Athymalus

for many years but are rarely sharp. The first spines produced on each branch are usually sterile and unbranched and some of these may become sharp. Short-shoots that bear cyathia are generally rather irregularly branched near their apex and these shoots then persist as a rigid shaft with several short and somewhat thinner but still rigid radiating spikelets around its apex. In mature plants the fertile short-shoots may occasionally be followed by a whorl or two of sterile spines, which are often unbranched and may be longer than the fertile short-shoots. Euphorbia pillansii flowers from May to July and then the tops of the branches are usually covered with the small cyathia. In both the male and female plants, the small erect glands are an inconspicuous, dull green that closely matches the colour of the branches. While the female cyathia are barely noticeable, the flowering of the male plants can be more conspicuous. The male cyathia are produced in such numbers, all maturing roughly simultaneously, that the yellow of the anthers and the white of the bracteoles are quite striking against the dull colour of the plant. The male cyathia emit a somewhat foetid, sweetish odour that has not been detected in the females. The female cyathia, which are about half the diameter of the males and are somewhat wrapped up in bracts, remain inconspicuous. In the shape of the plant and the branches, E. pillansii is similar to the widespread E. heptagona. However, it differs clearly in its considerably stouter branches. The branches are usually also fewer in number and not as densely packed as in E. heptagona (at least in its western habitats). In E. pillansii the angles along the branches are lower and broader and they also have much less prominent grooves between them than

2.1 Sect. Anthacanthae

155

Fig. 2.165.  Euphorbia pillansii, with Gibbeaum pachypodium, Ockertskraal, south of Ladismith, South Africa, 1 Dec. 2006 (© PVB).

Fig. 2.166.  Euphorbia pillansii, male cyathia, PVB 7332, Plathuis, west of Ladismith, South Africa, 2 May 2005 (© PVB).

Fig. 2.167.  Euphorbia pillansii, female cyathia, PVB 7332, Plathuis, west of Ladismith, South Africa, 2 May 2005 (© PVB).

does E. heptagona. The two species differ most clearly in the stouter, shorter, and usually many-forked, persistent spines in E. pillansii. In respect of the branched peduncles, E. pillansii is more reminiscent of E. pentagona from the Eastern Cape, where the peduncles, but not the sterile short-shoots, are forked. However, these two species differ more markedly in the colour of the branches and E. pentagona has much longer leaf-rudiments than E. pillansii, as well as differently shaped cyathial glands and considerably larger cyathia. The forked spines of E. pillansii also suggest the more stoutly-branched and altogether more massive E. stellispina. In E. stellispina branching of the peduncle takes place very close to its apex to form a whorl of sterile spinelets just beneath the cyathium. In E. pillansii each branch of the peduncle is fertile and branching is considerably more irregular, with the branches more spread out around the

apex of the shoot and more variable in length. Consequently, the spine-complex is neither as dense nor as regularly shaped as in E. stellispina. Both the spines and cyathia are glabrous in E. pillansii (finely pubescent in E. stellispina) and the two differ also in the shape of the cyathial glands. History Euphorbia pillansii is named after Neville S.  Pillans, who collected the type material between Ladismith and Muiskraal on the Klein Karoo in August 1907. Pillans’ specimen was identified in the Bolus Herbarium as Euphorbia stellispina, but N.E.  Brown realised that it was different and gave it a new name. A specimen of E. pillansii was also figured by Phillips (1929a) as ‘Euphorbia stellaespina’ but, otherwise, E. pillansii has gained little attention and has been relatively rarely collected.

156

2  Euphorbia subg. Athymalus

Fig. 2.168.  Euphorbia pillansii, capsules, PVB 7332, Plathuis, west of Ladismith, South Africa, 28 Sep. 2014 (© PVB).

Fig. 2.169.  Euphorbia pillansii. A, side view of male cyathia (scale 2 mm, as for B, C). B, side view of female cyathia. C, side view of female cyathium. D, anthers and bracteoles (scale 1 mm, as for E). E, female floret. Drawn from: PVB 7332, Plathuis, west of Ladismith, South Africa (© PVB).

Euphorbia polygona Haw., Misc. Nat.: 184 (1803). Neotype (Bruyns 2012): South Africa, Cape, Witpoortsberg, 2000– 3000’, Aug., Drège 8212 (S; BM, HBG, K, MEL, MO, P, TCD, W, iso.).

Euphorbia polygonata G.Lodd. Bot. Cab. 14: t. 1334 (1828). Lectotype (designated here): Bot. Cab. 14: t. 1334. Euphorbia horrida Boiss., Cent. Euphorb.: 27 (1860). Euphorbia polygona var. horrida (Boiss.) Schnabel,

2.1 Sect. Anthacanthae

Euphorbia World 9: 10 (2013), nom. illegit. Lectotype (Bruyns 2012): South Africa, Cape, Witpoortsberg, 2000– 3000’, Aug., Drège 8212 (S; BM, HBG, K, MEL, MO, P, TCD, W, iso.). Euphorbia horrida var. striata A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 964 (1941). Euphorbia polygona var. striata (A.C.White & al.) Schnabel, Euphorbia World 9: 20 (2013). Type: South Africa, Cape, 15 miles north of Steytlerville, Lückhoff 123 (missing). Euphorbia horrida var. noorsveldensis A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 965 (1941). Euphorbia polygona var. noorsveldensis (A.C.White, R.A.Dyer & B.Sloane) Schnabel, Euphorbia World 9: 20 (2013). Type: South Africa, Cape, 1.3 miles north of Jansenville, Aug. 1939, Dyer 4010 (PRE, holo.). Euphorbia horrida var. major A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 965 (1941). Euphorbia polygona var. major (A.C.White & al.) Schnabel, Euphorbia World 9: 18 (2013). Type: South Africa, Cape, Kruidfontein, 19 miles from Willowmore towards Rietbron, Aug. 1939, Dyer 4041 (missing). Euphorbia anoplia Stapf, Bot. Mag. 148: t. 8947 (1923). Euphorbia polygona var. anoplia (Stapf) Schnabel, Euphorbia World 9: 10 (2013). Type: cultivated plant of unknown origin, possibly from South Africa (K). Euphorbia polygona var. nivea Schnabel, Euphorbia World 7: 25 (2011). Type: South Africa, Long Kloof, Kleinrivier, 508 m, 15 Nov. 2010, Schnabel 1 (GRA, holo.). Euphorbia polygona var. hebdomadalis Schnabel, Euphorbia World 8: 12 (2012). Type: South Africa, Western Cape,

Fig. 2.170.  Distribution of Euphorbia polygona (© PVB).

157

Seweweekspoort, 672 m, 16 May 2012, Schnabel 5 (NBG, holo.). Euphorbia polygona var. exilis Schnabel, Euphorbia World 8: 20 (2012). Type: South Africa, Eastern Cape, foothills of Kouga Mountains, 65 m, 16 Nov. 2011, Schnabel 4 (GRA, holo.). Euphorbia polygona var. minor Schnabel, Euphorbia World 9: 13 (2013). Type: South Africa, Western Cape, 23 km north of Oudtshoorn near Koos Raubenheimer Dam, 717 m, 17 Nov. 2012, Schnabel 6 (NBG, holo.). Euphorbia polygona var. alba Schnabel, Euphorbia World 10: 24 (2014). Type: South Africa, Western Cape, ± 7 km north of Uniondale, 850 m, 11 Nov. 2013, Schnabel 7 (NBG, holo.). Euphorbia polygona var. ambigua Schnabel, Euphorbia World 10: 24 (2014). Type: South Africa, Eastern Cape, Burbank, 23 km west of Klipplaat, 800 m, 13 Nov. 2013, Schnabel 8 (GRA, holo.). Euphorbia cucumerina Willd., Sp. Pl., ed. 4, 2: 886 (1799). Lectotype (designated here): South Africa, F. le Vaillant, Reise in das Innere von Afrika 3: 410, t. 5 (1797). Unisexual (rarely bisexual) spiny (occasionally spineless) many-branched glabrous succulent shrub 0.15–2  m tall, branching from near base (and rarely above) of similar stem, with fibrous roots around base of stem and near bases of some branches, some branches also spreading underground with short rhizomes 20–100 mm long. Branches erect, clavate, 0.1–2  m × 40–200 (300) mm, simple to occasionally rebranched, covered with tubercles, grey-green (sometimes nearly white with heavy waxy layer); tubercles low and coni-

158

cal, 2–6 mm long laterally flattened and vertically fused into 7–20 prominent obtuse to acute often wing-like and undulating angles, with shallow to deep grooves between angles, with 1–5 (7) simple spines in their axils (spines sometimes absent), with fine horizontal ridges running up to bases of leaf-rudiments; spines from persistent fertile or sterile shortshoots, straight and spreading, rigid, 4–40 × 1–3 mm, with several minute alternating ovate to spathulate finely pubescent bracts to 2 × 1 mm mainly towards apex, initially suffused with red and finely pubescent, drying to brown to grey and glabrous; leaf-rudiments on tips of new tubercles towards apices of branches and stem, 1–2 × 1 mm, spreading, fleeting, deltate, acute, sessile. Synflorescences many per branch towards (but not at) apex, each in axil of tubercle on glabrous peduncle 2–12 mm long with 1–3 unisexual cyathia; cyathia cupular, glabrous outside, 4–7  mm broad, with 5 finely pubescent reddish green lobes with deeply divided margins, green to red or dark purple; glands 5, transversely elliptic, 2 mm broad, slightly separated, green to purple, slightly concave and smooth above, outer margins entire; stamens with pubescent pedicels, bracteoles filiform and pubescent; ovary

2  Euphorbia subg. Athymalus

globose, white-tomentose, raised on short pedicel < 0.5 mm long; styles ± 2 mm long, branching in upper half. Capsule 5–6  mm diam., almost spherical, pubescent, grey-green to purple, sessile. Distribution & Habitat Euphorbia polygona is widespread (Fig. 2.170) in the sandstone mountains of the southern and eastern parts of the former Cape Province, in ranges such as the Suurberg, Groot Winterhoek, Kouga and Baviaanskloof Mountains. In these, it is of fairly general occurrence from near Uniondale in the west to Grahamstown and Bathurst in the east, usually growing in shallow soils on drier, rocky outcrops in the wetter areas. It is also found on the slopes of the considerably drier sandstone mountains bordering the Great Karoo between Prince Albert, Willowmore, Steytlerville and Kommadagga. From west and north-west of Willowmore to north-east of Jansenville (on the southern Great Karoo) it also occurs sporadically in flat to hilly country on shales to the north of the sandstone mountains. Further west it is known from rather isolated populations along the Swartberg at Seweweeks

Fig. 2.171.  Euphorbia polygona, stout branches ± 18 cm thick, PVB 6306, SE of Rietbron, South Africa, 1 Jul. 1995 (© PVB).

Poort, along a few ridges south of the Cango Caves near Oudtshoorn, Meiringspoort and along the rocky edges of some of the north-facing gorges in the Rooiberg to the south of Calitzdorp. Euphorbia polygona is most common on dry and hot, north- or east-facing, rocky slopes where the soils are derived from sandstones. In such spots it occurs with a wealth of other succulents, often with Aloe ferox, many species of Crassula, Cotyledon woodii, the ubiquitous Portulacaria afra and often with several stapeliads.

It also occurs on shales, though somewhat less frequently. West and north of Jansenville, E. polygona occurs together with E. heptagona and E. radyeri, with the lattermost dominant in some spots and E. polygona dominating in others. Those growing in the flats and on low hills between Willowmore and Rietbron (as in Fig.  2.171) occur among short karoo-bushes, sometimes with Ceropegia (Hoodia) pilifera subsp. annulata and E. ferox and here they are often the tallest shrubs.

2.1 Sect. Anthacanthae

Fig. 2.172.  Euphorbia polygona, stout branches ± 15 cm thick, PVB 11504, Rooiberg, south of Calitzdorp, South Africa, 20 Oct. 2009 (© PVB).

Fig. 2.174.  Euphorbia polygona, large shrub ± 2 m diam. with branches ± 12 cm thick (Aloe striata flowering in front and behind), PVB 12991, west of Jansenville, South Africa, 4 Sep. 2015 (© PVB).

159

Fig. 2.173.  Euphorbia polygona, dense mound wedged between boulders with branches ±  12  cm thick, PVB 6323, Boesmanspoortberge, NW of Willowmore, South Africa, 24 Oct. 2009 (© PVB).

160

2  Euphorbia subg. Athymalus

Fig. 2.175.  Euphorbia polygona, nearly 2 m tall with branches ± 12 cm thick, PVB 9789, Groot River Poort, SE of Steytlerville, South Africa, 6 Apr. 2004 (© PVB).

Diagnostic Features & Relationships Plants of E. polygona are often large, forming impressive and dense, formidably spiky clumps that, in some areas, reach 2 m tall and at least 1 m broad. The most massive specimens (as in Fig. 2.174) are those found among the low hills of the southern Great Karoo to the north-west and west of Willowmore. Here, not only are the plants large (though usually with few branches), but the branches are also particularly stout, often reaching over 150  mm in diameter. Very large plants (as in Fig. 2.175) may also develop in the sandstone mountains (especially of the Groot Winterhoek and Suurberg), but here the branches are considerably more slender than those in the lower-lying karroid areas near Willowmore. Plants in the sandstone mountains (as in those north-west and north-east of Willowmore, Fig. 2.173) are usually relatively small. These smaller versions are generally very densely branched, forming cushion-like mounds between rocks. The smallest of all are those found south of the Cango Caves, where many of the plants are not more than 150 mm tall (some mature specimens only 80 mm tall), though densely branched and forming tight cushions between rocks. In most plants, the tubercles are joined into tall, usually quite thin and often conspicuously undulating angles along the length of the branches. Grooves that can be as much as 25 mm deep separate adjacent angles. The branches generally have a grey-green colour, but they are sometimes nearly white (especially in the mountains between Joubertina and Studtis, as in Fig. 2.177). This pale colour is caused by a thicker-than-usual layer of wax covering the surface, which wears off if the plants are grown in wetter climates. Variation in the thickness of this wax layer may also give rise to many,

Fig. 2.176.  Euphorbia polygona, branches ±  8  cm thick, among Passerina and sandstone outcrops, PVB 12151, Kouga Mountains, SE of Studtis, South Africa, 19 May 2012 (© PVB).

2.1 Sect. Anthacanthae

161

Fig. 2.177.  Euphorbia polygona, branches very pale and ±  8 cm thick, surrounded by bushes of Passerina, PVB 12151, Kouga Mountains, SE of Studtis, South Africa, 19 May 2012 (© PVB).

regular ascending darker lines running out from the grooves between the angles to the axil of each tubercle. Where present, the spines in E. polygona are stout and rigid, but they not especially sharp. In each axil the first spine to develop is usually (though not always) a sterile shortshoot and it is generally longer than any others that may arise later in the same axil. Up to five spines may develop in each axil. There are also some areas, especially on the hills between Port Elizabeth and Humansdorp, where the branches may be entirely without spines. Especially dense brushes of spines along the angles are found in the thick-branched plants that occur between Willowmore and Toorwaterpoort (as in the picture on the front of this volume). In all of them, initially the spine is soft and red, with several tiny bracts along its length, but it hardens off very soon, becoming rigid, grey and woody, in which state it persists for many years. Traces of the bracts along the spine remain in the form of the small basal tubercles, so that the spines are not smooth once they dry out. Leaf-rudiments are especially small in E. polygona and are present in the crown of the branch as small, recurved, brown scales that soon dry out and gradually wear off lower down the branch. Flowering in E. polygona usually takes place from October to January, though some plants flower throughout the summer and capsules are often present when the plant is still in flower. Plants in the west (near Cango, Rooiberg and Seweweekspoort) may flower from June to October as well. Fertile short-shoots usually arise in a ring a little away from the apex of the branch and mainly develop in the lateral axils (relative to the axis of the branch) of the first and other sterile short-shoots. In years of suitably-timed rainfall, the whole

Fig. 2.178.  Euphorbia polygona, spineless branches ±  6  cm thick (and some of them rhizomatous around the plant), PVB 6864, conglomerates near Hankey, South Africa, 24 Oct. 2012 (© PVB).

162

2  Euphorbia subg. Athymalus

Fig. 2.179.  Euphorbia polygona, rhizomatous habit of some young branches, PVB 12151, Kouga Mountains, SE of Studtis, South Africa, 19 May 2012 (© PVB).

top of each branch of the plant is covered with cyathia but, when the rainfall is not optimal, the cyathia mature in smaller numbers, mainly on the sides of the branches facing the sun. The cyathia give off an odour of something sweet mixed with rotten fruit, just like that in E. heptagona and they attract

considerable numbers of ants, bees and flies. The peduncles are much shorter than the spines and have several obvious bracts along their length. In the males the bracts are spread out along the short-shoot (longest close to the cyathium and shortest near the base) and the short-shoot is longer than in

Fig. 2.180.  Euphorbia polygona, dull yellow-green glands (male), PVB 11504, Rooiberg, south of Calitzdorp, South Africa, 30 Jul. 2013 (© PVB).

Fig. 2.181.  Euphorbia polygona, glands dull green suffused with purple (male), PVB 6323, Boesmanspoortberge, NW of Willowmore, South Africa, 24 Oct. 2009 (© PVB).

the females. In the females the bracts cluster beneath the cyathium and partly hide it. The cyathia vary from green (in the west) to red or even dark purple in the east and may have green (in the west), red or purple glands as well. The female cyathium is short and so the densely pubescent (and consequently rather whitish) ovary projects somewhat from it, even though it has only a very short pedicel beneath it.

Occasionally, plants that are predominantly female have been found to produce some male cyathia, especially towards the end of the flowering period, so the sexuality of the plants is not fixed (though this has not been observed in male plants). Among the species of ser. Meleuphorbia, Euphorbia polygona and E. stellispina have by far the stoutest branches,

2.1 Sect. Anthacanthae

163

Fig. 2.182.  Euphorbia polygona, flowering profusely, glands dull green deeply suffused with purple (male), PVB 11529, SE of Rietbron, South Africa, 24 Oct. 2009 (© PVB).

Fig. 2.183.  Euphorbia polygona, deep purple glands (male), PVB 12991, west of Jansenville, South Africa, 4 Sep. 2015 (© PVB).

though the very thick-branched forms of E. polygona from the Karoo north-west and west of Willowmore are much stouter than any in E. stellispina. Euphorbia polygona is easily separated from E. stellispina in that the spines in E. polygona are always simple (only occasionally absent) and are always forked in E. stellispina (where they are never absent). The covering of spines is also more uniform in E. polygona, while in E. stellispina they are often produced in ring-like patches along the branches and may be absent between these rings. Euphorbia stellispina also has much more prominent leaf-rudiments, which are clearly visible in neat rows at the apex of the branch when it is actively grow-

ing. The angles along the branches in E. stellispina are far less emphasized, with only shallow grooves between them. Euphorbia polygona is unusual among its relatives for its often red to purple cyathia with purple glands: in all the others (apart from E. pulvinata) the cyathia are green or yellow. It is also unusual for the fact that up to five spines may arise in the axil of each tubercle, while in most of the others there is only one spike in each axil. This makes for very ample protection of the plant against predation. Euphorbia polygona often occurs together with or near to E. heptagona and E. pentagona. Many hybrids between them may be observed and some of these (probably involving E.

164

2  Euphorbia subg. Athymalus

Fig. 2.184.  Euphorbia polygona, deep purple glands (female), PVB 12991, west of Jansenville, South Africa, 4 Sep. 2015 (© PVB).

Fig. 2.185.  Euphorbia polygona, capsules, sandstone outcrops east of Joubertina, South Africa, 26 Oct. 2018 (© PVB).

pentagona) were named E. inconstantia by R.A.  Dyer. Branches in such hybrids are generally of an intermediate thickness between the two species, with more angles than in E. heptagona or E. pentagona (and fewer than in E. polygona) and they have one to several spines in each axil along the branches. There is no evidence that they form breeding populations and so they are not here treated as a separate species. However, in some cases intermediates between these hybrids and their parents are found, so some of them are fertile. Detlef H. Schnabel has published many accounts of various forms of this species in an attempt to document its variability. This culminated in a book on this species alone

(Schnabel 2018). There seems to be a great deal of duplication in the names he proposed and, in several cases, he suggested names before he had seen the plants in habitat, adding to the unconvincing nature of these accounts. His ‘revision’ of the complex (Schnabel 2013) did not include citation of any specimens examined, so it remains unclear on what the ‘conclusions’ contained in it were based, apart from the very modest number of collections he made himself. Furthermore, several ‘new’ varieties and forms have appeared since this ‘revision’, leading to the impression that it was premature. White et al. (1941) recognized several varieties of E. horrida and some of these may be worth maintaining as subspecies of E. polygona, there being no obvious grounds for maintain-

2.1 Sect. Anthacanthae

ing E. horrida and E. polygona as separate species, as Schnabel also concluded. My own collections suggest that most of the forms found on the sandstones with acidic soils are difficult to separate into distinct taxa at any rank. Some are much larger than others. Those that occur high on slopes of mountains are often smaller (frequently not exceeding 30 cm tall, but with occasional plants still exceeding 50 cm tall), with shorter branches very densely packed into tight clumps (often very pale with a heavy covering of wax on the surface). Those occurring on lower slopes are more laxly branched with often much longer branches that regularly reach and exceed 1 m long. In the west (Meiringspoort and westwards, probably all referable to E. polygona var. minor), the cyathial glands are green and in the east they are red to purple-red (E. polygona var. polygona and ‘horrida’). Among the forms found on sandstones, some nearer the coast around Port Elizabeth have quite slender stems (40–80 mm thick) and some of these may even lack spines (Fig. 2.178). Those found on shales often have especially thick branches (which reach 20  cm and sometimes even nearly 30 cm thick), they may be taller than usual (and regularly exceed 1.5  m tall) and have green to brownish green (darker in the young cyathia, later becoming greener)

165

or purple glands (E. polygona var. major and var. noorsveldensis). History Haworth described Euphorbia polygona from cultivated material that came ostensibly from the Cape and had been introduced to England before 1790, probably from collections made by Francis Masson. Neither illustrations nor material preserved from these plants is known to exist. The earliest documented specimens are those collected by J.F. Drège in the ‘Witpoortsberg’ (probably the present-day Swanepoelspoort, north-east of Willowmore), which he first visited in August 1829 and again in August 1832, on both occasions in the company of his brother Carl. This was selected as a neotype in Bruyns (2012), but Carter (2014) selected another neotype for E. polygona, so that the same specimen does not serve as types for both E. horrida and E. polygona. The earlier neotypification is maintained here. There is good evidence (Dyer 1949: 14) that the first record of E. polygona is that of François le Vaillant (1795, Vol. 1: t. VI; 1797, Vol. 3: t. 5), who wrote here about his travels between 1783 and 1785. Willdenow’s name E. cucumerina was based on the figure in the German edition (Le

Fig. 2.186.  Euphorbia polygona. A, side view of male cyathium (scale 2 mm, as for B, D). B, side view of female cyathium. C, anthers and bracteoles (scale 1 mm). D, female floret. Drawn from: A, C, PVB 6323, Boesmanspoortberge, NW of Willowmore, South Africa. B, D, PVB 6864, near Hankey, South Africa (© PVB).

166

2  Euphorbia subg. Athymalus

Vaillant 1797: t. 5). Dyer (1949: 14) pointed out that Le Vaillant’s illustration is a completion of an unfinished sketch of Robert Gordon’s which is among Paterson’s collection of paintings (see the similar cases of E. meloformis and E. stellata) and so it is likely that this plant was not collected by Le Vaillant. Whatever its origin, the identity of this figure has always been somewhat mysterious, but the coloured version of the plate shows the purplish-red cyathia and the simple spines typical of many plants of E. polygona. On the basis of these facts, Dyer suggested that this very possibly came from a single, fairly short, sparingly spined branch of E. polygona. Le Vaillant maintained that this plant was observed between the Groene River and the Koper Berg (near Springbok) in Namaqualand. However, his localities could be unreliable (Jordaan 1973), as in the case of E. stellata, which he claimed came from southern Namibia. Since no members of ser. Meleuphorbia occur in Namaqualand, his locality for E. cucumerina can be discounted. It is also possible that the name E. polygonata was merely a misspelling of E. polygona, but it is nevertheless listed above. In its westernmost known locality, namely Seweweekspoort, E. polygona was first recorded by Max Otzen in July 1940 (specimen at K) and a photograph of this by Herbert Krüger appeared in White et al. (1941).

± 1 mm, with several small alternating glabrous spathulate bracts to 1 × 1 mm along length, initially green or red, drying to grey, glabrous to minutely pubescent; leaf-rudiments on tips of new tubercles towards apices of branches, 1–5 × 1 mm, spreading, fleeting, linear to lanceolate, acute, sessile. Synflorescences many per branch towards apices, each solitary in axil of tubercle on red to green (rarely branched) peduncle 1–6 (15) mm long (usually longer in males than females) with 1 terminal cyathium (sometimes 2–4 more developing from axils of bracts beneath it), glabrous to pubescent, with 2–3 broadly lanceolate acute to obtuse glabrous to pubescent apically toothed bracts (often with marginal cilia) 2–3 mm long beneath terminal cyathium and 1–2 bracts further down along peduncle; cyathia cupular, glabrous, 1.5–2.5 × 2–5 mm (almost twice as broad in male as in female), with 5 lobes with deeply toothed margins, green; glands 5 (6), transversely elliptic, 1–1.5  mm broad, erect, green to brown-green or yellow-green often with darker margins, slightly depressed in centre, outer margins entire, stamens with pubescent pedicels, bracteoles filiform and sparsely pubescent; ovary globose, glabrous, raised on very short pedicel < 0.2 mm long; styles 1.5–2 mm long, branched to middle and recurved only near apex or widely spreading, sparsely pubescent. Capsule 3–5  mm diam., obtusely 3-angled, glabrous, grey-green to red, sessile.

Euphorbia pseudoglobosa Marloth, S.  African Gard. 19: 191 (1929). Type: South Africa, Cape, near Krombeks River, Riversdale distr., Sept. 1933, Muir 4089 (PRE, holo.).

Euphorbia pseudoglobosa is restricted to the WorcesterRobertson Karoo, the Klein Karoo and parts of the coastal plain to the south of these areas. On the Klein Karoo it is similar to its geographical neighbour, E. susannae. Euphorbia susannae differs by the fact that the stem just reaches the surface of the ground, where it is surrounded by a rosette of branches, with each of them also just reaching the surface of the ground. In E. pseudoglobosa the stem is not visible above the ground (being pulled below it as the plant ages) and the branches usually protrude from the surface. The branches are usually much shorter in E. susannae and they do not taper into a slender neck below the ground by which they are joined to the stem, as one finds in E. pseudoglobosa. Euphorbia susannae is also distinguished by the much more prominent tubercles which taper to slender tips (tubercles flattened and usually blunt in E. pseudoglobosa). Superficially, subsp. pseudoglobosa is also similar to small forms of E. tubiglans, but the two differ clearly in the much larger, widely spreading bracts that subtend the cyathia in E. tubiglans and the cyathial glands that are flat to slightly depressed in the centre in E. pseudoglobosa (and with a tubular shape in E. tubiglans). Previously E. pseudoglobosa was treated as separate from E. nesemannii. However, in the western part of the coastal plain (between Bredasdorp and Heidelberg), E. pseudoglobosa is much more variable than previously realized and examples of this are shown in Fig. 2.188. Branches

Unisexual spineless to spiny few- to many-branched dwarf glabrous succulent to small shrub, subterranean and forming rosette 20–50 (150) × 50–150 mm to shrublet 0.1– 0.3 (0.7) × 0.1–0.5 m, branching from around apex of fusiform wholly or partly subterranean stem 20–30  mm thick tapering into tap root with fibrous roots. Branches 1–15, erect to spreading, globose to clavate with upper surfaces just protruding from surface of ground to cylindrical and tapering below into narrow subterranean neck 10–30  mm long above stem (rarely rhizomatous), mostly simple, 5- to 10- (14) angled, 20–200 (700) × (8) 10–30 (40) mm, partly covered with obscure tubercles, smooth, green to browngrey- or red-green (longer branches corky below with age); tubercles obtusely conical and usually very inconspicuous, 4–8  mm long with transversely rectangular to hexagonal base, vertically (rarely spirally) fused into (4) 5–10 (14) low obtuse to acute angles, with shallow vertical broadly triangular grooves between angles (sometimes only near apex) and sometimes with fine horizontal lines separating tubercles, occasionally with few solitary simple spines in their axils (always present in subsp. nesemannii); spines from persistent sterile and fertile short-shoots, straight, rigid, 5–20 ×

2.1 Sect. Anthacanthae

167

Fig. 2.187.  Distribution of Euphorbia pseudoglobosa (▼ = ssp. nesemannii; ● = ssp. pseudoglobosa; ▲ = ssp. vlokii) (© PVB).

of plants that are sheltered may reach 30 cm long and are much taller than those that are exposed (some of which are only 1 cm tall). These longer branches have more angles (up to 14), are often thicker and bear many persistent spikes. The tallest and thickest-branched plants found on outcrops of shale are indistinguishable from E. nesemannii and they grow together with others with shorter, thinner and fewerangled branches that are typical of E. pseudoglobosa. The branches in E. nesemannii on the Worcester-Robertson Karoo are mostly adorned with persistent peduncles and

Fig. 2.188.  Euphorbia pseudoglobosa ssp. pseudoglobosa, variation in length of branches and the presence of persistent peduncles in a single population, PVB 13732, Beyerskraal, Bredasdorp, South Africa, 5 Jul. 2019 scale bar indicates 3 cm (© PVB).

this extends to the populations of E. pseudoglobosa on the coastal plain between Bredasdorp and Heidelberg where, especially in the west, many plants have persistent peduncles, though in each population there are also some plants whose branches lack them entirely. Consequently, E. nesemannii is now treated as a third subspecies of E. pseudoglobosa. Of these three subspecies, one occurs on the western Klein Karoo and on the coastal plain between Bredasdorp and Riversdale, another (involving the largest forms of E.

168

pseudoglobosa) is mainly restricted to the WorcesterRobertson Karoo and the third is only known near

2  Euphorbia subg. Athymalus

Dysselsdorp, east of Oudtshoorn on the eastern part of the Klein Karoo. They may be separated as follows:

1. Branches mostly > (50) 80 mm long and frequently 100–300 mm long............................................................................2. 1. Branches mostly < 80 mm long (usually only longer than 50 mm in well-sheltered plants), often only 10–30 mm long and partly submerged in soil.........................................................................................................................subsp. pseudoglobosa 2. Branches with persistent sterile peduncles, (6) 9- to 15-angled, styles ascending..................................subsp. nesemannii 2. Branches without persistent sterile peduncles, usually 5- to 8-angled, styles widely spreading........................subsp. vlokii

Euphorbia pseudoglobosa subsp. pseudoglobosa Euphorbia frickiana N.E.Br., Cact. Succ. J. (US) 2: 491 (1931). Type: South Africa, Riversdale div., Ferguson comm. Frick (K, holo.). Euphorbia juglans Compton, J.  S. African Bot. 1: 126 (1935). Euphorbia pseudoglobosa var. juglans (Compton) Schnabel, van Veldh. & J.G. Marx, Euphorbia World 14: 13 (2018). Type: South Africa, Cape, about 20 miles west of Ladismith, Feb. 1932, Compton 3951 (BOL, holo.). Euphorbia pseudoglobosa var. oshoekensis Schnabel, van Veldh. & J.G.Marx, Euphorbia World 14: 17 (2018). Type: South Africa, west of Ladismith, Oshoek, 610 m, 10 Nov. 2016, Schnabel 17 (NBG, holo.). Branches 20–50 (150) × 12–40 mm, mostly protruding from soil for ± same distance as length of subterranean part, 5- to 7- (9) angled, angles often rounded, occasionally with blunt persistent peduncles 4–15 mm long in axils of tuber-

Fig. 2.189.  Euphorbia pseudoglobosa ssp. pseudoglobosa, walnutshaped branches among small stones, PVB 2613, Kruisrivier, NE of Montagu, South Africa, 11 Nov. 1995 (© PVB).

cles. Synflorescences finely pubescent to glabrous, occasionally one or two additional cyathia arising in axils of upper bracts; cyathium often finely pubescent outside, with spreading glands; styles ascending and spreading near apices. Distribution & Habitat Subsp. pseudoglobosa is recorded on the western side of the Klein Karoo between Lemoenshoek (east of Barrydale), the southern foot of the Anysberg and Ouberg Pass (near Montagu). Several collections (of which the type is one) have also been made on the southern coastal plain between Bredasdorp and Riversdale. Subsp. pseudoglobosa grows in arid, exposed spots on flat to gently sloping ground, sometimes in crevices on low shale outcrops. On the Klein Karoo it may be found in flat patches among quartz pebbles (sometimes with Gibbaeum pubescens), in crevices between small vertical plates of shale on low shale outcrops (where it sometimes grows with the unusual

2.1 Sect. Anthacanthae

169

Fig. 2.190.  Euphorbia pseudoglobosa ssp. pseudoglobosa, male plant, among stones, PVB 2613, Kruisrivier, NE of Montagu, South Africa, 2 May 1987 (© PVB).

Gibbaeum nebrownii, Rhinephyllum muirii and other small succulents) or in hard loam among small sandstone rocks and very small bushes, here usually with a wealth of other small succulents. Habitats on the coastal plain (south of the

Langeberg) are usually exposed, pebbly alluvial deposits or patches of quartz-gravel on stony loam overlaying shales. These spots are also rich in small succulents.

Fig. 2.191.  Euphorbia pseudoglobosa ssp. pseudoglobosa, with slightly more prominent angles on branches, PVB 2463, east of Kareevlakte, South Africa, Jan. 1985 (© PVB).

Fig. 2.192.  Euphorbia pseudoglobosa ssp. pseudoglobosa, with much more prominent angles on branches, PVB 2463, east of Kareevlakte, South Africa, Jan. 1985 (© PVB).

Diagnostic Features & Relationships In subsp. pseudoglobosa most of the plant is buried in the ground. Initially, the stem is solitary and it projects very slightly from the ground. As it develops it expands, especially below the surface, developing a strong, gradually tapering tap-root, with fibrous roots along its length. With

the development of the branches, the apex of the stem usually becomes submerged (at a depth of 10–50 mm) and disappears entirely from view, pulled down into the ground by the considerable tap-root. In shape and form the stem is similar to the branches, though it is generally thicker and longer than they are.

170

2  Euphorbia subg. Athymalus

Fig. 2.193.  Euphorbia pseudoglobosa ssp. pseudoglobosa, walnutshaped branches with occasional persistent peduncles, PVB 13477, near mouth of Breede River, South Africa, 3 Dec. 2017 (© PVB).

Fig. 2.194.  Euphorbia pseudoglobosa ssp. pseudoglobosa, slightly longer walnut-shaped branches, PVB 13477, near mouth of Breede River, South Africa, 3 Dec. 2017 (© PVB).

Mostly the branches are at least partly submerged in the ground and, when growing in the open, only the tips are visible. Then their lowly habit and their colour matches the surrounding stones and gravel very closely. If sheltered, they can become quite long (to 10  cm or more) and etiolated. Each branch is elliptic in outline and tapers beneath into a narrow, slender neck up to 3 cm long before joining the stem. The tubercles on the branches are usually very obscure and fused into obtuse angles. When turgid this lends the branches on some plants their walnut-like shape, which Compton used to name this subspecies. In many localities on the Klein Karoo, though, this feature is variable, so that one finds

plants in which the angles along the branches are almost obsolete and others where the angles are quite sharply defined. In plants from south of the Langeberg there is less variability in the prominence of the angles. Here the old peduncles may also persist as blunt spines after the cyathia have fallen off. This does not happen on the Klein Karoo, where plants are spineless. Flowering in subsp. pseudoglobosa takes place between April and July and is often very prolific, with the tips of the branches usually covered for a short while with cyathia. Both male and female cyathia are small, with the male cyathia about twice the diameter of the females and borne

2.1 Sect. Anthacanthae

171

Fig. 2.195.  Euphorbia pseudoglobosa ssp. pseudoglobosa, female plant, with angled branches, PVB 6704, Lemoenshoek, east of Barrydale, South Africa, 4 Jun. 2012 (© PVB).

on longer peduncles. Further cyathia occasionally develop in the axils of the bracts just beneath the first cyathium. The female cyathia are on short peduncles and are often partly hidden by the bracts around them. Both male and female cyathia emit quite a strong, sweet scent. The cyathia are

ornamented with small, flat, usually not quite contiguous, greenish entire glands. Those on the Klein Karoo are usually finely pubescent while those from the southern coastal plain are often glabrous but may be pubescent too.

Fig. 2.196.  Euphorbia pseudoglobosa ssp. pseudoglobosa (A, B, D, F, G) and ssp. vlokii (C, E, H). A–C, side view of male cyathium (scale 2 mm, as for B, C–E). D, E, side view of female cyathium. F, anthers and bracteoles in dissected cyathium (scale 1 mm, as for G, H). G, H, female floret. Drawn from: A, B, D, F, G, PVB 6704, Lemoenshoek, east of Barrydale, South Africa. C, E, H, PVB 12582, near Dysselsdorp, South Africa (© PVB).

172

History Subsp. pseudoglobosa was discovered by J.  Muir in September 1928 near the Krombeks River, south of Heidelberg. The plants that Compton described as E. juglans were discovered in February 1932 on the Klein Karoo west of Ladismith, along the southern foot of the Anysberg. Euphorbia pseudoglobosa subsp. nesemannii (R.A.Dyer) Bruyns, comb. et stat. nov. Euphorbia latimammillaris Croizat, Cact. Succ. J. (US) 4: 333, fig. 5 (1933). Lectotype (selected here): South Africa, Cape, Worcester, Pillans (K). Euphorbia nesemannii R.A.Dyer, Bull. Misc. Inform. 1934: 267 (1934). Lectotype (Bruyns 2012): South Africa,

2  Euphorbia subg. Athymalus

Cape, koppie west of Robertson, 300’, Jul. 1930, Nesemann sub Dyer 2441 (GRA; K, iso.). Many-branched shrub 0.1–0.3 (0.7) × 0.1–0.5 m, branching from base and above from short thicker eventually partly subterranean nearly cylindrical stem (rarely with slender rhizomes from stem). Branches erect to spreading, cylindrical to slightly club-shaped and 6- to 14-angled, 60–200 (700) × (8) 15–40 mm, with solitary simple blunt rigid persistent spines 5–15 × ± 1 mm in axils of tubercles. Synflorescences glabrous, with 1 terminal cyathium (occasionally another 2  in axils of bracts); cyathia with spreading glands; styles 1.5–2  mm long, ascending.

Fig. 2.197.  Euphorbia pseudoglobosa subsp. nesemannii, ± 0.2 m tall with branches joined to a short underground stem, PVB 13050a, road to Scherpenheuwel, SE of Worcester, South Africa, 22 Oct. 2015 (© PVB).

Distribution & Habitat Subsp. nesemannii is mainly found on the WorcesterRobertson Karoo from Worcester to Ashton and Bonnievale (though it is much less common around Bonnievale than further west), with a single record from Montagu. There are also two records from the southern coastal plain at Stormsvlei and another north-east of Bredasdorp which belong to this subspecies rather than to subsp. pseudoglobosa. Subsp. nesemannii is common on the low shaly hills between Worcester and Robertson as well as in the flats between them. They are also found on some of the higher hills in the area. Habitats vary from steep, dry northern slopes (as on the Gannaberg and the Rooiberg) to flat areas (especially around Robertson), both in hard, loamy ground and in more sandy soil (as near Scherpenheuwel). It usually occurs with many other succulents (including Euphorbia burmanni and E. rhombifolia) among shrubs of Pteronia

paniculata, but many of its low-lying habitats have been and continue to be destroyed by viticulture and burgeoning townships. Diagnostic Features & Relationships Subsp. nesemannii may form quite robust shrubs which consist of few to many erect to somewhat sprawling branches arising from around the apex of a short partly subterranean stem. Mostly, the branches are 0.15 to 0.3  m tall, around 30 mm thick, with about 10 angles. For most of their length they bear short, spike-like and simple, persistent sterile shortshoots which gradually wear off to disappear entirely towards the base of the branch. Specimens growing higher up on steep slopes develop into smaller clumps only 150  mm tall and broad (as in Fig. 2.199). In such situations the branches are often much more slender, only 6-angled and they are paler than those of larger and more robust plants found lower down.

2.1 Sect. Anthacanthae

173

Fig. 2.198.  Euphorbia pseudoglobosa subsp. nesemannii, in stabilized sand with Ruschia and Pteronia paniculata, PVB 13050a, road to Scherpenheuwel, SE of Worcester, South Africa, 22 Oct. 2015 (© PVB).

Fig. 2.199.  Euphorbia pseudoglobosa subsp. nesemannii, male plant ± 0.1 m tall, slenderand attractively-branched form on steep slopes of Rooiberg, west of Robertson, South Africa, 9 Jun. 2015 (© PVB).

In subsp. nesemannii, flowering usually takes place from May to July, often irrespective of whether the winter-rains have begun. Cyathia are borne on short-shoots that are usually shorter than the sterile short-shoots. Each cyathium is surrounded at the top of the short-shoot by bracts that are slightly larger than those along the shoot, but which do not exceed the cyathium. Somewhat cupular in shape, the cyathia are relatively small, with the male cyathia considerably more prominent than the female cyathia and almost twice their diameter. The plump glands are small and deep green to bright yellow-green above, contrasting slightly with the

duller grey-green of their sides and of the rest of the cyathium. Once flowering is over the male and unpollinated female cyathia fall off and their peduncles persist as short, blunt spikes. The capsules have mostly exploded by the end of October and their peduncles also remain as short, blunt spikes. The westernmost representative of ser. Meleuphorbia, subsp. nesemannii is of similar appearance to small plants of E. pillansii, as they occur just north of Montagu. However, it is readily separated from E. pillansii (where the branches are usually slightly thicker) by the short and simple spines. It is

174

2  Euphorbia subg. Athymalus

Fig. 2.200.  Euphorbia pseudoglobosa subsp. nesemannii, male cyathia, PVB 13050a, road to Scherpenheuwel, SE of Worcester, South Africa, 5 Jul. 2019 (© PVB).

also somewhat similar to E. heptagona. However, the branches (especially in plants growing in flat areas) are often much thicker than those of E. heptagona, with more angles that are less prominent and with fewer, shorter spines. In E. heptagona, too, the stem is not as short as in subsp. nesemannii. In subsp. nesemannii, the cyathia are borne on peduncles that are far shorter than in E. heptagona, so that they are much more densely clustered around the tips of the branches (though they are not as short as in E. ferox).

Fig. 2.201.  Euphorbia pseudoglobosa subsp. nesemannii, female cyathia, PVB 13050a, road to Scherpenheuwel, SE of Worcester, South Africa, 23 Jun. 2019 (© PVB).

History Subsp. nesemannii was first recorded by Louisa Kensit in July 1901 and again by W.F. Purcell in 1914. A collection of N.S. Pillans from near Worcester was included by N.E. Brown under ‘E. fimbriata’. Among the specimens preserved at PRE from the Marloth herbarium, Marloth had placed some of subsp. nesemannii under E. captiosa (= E. ferox) and some under E. cereiformis (see under E. mammillaris). A specimen of subsp. nesemannii collected by J. Smith from Angora, near Bonnievale, was included by Marloth under his rather muddled concept of E. tubiglans (PRE records). In March 1930, A. Nesemann brought material to both R.A. Dyer and Marloth (PRE records) and it was from these plants that Dyer described it. As Croizat (1965: 640) pointed out, his name E. latimammillaris has precedence over Dyer’s E. nesemannii. Croizat (1933) published this name partly to replace the E. fimbriata of N.E. Brown (1915: 347), which did not agree in several aspects with Scopoli’s E. fimbriata. Of the various citations that Criozat gave here, only that by Brown included the citation of a specimen (namely Cape,

2.1 Sect. Anthacanthae

175

Fig. 2.202.  Euphorbia pseudoglobosa subsp. nesemannii, capsules, PVB 13050a, road to Scherpenheuwel, SE of Worcester, South Africa, 22 Oct. 2015 (© PVB).

Fig. 2.203.  Euphorbia pseudoglobosa subsp. nesemannii. A–C, side view of male cyathium (scale 2 mm, as for B–D). D, side view of female cyathium. E, F, anthers, bracteoles and sterile female in dissected male cyathium (scale 1 mm, as for F–H). G, female floret with bracteoles in dissected female cyathium. H, base of female floret. Drawn from: PVB 13050a, road to Scherpenheuwel, SE of Worcester, South Africa (© PVB).

Worcester, Pillans (K)) and so this is chosen here as the lectotype of Croizat’s E. latimammillaris. This specimen belongs to subsp. nesemannii. Euphorbia pseudoglobosa subsp. vlokii Bruyns, Haseltonia 25: 40 (2018). Type: South Africa, Western Cape, 6 km south of Dysselsdorp, 450 m, 7 Aug. 2013, Bruyns 12582 (BOL, holo.).

Euphorbia pseudoglobosa var. dysselsdorpensis Schnabel, van Veldh. & J.G.Marx, Euphorbia World 14: 14 (2018). Type: South Africa, ± 23 km east of Oudtshoorn, 450 m, 8 Nov. 2016, Schnabel 16 (NBG, holo.). Branches 50–200 × 15–25 mm, mostly protruding from soil for much greater length than subterranean part, (4) 5- to 8- (9) angled, angles often prominent and not rounded. Synflorescences glabrous. cyathia solitary on each peduncle;

176

cyathium glabrous outside, with erect glands pressed against lobes; styles widely spreading above ovary. Distribution & Habitat Subsp. vlokii is known from a single population near Dysselsdorp, east of Oudtshoorn. Here it grows among large shrubs of Ruschia pungens, scattered trees of Dodonea,

2  Euphorbia subg. Athymalus

Carissa bispinosa and various species of Pteronia on slopes covered with deposits of an alluvial nature (with many rounded pebbles both on the surface and below it to some depth) in red loam. Succulents are also plentiful here and it occurs together with E. ferox subsp. calitzdorpensis, E. heptagona, Senecio radicans and several species of Crassula.

Fig. 2.204.  Euphorbia pseudoglobosa ssp. vlokii, branches ± 6 cm long, PVB 12582, near Dysselsdorp, South Africa, 25 May 2014 (© PVB).

Fig. 2.205.  Euphorbia pseudoglobosa ssp. vlokii, branches ±  15  cm long, PVB 12582, near Dysselsdorp, South Africa, 25 May 2014 (© PVB).

Diagnostic Features & Relationships Plants of subsp. vlokii usually have branches up to 10 cm tall and even to 20 cm long on occasion, arising from the hidden stem. Most plants have 5- to 8-angled branches but plants with 4-angled branches are also occasionally found. The angles on the branches are more pronounced than is usually the case in subsp. pseudoglobosa so that the branches are never walnut-shaped as in subsp. pseudoglobosa. Florally the two subspecies differ by the glabrous peduncle and cyathium of subsp. vlokii (usually pubescent in subsp. pseudoglobosa on the Klein Karoo), the ascending cyathial glands that are pressed against the lobes (more

spreading in the other two subspecies) and by its more spreading styles in the females. Subsp. vlokii has been found to flower over most of the year: in habitat cyathia were observed in August 2013 and in May 2014 while in cultivation they are produced over most of the summer and into the winter. The cyathia emit only a faint sweetish odour, if any at all. At the type locality hybrids were observed with E. heptagona. These had longer branches than subsp. vlokii that rebranched frequently, with considerably longer red peduncles (though shorter than in E. heptagona) and smaller bracts. Many of these peduncles persist as hardened, blunt spikes.

Fig. 2.206.  Euphorbia pseudoglobosa ssp. vlokii, some branches ± 18 cm long, PVB 12582, near Dysselsdorp, South Africa, 25 May 2014 (© PVB).

Fig. 2.207.  Euphorbia pseudoglobosa ssp. vlokii, female cyathia, PVB 12582, near Dysselsdorp, South Africa, 14 Nov. 2015 (© PVB).

2.1 Sect. Anthacanthae

History Subsp. vlokii was first observed by Jan Vlok at the type locality on 5 April 1993. This remains the only place where it is known. Euphorbia pulvinata Marloth, Trans. Roy. Soc. South Africa 1: 315 (1909). Type: South Africa, Cape, Queenstown, Marloth 4372 (missing). Neotype (Bruyns 2012): South Africa, Cape, Queenstown, Nov. 1898, Galpin 2527 (PRE; K, iso.). Unisexual spiny many-branched glabrous mound-forming succulent 0.05–0.3 × 0.2–2 m, branching and rebranching extensively mainly from base and rarely above from similar stem, with fibrous roots near bases of stem and many of branches. Branches spreading to erect, initially globose becoming cylindrical to slightly club-shaped and 6- to 10 (12)-angled usually with slightly depressed apex, 20–200 × 20–55 mm, rebranching from base, partly covered with tubercles, smooth, shiny green becoming brown and somewhat woody lower down with age; tubercles conical, 3–6 mm long and vertically fused into 6–10 (12) prominent acute angles, with broad triangular vertical grooves between angles and usually without horizontal lines separating tubercles, many with solitary simple spines in their axils; spines from persistent sterile short-shoots, straight, woody, rigid, 6–20 × 1–1.5 mm, with several small alternating glabrous deltate bracts 1–1.5 × ± 1 mm along length, initially glaucous pink to red, drying to brown then grey, glabrous; leaves on tips of new tubercles towards apex of branches and stem, 2–50 × 2–7 mm, erect to spreading, soon deciduous, linear-lanceolate to linear-oblong or narrowly spathulate, grey-green, slightly channelled above, acute, slightly fleshy, sessile. Synflorescences many per branch near apex, each solitary in axil of tubercle on unbranched grey-green to reddish green shoot 1–8 mm long (and later 2–4 mm thick) with 1 terminal unisexual cyathium (sometimes 2 more arising in axils of uppermost bracts), glabrous to sparsely pubescent, with 2–4 broadly ovate finely and sparsely ciliate and sometimes sparsely pubescent purple-red to purplish green bracts 2–5  mm long just beneath cyathium hiding shaft of shoot, others below reddish green and smaller (those around female cyathium narrower than around male); cyathia conical-cupular, glabrous or very sparsely pubescent, 3–6 mm broad (often broader and longer in male than female, ± 2.5 mm long below insertion of glands), with 5 purple-red to pink (green) whitepubescent lobes with prominently toothed margins, purplered to green; glands 5 (6), transversely elliptic to almost circular, 1.5–2 mm broad, spreading, mostly not contiguous, dark purple-red to green with purple or red margins to yellow-green, smooth to pitted on upper surface, outer margins

177

entire, stamens with pubescent pedicels, bracteoles filiform and sparsely pubescent; ovary ellipsoidal to spherical, glabrous, nearly sessile; styles 1.5–7  mm long, reddish green, branched and spreading to recurved in upper half. Capsule 4.5–6  mm diam., almost spherical, glabrous, grey-green to purple, sessile. Distribution & Habitat Euphorbia pulvinata is very widely distributed along the eastern side of South Africa, from King Williams Town, Queenstown and the mountains around Cradock though the lower and drier parts of the Drakensberg and the drier valleys of its foothills to northern Natal and Swaziland. It is also widely distributed along the mountainous eastern flank of the former Transvaal where it reaches as far north as the Soutpansberg. It occurs at altitudes between 200 and 1800  m and in some areas (as around Tarkastad) is regularly covered with light falls of snow in winter and early spring. Plants are generally found on gentle to steep slopes with short grasses and low bushes. They usually grow where the ground is shallow, over submerged shale or sandstone slabs or among partially buried doleritic boulders and then they may fill the gaps between these boulders. In the northern part of its range, E. pulvinata is often found on exposed outcrops of sandstone, granite or dolerite, where it may have a more free-standing habit. They usually occur where the grass is less dense, but grass is frequently still present and larger plants are often heavily scorched by the fires that frequently rage through the grasslands during the dry season. Euphorbia pulvinata is by far the most widely distributed of all members of ser. Meleuphorbia and is both tolerant of and requires far more moisture than any of the others. It is also the only one that has advanced so far into the parts of South Africa that receive rainfall in summer. Diagnostic Features & Relationships Typically, plants of E. pulvinata are very densely branched and form low, cushion-like mounds. These mounds vary from 0.3  m to around 2  m broad. In such mounds the branches are so tightly packed as to form an almost continuous surface which undulates among and even over some of the surrounding rocks. In the northern part of its range (and especially in the Soutpansberg), the plants tend to be much less cushion-like and form somewhat ‘leggy’ shrubs with few, longer, more slender and more loosely arranged branches. Young branches in E. pulvinata are usually almost exactly spherical but they gradually grow longer to become cylindrical and this elongation allows the mound to expand. Only the apex of the branch remains green (frequently only

178

Fig. 2.208.  Distribution of Euphorbia pulvinata (© PVB).

Fig. 2.209.  Euphorbia pulvinata, laxly-branched plant ± 20 cm tall, PVB 12054, western Soutpansberg, South Africa, 3 Nov. 2011 (© PVB).

2  Euphorbia subg. Athymalus

2.1 Sect. Anthacanthae

179

Fig. 2.210.  Euphorbia pulvinata, densely-branched and leafy plant ± 30 cm diam., PVB 12068, NE of Wartburg, South Africa, 7 Dec. 2011 (© PVB).

Fig. 2.211.  Euphorbia pulvinata, huge mounds nearly 2 m diam. among doleritic outcrops, PVB 9392, near Steynsburg, South Africa, 17 Dec. 2002 (© PVB).

the upper 10 to 20 mm) and below this the lower parts gradually becoming hard and rigid, with a brown surface and a tough, woody consistency that is hard to cut. The uppermost, green parts of each branch usually have a brightly shiny, fresh green colour, though this is absent in plants in the Soutpansberg, where they are dull green. In most plants the branches typically have six to eight prominent, fairly sharp angles (formed by the almost complete vertical fusion of the

tubercles), separated by deep, broadly V-shaped grooves. The apex of each branch is usually slightly depressed. Euphorbia pulvinata is widely known as ‘mother-inlaw’s cushion’. This name is derived from its densely cushion-forming habit combined with the sharp and rigid spines that are produced along the angles in the axils of the tubercles and form a formidable, though not very dense protection for the upper surface of the mound. The spines are somewhat randomly produced, often with only every second

180

Fig. 2.212.  Euphorbia pulvinata, flat mats ± 1 m diam. with other succulents including E. mauritanica, PVB 9392, near Steynsburg, South Africa, 17 Dec. 2002 (© PVB).

Fig. 2.213.  Euphorbia pulvinata, big mound ± 1.5 m diam., Tarsus, near Stutterheim, South Africa, 18 Jan. 1992 (© PVB).

2  Euphorbia subg. Athymalus

or third tubercle giving rise to a spine. Each spine starts off among the leaves as a soft, pinkish short-shoot with a few minute bracts along it. Within one centimetre of the apex of the branch, this shoot has hardened off to become needlesharp and rigid and remains like this until it wears away on the lower, woody parts of the branches. Although the spines offer the plant some protection, the tips of the branches are nevertheless frequently chewed off by grazing antelope (Fig. 1.54). The firm, slightly fleshy leaves are present around the apex of the branch during the growing season. In the southern part of its distribution (the Eastern Cape), they are usually between 5 and 10  mm long but they often reach 20–40  mm long in plants from the former Natal and Transvaal and they are by far the largest in ser. Meleuphorbia. In the Eastern Cape they rarely last further than the upper 10  mm of the branch, as growth of the branches is slow. However, in hotter and wetter regions of the former Natal and Transvaal they are present in the upper 20–40  mm of each branch and persist for most of the growing season (as in Fig. 2.210). Flowers are produced between July and early September (though flowering in cultivation, while rare, may take place until March or April). Typically, the cyathium in E. pulvinata is purple-red or green suffused with purple, with purple-red bracts, lobes and small, dark purple-red glands. Clusters of the cyathia around the tops of the branches contrast strongly against the bright green of the tops of the branches and their dark colour also contrasts strongly with the white hairiness of the bracteoles and the yellow anthers in the male cyathia. The female cyathia (typically narrower than the males) are also

2.1 Sect. Anthacanthae

181

Fig. 2.214.  Euphorbia pulvinata, PVB 11893, female cyathia, between Boyne and Haenertzburg, South Africa, 21 Jan. 2013 (© PVB).

Fig. 2.215.  Euphorbia pulvinata, male cyathia (plant with scale-insect infestation on its branches), PVB 11056, west of Witmos, south of Cradock, South Africa, 4 Aug. 2011 (© PVB).

dark purple-red and the styles are usually also somewhat reddish. In plants from the northern parts of its distribution with much larger leaves, the bracts around the cyathia are also larger, partly hiding the cyathia. In these plants there is also considerably less difference in size between the male and female cyathia, the female florets are much more robust and the cyathial glands are smooth (and not pitted). In the Eastern Cape, fruit are present until early summer (November), usually in small clusters around the apices of the branches but they may be present much later further north. Euphorbia pulvinata is usually readily distinguished from other members of ser. Meleuphorbia by its much more

broadly and densely cushion- or mound-forming habit, with the stems tightly interlocking in the mound. Fresh young growth at the tips of the stems is more shiny and bright green than in related species (though bright green is also found in E. pentagona) and the leaf-rudiments are generally larger and persist for longer than in any of the others. In some areas E. ferox and E. pulvinata may be difficult to distinguish. Euphorbia ferox is mostly a plant of arid, fairly low-lying areas well to the south and west of where E. pulvinata occurs. Nevertheless, in the mountains towards the eastern side of the drier areas where E. ferox occurs, plants have been recorded which have similar, fairly slender stems

182

2  Euphorbia subg. Athymalus

Fig. 2.216.  Euphorbia pulvinata, capsules, PVB 11056, west of Witmos, south of Cradock, South Africa, 26 Oct. 2012 (© PVB).

(though variable in this respect too) with rounded angles but they may occasionally form very much larger and very tightly packed, low mounds (with some few-stemmed, diffusely branched specimens among them). In some of these the stems are 8–12 angled (on the low side for E. ferox and on the high side for E. pulvinata), the spines vary from 50 mm long on some plants to only 10–20 mm long on others (rather short for E. ferox and more like E. pulvinata) and the stems are grey-green to bright shiny green (i.e. varying between the typical colours for E. ferox and E. pulvinata (Fig. 2.218)). They often have rounded angles more like E.

ferox and the apex of each branch is sometimes depressed, sometimes flat or slightly protruding. In some plants the branches are quite tall and loosely packed (typical of E. ferox), while in others the plant is cushion-like and the branches are short and very densely packed (as in E. pulvinata). The densely packed plants are generally also those with fewer, sharper angles, shorter spines and a brighter green colour and are more similar to E. pulvinata. The leafrudiments may also be rather longer than one usually finds in E. ferox, but shorter than those typical of E. pulvinata. Florally these plants are also intermediate. In some popula-

Fig. 2.217.  Euphorbia pulvinata. A, side view of male cyathium (scale 2 mm, as for B, C). B, C, side view of female cyathium. D, anther and bracteoles (scale 1 mm, as for E, F). E, F, female floret. Drawn from: A, B, D, E, PVB 11056, west of Witmos, south of Cradock, South Africa. C, F, PVB 11893, between Boyne and Haenertzburg, South Africa (© PVB).

2.1 Sect. Anthacanthae

183

Fig. 2.218.  Euphorbia pulvinata-ferox intermediate, mound ± 1 m diam., PVB 11030, near Fauresmith, South Africa, 26 Apr. 2008 (© PVB).

tions some specimens have purple-red cyathia while in others they are green. Many of these specimens make quite a show when in flower, with especially large numbers of cyathia on most of the branches. Such plants are especially common in the mountains north-west of Cradock and between Colesberg and Fauresmith (see Figs.  2.101 and 2.103). These have often been referred to as E. aggregata, a view that was especially propagated by White et al. (1941), though Berger left no type specimen of E. aggregata and this name is therefore difficult to apply with certainty to any known plants. History The first record of E. pulvinata was made by Ernest E. Galpin in November 1898 near Queenstown, where it was ‘common on the sides and summits of mountains’. Another early collection was made by Burtt-Davy in July 1908. It was collected near Queenstown again by Marloth and he based his description on this collection of which, however, no preserved material has been located. Euphorbia stellispina Haw., Philos. Mag. Ann. Chem. 1: 275 (1827) as ‘stellaespina’. Lectotype (Bruyns 2012): South Africa, Cape of Good Hope, received 1822, Bowie (painting number 803/324 by T. Duncanson at K). Euphorbia stellispina var. astrispina (N.E.Br.) A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 716 (1941). Euphorbia astrispina N.E.Br., Fl. Cap. 5 (2): 355 (1915). Type: South Africa, Beaufort West distr., Willowmore side, Brauns sub Schӧnland 1711 (K, holo.). Unisexual spiny many-branched glabrous succulent shrub 0.3–1 m tall, branching from base (rarely re-branching above)

of similar stem with fibrous roots around base of stem and on bases of some branches. Branches erect, clavate, 0.1–5 (1) m × 30–100 mm, simple to rarely rebranched, covered with tubercles, grey-green (sometimes horizontally marked with darker green or purple bands) towards tips and becoming covered with grey to brown bark below; tubercles conical, 3–7  mm long with transversely rectangular to hexagonal base, vertically fused into 8–18 low and rounded to tall and acute angles, with broad and shallow to deeply V-shaped grooves between angles and fine horizontal lines separating tubercles, with solitary branched spines in their axils; spines near apex of persistent fertile or sterile short-shoots, shortshoots straight to slightly ascending, rigid, (2) 4–20 × 1.5–3 mm, with several small alternating apically pubescent leaflike bracts with ciliate margins to 1.5 × 2 mm along length and in whorl of 3–6 immediately beneath cyathium (if present) in each of whose axils a single spreading to slightly ascending rigid spinelet 4–20 × 1–1.5  mm develops giving rise to whorl of 3–6 spines around apex of short-shoot, initially green to pink or red and finely pubescent, drying to brown or grey; leaf-rudiments on tips of new tubercles towards apices of branches and stem, 3–10 × 1–3 mm, ascending then recurved, fleeting, narrowly lanceolate, acute, sessile. Synflorescences many per branch towards apex, each in axil of tubercle on peduncle with 1 terminal unisexual cyathium and whorl of sterile spines subtending cyathium; cyathia cupular, finely pubescent outside, 3.5–5 mm broad, with 5 externally pubescent lobes with crenulate margins, green to red; glands 5, transversely elliptic, 1.3–1.5  mm broad, widely separated, deep green to purple, slightly depressed in centre, outer margins entire; stamens with pubescent pedicels, bracteoles filiform and sparsely pubescent; ovary globose, finely pubescent, raised on short pedicel

184

±  0.5  mm long; styles ±  2  mm long, divided to just below middle. Capsule 5–7 mm diam., very obscurely 3-angled and almost spherical, finely pubescent, grey-green to purplish, sessile. Distribution & Habitat Euphorbia stellispina inhabits the Great Karoo and is especially associated with doleritic outcrops of the Karoo series.

2  Euphorbia subg. Athymalus

It is of very wide occurrence, on the southern parts of the Great Karoo from Laingsburg to Beaufort West and eastwards to near Willowmore. Further north it has been recorded from the eastern slopes of the Hantam Mountains (north-east of Calvinia), to the hills around Williston and eastwards to Hopetown and Colesberg. In the southern parts of its distribution, E. stellispina grows in stony flat areas and on low ridges with small bushes,

Fig. 2.219.  Distribution of Euphorbia stellispina (© PVB).

occasional shrubs of Rhigozum obovatum and sometimes a wide selection of other succulents. They are also found on steeper hillsides strewn with doleritic boulders. Although Fig. 2.220.  Euphorbia stellispina, ± 30 cm tall, PVB 9990, north of Laingsburg, South Africa, 30 May 2005 (© PVB).

young plants are mostly encountered sheltering in a bush or alongside a rock, many of the larger and older specimens grow fully in the open and are often the largest shrubs pres-

2.1 Sect. Anthacanthae

ent in the area. In the north-easterly parts of its distribution (around Colesberg), where frosts are severe, E. stellispina occurs in isolated patches on steep, north-eastern aspects of rugged hills densely strewn with doleritic boulders. The sur-

185

rounding vegetation is denser here, with many grasses and scattered trees and a few other succulent species (such as Aloe broomii) are present.

Fig. 2.221.  Euphorbia stellispina, ± 1 m diam., PVB 11638, NW of Colesberg, South Africa, 9 Jan. 2012 (© PVB).

Diagnostic Features & Relationships Plants of E. stellispina form many-stemmed, spiky shrubs that superficially bear a strong resemblance to a cactus. In the southern part of its distribution, plants are rarely more than 450  mm tall and broad. However, around Colesberg they develop into massive, seemingly impenetrable mounds

Fig. 2.222.  Euphorbia stellispina, ± 30 cm diam., PVB 12679, SW of Fraserburg, South Africa, 29 Mar. 2014 (© PVB).

up to 1 m high and 1.5 m in diameter. Its very stout, manyangled branches are heavily armed with a covering of persistent, spinescent peduncles, making closer approach to the surface of the branches nearly impossible. Nevertheless, the branches are very flimsily attached at their bases (often being held in position in the clump partly by the network of spines)

186

and they often root when in contact with the ground, readily doing so if detached and brought into cultivation. They have a remarkable number of angles (up to 18  in many plants), which are low, broad and rounded in the southern part of the

2  Euphorbia subg. Athymalus

distribution and taller, thinner and more acute in the northeastern areas. In plants with sharp angles, the branches may also be horizontally mottled with dark green to purple.

Fig. 2.223.  Euphorbia stellispina, male plant ± 30 cm tall, PVB 11102, NE of Laingsburg, South Africa, 2 Aug. 2011 (© PVB).

In E. stellispina the spines develop from around the apex of the short-shoots (both fertile and sterile). They begin to appear when the plant is only 25 mm tall, so that even small specimens have their apex well protected by a crown of spikes. The short-shoot is quite stout and is of variable length. The longest short-shoots are usually produced in a single whorl at the beginning of the summer and appear always to be sterile. They are often over twice as long as the fertile short-shoots, which form a broader band between successive whorls of sterile short-shoots. Usually the fertile short-shoots are at least 10  mm long, but in material from around Colesberg they may be only 4 mm long or less and are partly hidden among the tubercles. If the early stages of the development of the spine-complex are observed, one will see that each short-shoot has several small bracts towards the base, each of which arises on a slight tubercle. Towards the apex of the short-shoot (and just beneath the base of the cyathium, if it is present) the bracts are longer and are collected into what appears to be a whorl. A spine develops in the axil of each of these longer bracts. Initially each spine is soft, green to red and finely pubescent towards the tip and it rapidly grows out to its full length. It remains soft until fully grown but, after that, it soon hardens into a rigid, grey to brown spike. None of the spikes bear any bracts, nor do they produce a cyathium at their tip. After the cyathium dries off

and falls away, the spikes remain at the tip of the short-shoot and persist for several years. In Euphorbia stellispina flowering usually takes place between late July and early September to mid-October. In good years, a roughly circular area up to about 50 mm in diameter at the top of many of the branches becomes covered with the sweetly scented cyathia. Flowering usually reveals that male plants are more common than females in many populations and the males are more conspicuous when in flower. This is brought about by their bearing larger numbers of cyathia (usually densely covering the tops of the branches), by the cyathia being larger and by the paler lobes and anthers in each cyathium. The almost completely round, finely pubescent capsules soon develop on female plants in a dense cluster at the tips of the branches and the seeds are mostly released by the end of October or mid-November. It has been reported (John Bell, pers. comm. 2008) that the tips of the branches of E. stellispina may be chewed away by eland. This leads to the suspicion that these buck (and possibly black rhinos as well) may, on occasion, consume the nearly ripe capsules along with the rest of the branch’s tip, later excreting the capsules and seeds again. This may lead to wider dispersal of the seeds than can be achieved by the explosion of the capsules and may have brought about its remarkably wide and disjointed distribution across the Great Karoo.

2.1 Sect. Anthacanthae

187

Fig. 2.224.  Euphorbia stellispina, male cyathia, PVB 11102, NE of Laingsburg, South Africa, 2 Aug. 2011 (© PVB).

Fig. 2.225.  Euphorbia stellispina, female cyathia, PVB 11102, NE of Laingsburg, South Africa, 2 Aug. 2011 (© PVB).

The spinescent peduncles, characteristic of many species of ser. Meleuphorbia, are simple in all of them except E. pentagona, E. pillansii and E. stellispina. Euphorbia stellispina is unique in that the fertile peduncle is terminated by a cyathium around which a whorl of hard, sterile spikes develops. In contrast to this, in both E. pentagona and E. pillansii the branches of the peduncle are fertile too. In E.

stellispina the spikes at the apex of the peduncle spread away from their epicentre and together make an almost impenetrable network over the stem. Their slightly ascending orientation around the capsule makes access to the capsule difficult as well, adding to its protection from predation until ripe. The cyathium in E. stellispina is much more rounded towards the base than in E. polygona (and most

Fig. 2.226.  Euphorbia stellispina, capsules, PVB 11102, NE of Laingsburg, South Africa, 2 Aug. 2011 (© PVB).

similar to the shape seen in E. clandestina) and the peduncle is both thicker and more clearly tuberculate in E. stellispina. History Euphorbia stellispina was described in 1827 by Haworth from plants in cultivation at Kew. Although paintings were

made by Thomas Duncanson of the specimens that Haworth saw, no material was preserved. Haworth mentioned that the material he saw came from ‘C.B.S.’ i.e. the Cape of Good Hope and on Duncanson’s painting (Fig. 2.228) it was mentioned that the plants were brought to Kew by James Bowie. He may have collected this in the spring of 1821, when he travelled into some of the drier parts of the Karoo.

188

2  Euphorbia subg. Athymalus

Fig. 2.227.  Euphorbia stellispina. A, B, side view of male cyathium (scale 2 mm, as for B, C). C, side view of female cyathium. D, anthers and bracteoles (scale 1 mm, as for E). E, female floret. Drawn from: A, B, D, PVB 3140, near Rietbron, South Africa. C, E, PVB 9990, north of Laingsburg, South Africa (© PVB).

Fig. 2.228.  Euphorbia stellispina, South Africa, Cape of Good Hope, received 1822, Bowie. Watercolour 803/324 by T. Duncanson, the lectotype of E. stellispina (© Royal Botanic Gardens, Kew).

Euphorbia stellispina was known more than 100 years before 1827. It was figured (apparently by Heinrich Claudius) in an unpublished manuscript of Jan Commelijn on plants that were ostensibly observed on 25 October 1685 by the expedition of Simon van der Stel to the Copper Mountains of Namaqualand (Wijnands et al. 1996). A similar figure also appeared as Folio 82 in the Codex Witsenii (Wilson et  al. 2002; also Folio 302  in the volume of the Codex witsenii known as Icones plantarum et animalium, Macnae and Davidson 1969). It also appeared in slightly modified form in Plukenet (1692: t. 231, fig. 1) and Petiver (1709–11: t. 90, fig. 2), though Croizat (1934: 78) believed that these represented the Moroccan E. officinarum. Haworth did not mention any of these early figures when he described E. stellispina. Consequently, White et  al. (1941: 707) were incorrect when they stated that the type locality of E. stellispina was ‘Namaqualand Distr., near Springbok’. In fact, though, E. stellispina does not occur in Namaqualand or near Springbok, coming closest to it in the eastern parts of the Hantam Mountains, to the northeast of Calvinia. Since the expedition to the Copper Mountains did not pass though this area, one must assume that either this plant was brought to the members of the expedition or that it was drawn from a specimen that had been obtained elsewhere and was already in cultivation in Cape Town.

2.1 Sect. Anthacanthae

Euphorbia susannae Marloth, S.  African Gard. 19: 191 (1929). Lectotype (Bruyns 2012): South Africa, Cape, Phisantefontein, 1500’, fl. 1 Oct. 1923, Muir 2762 (BOL; BOL, K, PRE, iso.). Unisexual spineless few- to many-branched glabrous dwarf succulent, mainly subterranean and with only tips of stem and branches in rosette flush with surface and occasionally protruding 10–30 mm above ground, 50–200 mm broad, with age branching from upper parts of swollen fusiform stem 20–60 mm thick gradually tapering below into long tap root from which fibrous roots arise. Branches 5–10, erect, mostly subterranean (upper surfaces often flush with surface of ground), globose to shortly clavate and tapering below into narrow join to stem, mostly simple, 12- to 16-angled, 20–50 × 10–35 mm, wholly covered with prominent tubercles, glabrous and smooth, pale to dark green; tubercles acutely conical, slender and often recurved, 4–8  mm long with rectangular to hexagonal base, vertically arranged and slightly fused into 12–16 acute angles, with shallow vertical grooves and lines between angles and without fine horizontal lines separating tubercles; spines absent; leaf-rudiments on tips of new tubercles at apices of branches, 2–5 × < 1 mm, erect, fleeting, filiform, acute, sessile. Synflorescences many per branch towards apex, usually solitary in axil of tubercle on green peduncle 1–4  mm long (usually longer in males than females) with 1 terminal cyathium and in males sometimes with 2–3 more developing from axils of bracts beneath terminal cyathium, glabrous, with 2–3 oblanceolate acuminate apically toothed glabrous bracts (with marginal cilia)

Fig. 2.229.  Distribution of Euphorbia susannae (© PVB).

189

1.5–2  mm long beneath terminal cyathium and 1–2 bracts further down along peduncle; cyathia cupular, glabrous, 1.5– 2.5 × 2–5 mm (almost twice as broad in male as in female), with 5 lobes with deeply toothed margins, green; glands 5, transversely elliptic, 1–1.5  mm broad, erect to spreading, slightly to widely separated, green to brown-green often with darker margins, slightly depressed in centre, outer margins entire, stamens with glabrous pedicels, bracteoles filiform and sparsely pubescent along margins; ovary globose, glabrous, sessile; styles 1.5–2 mm long, branched and spreading to middle and recurved only near apex, glabrous. Capsule 2–4  mm diam., obtusely 3-angled, glabrous, grey-green to red, sessile. Distribution & Habitat Euphorbia susannae is of restricted occurrence on the Klein Karoo along the base of the northern slopes of the Langeberg, between the eastern foot of the Warmwaterberg and the vicinity of Muiskraal. Euphorbia susannae usually grows in flat to gently sloping spots in hard loam overlaying shale among low shrublets and patches of quartz- or shale-gravel. White et  al. (1941) also showed a plant among shale stones on a hillside near Barrydale, but this is unusual and a rather atypical habitat. Plants may grow in the open or in the shelter of low shrublets and, whether in the open or sheltered, they are generally inconspicuous. Many of the habitats along the northern foot of the Langeberg from Barrydale to near Muiskraal receive substantial amounts of fine, mist-like rainfall from cloud coming over the mountains. Euphorbia susannae and several

190

highly succulent species of Gibbaeum of the Aizoaceae are only found here and disappear as soon as one progresses beyond the range of this mist-like precipitation. Diagnostic Features & Relationships With its stem and branches usually wholly sunken into the ground and only showing their tips above it, E. susannae is a

2  Euphorbia subg. Athymalus

very characteristic species. Plants may become quite large and specimens over 150 mm in diameter are not uncommon. In the centre of each plant is a thick stem which tapers off gradually into a slender tap-root. The upper surface of this stem is flattened and flush with the surface of the ground, though it begins life above it. Around the margins of the slightly flattened upper surface, just below where it descends

Fig. 2.230.  Euphorbia susannae, mature female plant ± 20 cm broad, PVB 11236, alongside BarrydaleLadismith road, South Africa, 5 Oct. 2008 (© PVB).

Fig. 2.231.  Euphorbia susannae, two plants among quartz-pebbles, with branches sunken into the ground and only their tips showing, PVB 11396, east of Lemoenhoek, Barrydale-Muiskraal road, South Africa, 23 Jun. 2009 (© PVB).

into the ground, a rosette of branches develops. Each branch is compressed-spherical and also slightly flattened above, with a conical and very short join to the stem. The uppermost part of each branch is usually also roughly flush with the surface of the ground. Both the upper surfaces of the branches

and of the stem remain visible for the whole of the life of the plant. Most of the plant is covered with prominent, slender and tapering tubercles which are vertically arranged (sometimes only somewhat vaguely) into rows and are fused together vertically into angles mainly towards their bases, if

2.1 Sect. Anthacanthae

at all. Each tubercle is tipped initially with a slender, hairlike leaf-rudiment which soon falls off. The exposed and visible surface of the plant is uniformly pale to dark green, becoming yellow if very exposed to the sun.

191

Euphorbia susannae flowers from April to June. The cyathia are unisexual and are considerably larger in the males than in the females. In both sexes the cyathia arise on short peduncles, though the peduncles are shorter in the females than in the males so that, whereas the male cyathia project

Fig. 2.232.  Euphorbia susannae, young plant with Senecio radicans, Phisantefontein, alongside Barrydale-Ladismith road, South Africa, 30 Nov. 2006 (© PVB).

well beyond the tubercles, the females are shorter than and remain among the tubercles. Female cyathia are usually also solitary, whereas there may be up to four male cyathia per peduncle. The males are also brighter yellow from their pollen and emit a strong, honey-like sweet scent, while the females are green and foetid or apparently scentless and match the colour of the branches and stem closely. In habitat

Fig. 2.233.  Euphorbia susannae, male cyathia, PVB 11236, alongside BarrydaleLadismith road, South Africa, 12 Apr. 2012 (© PVB).

the mature capsules are very dark green, almost black, but this is not usually found in cultivated specimens. Euphorbia susannae is closely related to E. heptagona and E. pseudoglobosa (Bruyns et al. 2006). It is obviously very different in appearance from E. heptagona and its appearance and habit are more similar to that of E. pseudoglobosa subsp. pseudoglobosa, which occurs in the same general area as

192

2  Euphorbia subg. Athymalus

Fig. 2.234.  Euphorbia susannae, female plant with capsules, sunken in ground among quartz-pebbles, PVB 11396, east of Lemoenhoek, Barrydale-Muiskraal road, South Africa, 23 Jun. 2009 (© PVB).

Fig. 2.235.  Euphorbia susannae, capsules, PVB 11396, east of Lemoenhoek, Barrydale-Muiskraal road, South Africa, 23 Jun. 2009 (© PVB).

well, though it has never been recorded together with E. susannae. The branches in E. susannae are generally almost spherical, so are much shorter relative to their breadth than those of subsp. pseudoglobosa, where they are more ellipsoidal. While they still taper beneath to where they join with the stem in E. susannae, the very slender neck without tubercles that is mostly present in subsp. pseudoglobosa is absent in E. susannae. On the stem and the branches in E. susannae the tubercles are arranged into far more and closely packed rows than those of subsp. pseudoglobosa. Consequently, in E. susannae the whole upper surface of the plant is densely covered with tubercles, while in subsp. pseudoglobosa the tubercles are confined to very clear and well-defined angles,

which are widely spaced out around the branch. In subsp. pseudoglobosa the tubercles are vertically fused into clear and well-defined angles, while in E. susannae, there is no significant vertical fusion among the tubercles and the angles are essentially absent. In addition, the tubercles are much longer and more numerous in E. susannae than in subsp. pseudoglobosa, where they are smaller and far fewer. The slender, almost filiform leaf-rudiments may be very similar in both though they can be short and deltate in subsp. pseudoglobosa. Both subsp. pseudoglobosa and E. susannae flower during late autumn and early winter. In both, the cyathia are small and unisexual, with the female cyathia almost half the diameter of the males and quite without any rudimentary

2.1 Sect. Anthacanthae

193

Fig. 2.236.  Euphorbia susannae. A, side view of male cyathium (scale 2 mm, as for B). B, side view of female cyathium. C, anthers and bracteoles (scale 1 mm, as for D). D, female floret. Drawn from: PVB 11236, alongside Barrydale-Ladismith road, South Africa (© PVB).

male florets. In E. susannae the cyathia are glabrous, while those of subsp. pseudoglobosa may be densely pubescent but both have very similarly shaped bracts just beneath the cyathia. In E. susannae the peduncles on female plants are mostly quite a bit shorter than the tubercles and so the female cyathia are partly hidden among the tubercles. This is not the case in subsp. pseudoglobosa. History Euphorbia susannae was first collected by John Muir sometime before 1 October 1923 and, presumably on being prompted by him, Marloth collected it again in November 1929, ostensibly at Lemoenshoek and named it after Muir’s wife, Susanna. The species is widespread in cultivation but has rarely been recorded in habitat. Euphorbia tubiglans Marloth ex R.A.Dyer, Bull. Misc. Inform. 1934: 268 (1934). Type: South Africa, Cape, near Steytlerville, Aug. 1929, Herre 1596 (PRE, holo.). Euphorbia jansenvillensis Nel, Jahrb. Deutsch. KakteenGes. 1: 32 (1935). Euphorbia tubiglans var. jansenvillensis (Nel) Schnabel, van Veldh. & J.G.Marx, Euphorbia World 13: 8 (2017). Type: South Africa, Cape, near Jansenville, Apr. 1932, Le Roux sub SUG 6550 (missing). Neotype (Bruyns 2012): South Africa, Cape, 1.5 miles east of Jansenville, Aug. 1939, Dyer 4012 (PRE; K, iso.).

Unisexual spineless few- to many-branched dwarf glabrous succulent, 20–150 (300) × 30–150 (500) mm, branching and often rhizomatous from around swollen apex of often 4- to 6-angled fusiform wholly subterranean stem 20–50 mm thick tapering below into tap root from which fibrous roots arise. Branches 1–10 (–50), erect to spreading, sometimes with subterranean and slender spreading base then abruptly thickening and protruding above ground otherwise wholly above ground, clavate to cylindrical and tapering below into narrow cylindrical neck 10–50 (150) mm long above stem, simple to forming branches in same way as stem, 4- to 5 (7)-angled, 20–300 × 10–30 mm, partly covered with obscure tubercles, smooth, glaucous green to dark green; tubercles conical to obtusely conical, 2–3 mm long with obscure base, vertically fused into 4–5 (7) prominent angles, with shallow vertical grooves between angles (sometimes only near apex); leaf-rudiments on tips of new tubercles towards apices of branches, 2–5 × 1–2 mm, spreading, fleeting, linear to linearsubulate, acute, sessile, sometimes with small tooth-like stipular outgrowths just below bases. Synflorescences several per branch towards apex, each in axil of tubercle on solitary red to green sometimes branched peduncle 1–4  mm long (usually longer in males than females) with 1 terminal cyathium, glabrous, with 3–5 broadly obovate reddish or greyish green glabrous apically finely toothed bracts (often with marginal cilia) 3–6 × 1–3 mm in rosette beneath terminal cyathium and 1–3 much smaller bracts further down

194

along peduncle; cyathia shallowly bowl-shaped or cupshaped, glabrous, 1.5–3 × 3.5–5 mm (almost twice as broad in male as in female), with 5 faintly reddish lobes with deeply toothed margins, greyish to reddish green; glands 5, almost circular, well separated, 0.5–1 mm broad, ascending, reddish green, initially flat then becoming longitudinally folded to form almost conical structure, outer margins entire to slightly crisped, stamens with finely and densely pubescent pedicels, bracteoles filiform and coarsely pubescent; ovary globose, glabrous, raised on very short pedicel < 0.2 mm long; styles

2  Euphorbia subg. Athymalus

2–3  mm long, branched except in lower third and widely spreading, glabrous. Capsule 4–6  mm diam., obtusely 3-angled, glabrous, dull grey-green to slightly reddish, sessile. Distribution & Habitat Euphorbia tubiglans is found in the dry parts of the Eastern Cape, around the foot of the Groot River Mountains from Kleinpoort and Steytlerville to south of Klipplaat and also in the hills around the town of Jansenville.

Fig. 2.237.  Distribution of Euphorbia tubiglans (© PVB).

Along the foot of the Groot River Mountains, plants are usually found wedged among stones on low, rocky outcrops (on tillites with vertically oriented strata just south of Klipplaat, sandstones in Kleinpoort, sandstone pebbles and other alluvial gravel around Steytlerville) with a short and very scanty cover of bushes. Around Jansenville, they grow on stony, gently sloping areas with scanty soils derived from shales, sheltering around the bases of shrubs of Rhigozum obovatum and more rarely under shrubs of the ubiquitous Euphorbia radyeri. Diagnostic Features & Relationships Plants of E. tubiglans vary greatly in size. Those around Klipplaat and Steytlerville may be especially small, with only a few short, branches up to 20  mm long protruding above the surface of the ground, with a larger, swollen stem beneath it. However, even here there is great variation in the length of the branches and, when sheltered, they may reach

20 cm long. Around Jansenville, plants may also be small, but some form clumps of up to 50 or more branches and may reach 0.5 m in diameter. The branches here may reach 30  cm tall, but they mostly remain inconspicuous, since they are always well hidden around the bases of larger shrubs. The stem usually has tubercles and angles towards its apex (which may occasionally even protrude from the soil in mature plants) and it is from among the axils of these tubercles that the branches develop. Around Klipplaat and Steytlerville the branches are erect on very short, slender bases attached to the stem. Around Jansenville this slender attachment may spread beneath the ground as a rhizome up to 10 cm long before the branch emerges from the soil. The branches may themselves give rise to several similar secondary branches around their tips. Roots develop readily on the lower, slender parts of the branches, which are able to become detached from the rest of the plant to form new and independent plants. Most of the branches are 5-angled, but

Fig. 2.238.  Euphorbia tubiglans, rhizomatous plant among branches of protecting shrub, longest branch ± 12 cm tall, PVB 11522, just east of Jansenville, South Africa, 3 Sep. 2015 (© PVB).

Fig. 2.239.  Euphorbia tubiglans, exposed plant with longer branches around 6 cm long, alongside a tuft of grass, PVB 12977, east of Jansenville, South Africa, 3 Sep. 2015 (© PVB).

Fig. 2.240.  Euphorbia tubiglans, plant among stones and with branches about 4 cm tall, PVB 12016, SE of Steytlerville towards Groot River Poort, South Africa, 6 Oct. 2011 (© PVB).

196

2  Euphorbia subg. Athymalus

Fig. 2.241.  Euphorbia tubiglans, longest branch ± 3 cm tall, rhizomatous plant among stones on small rocky outcrop, east of Steytlerville, South Africa, 13 Apr. 2017 (© PVB).

4-angled ones are regularly encountered (some branches even have 4-angled and 5-angled parts) and there may occasionally be 5- to 7-angled branches on one plant. The margins of these angles are often somewhat hardened as the tips of the tubercles become rounded, grey and slightly corky with age. When they are young, the branches are green and this soon changes (with the development of a waxy coat) to greyish and later brownish, sometimes covered lower down with a corky bark. The tip of each new tubercle bears a small leaf-rudiment which soon dries up and wears away. In plants from around Jansenville, just below the bases of the margins of the leaf-rudiment there are often a few (1–3 on each side) minute, tooth-like stipular rudiments. These are

much rarer in plants from near Klipplaat but may still be present. Flowering in E. tubiglans seems to take place at any time during the summer months when rain is received and plants flower from November until April in cultivation. While relatively few cyathia may sometimes be produced per branch, they may equally well be densely clustered around its apex. Each cyathium arises on a short peduncle with several bracts. These are largest and broadest in a conspicuous rosette around the apex of the peduncle. The larger bracts, which often have somewhat serrated margins, may form an almost calyx-like structure which hides the lower parts of the cyathium, or they may spread widely and

Fig. 2.242.  Euphorbia tubiglans, male cyathia, PVB 11522, just east of Jansenville, South Africa, 3 Sep. 2015 (© PVB).

Fig. 2.243.  Euphorbia tubiglans, male cyathia, PVB 11524, south of Klipplaat, South Africa, 24 Nov. 2011 (© PVB).

2.1 Sect. Anthacanthae

not conceal the cyathium at all. In E. tubiglans, the cyathia are noteable for their fairly small, well separated, green glands, their greyish green colour often with some reddish patches and the peculiar shape of the glands. Around their margins, the glands are entire to finely notched, with notches especially in plants around Jansenville. While the cyathium is still developing (i.e. before the florets emerge)

197

the glands are more or less flat and as the florets mature the glands become folded longitudinally into the almost conical or tubular shape that is typical for this species. Male cyathia are considerably larger than the females and the pollen is a very pale yellow, while the female florets have particularly long, widely spreading styles.

Fig. 2.244.  Euphorbia tubiglans, female cyathia with developing capsules, PVB 11524, south of Klipplaat, South Africa, 13 Nov. 2012 (© PVB).

Closely related to E. heptagona and allies, E. tubiglans is easily separated from most of them by its lowly and somewhat rhizomatous habit and the lack of spines. Some of its smaller forms (as in Fig. 2.240) look superficially similar vegetatively to E. pseudoglobosa subsp. pseudoglobosa. However, a closer examination and comparison reveals that they may be separated by the broader leaf-rudiments in E. tubiglans, which also has much larger bracts subtending the cyathia, the much larger cyathia (both male and female) and, also, by the much larger female florets. They differ also markedly in the shapes of their respective cyathial glands. Euphorbia jansenvillensis was said to differ from E. tubiglans by its much larger and more branched habit, by the fact that the cyathia are not present at the tip of the branches and by the lack of a deep-seated, thickened, tuberous rootstock (Nel 1935). The plants indeed mostly have much longer branches around Jansenville, but occasional, exposed individuals may have very short branches which are no different in length from those typical of E. tubiglans. In neither ‘species’ do the cyathia actually appear at the apex of the branch,

so this does not constitute a difference between them, though they may arise further away from the tip in material from Jansenville. Finally, plants from Jansenville do possess a central, often buried stem from which the branches arise and this may be somewhat thickened too, so the absence or presence of this stem cannot be used to separate them. Although White et al. (1941) maintained these two as separate species, there seems to be little doubt that they are the same, especially since Dyer (1937b) also included the smaller plants from Klipplaat under E. tubiglans. History Euphorbia tubiglans was first collected by H.  Herre in August 1929 near Steytlerville. Herre’s plants flowered at Stellenbosch University and in Marloth’s garden in January 1931. For these plants, Marloth proposed the name E. tubiglans, on account of the distinctive shape of the glands, but he died before sending this description off to press. Mrs Marloth passed on his notes to I.B. Pole-Evans and subsequently to Dyer, who eventually published it in September

2  Euphorbia subg. Athymalus

198

Fig. 2.245.  Euphorbia tubiglans. A, B, side view of male cyathium (scale 2 mm, as for B, C). C, D, side view of female cyathium (scale of D 1 mm, as for E–G). E, anthers and bracteoles. F, G, female floret (G in dissected cyathium with bracteoles). Drawn from: A, C, E, F, PVB 11522, just east of Jansenville, South Africa. B, PVB 11524, south of Klipplaat, South Africa. D, G, PVB 4913, Steytlerville, South Africa (© PVB).

1934 (wrongly given as 1935  in Bruyns 2012). Marloth seems to have had a rather muddled concept of E. tubiglans, since he included several collections of superficially similar plants that originated between Swellendam and Robertson in the Western Cape, which were quoted by Dyer (1934a) and later again by White et al. (1941). These should be referred to E. pseudoglobosa rather than to E. tubiglans, which is not known west of Klipplaat in the Eastern Cape. In October 1935, G.C. Nel described Euphorbia jansenvillensis in the fourth monthly instalment of the ‘Jahrbuch der deutschen Kakteen-Gesellsschaft’, of which the first instalment appeared in July 1935. He based this description on plants collected in April 1932 on the eastern outskirts of Jansenville by the pastor J.A. van Zyl le Roux, which flowered in the Botanical Garden at Stellenbosch University in November 1934. In 1939, Dyer collected E. tubiglans about 1.5 miles east of Jansenville and, despite considerable expansion of the town, it still occurs on both its eastern and its southern outskirts. A colony of plants a little further to the east has been

kept track of by the family Taljaard for at least the last 80 years (Gert Taljaard, pers. comm., 2015) and also continues to thrive.

2.1.2.3 Ser. Rhizanthium Euphorbia ser. Rhizanthium (Boiss.) Bruyns, Taxon 62: 1194 (2013). Euphorbia sect. Rhizanthium Boiss. in DC., Prodr. 15 (2): 10, 92 (1862). Euphorbia subsect. Rhizanthium (Boiss.) Denis in Euphorb. Iles Austr. Afr. 41, 66 (1921). Type (designated by Wheeler 1943): E. tuberosa L. Unisexual geophytes, with scarcely or not at all tuberculate subterranean stems (and branches in E. tuberosa), without persistent short-shoots or spines. Leaves soft, conspicuous, petiolate, deciduous and persisting for one growing season, often pressed to the ground. Cyathia unisexual, surrounded by conspicuous bracts.

2.1 Sect. Anthacanthae

Three species belong to this series and are restricted to the temperate parts of South Africa, where they are mainly associated with the winter-rainfall region of the south-west of South Africa. Only E. silenifolia extends outside the

199

Greater Cape Floristic Region into the Eastern Cape and may have leaves during summer. The other two are leafless in summer, growing and flowering during autumn, winter and spring.

Fig. 2.246.  Distribution of Euphorbia ser. Rhizanthium (showing number of species per half-degree square) (© PVB).

Key to the species of ser. Rhizanthium 1. Leaves tightly pressed to ground in flat rosette, dark green and slightly shiny above with impressed main and side veins, margins of leaves slightly recurved....................................................................................................E. ecklonii 1. Leaves erect to spreading and not pressed to ground in flat rosette, grey-green, margins of leaves slightly folded upwards........2. 2. Tuber usually elongated and frequently branched, leaves abruptly changing from blade into petiole, margins often undulating to strongly crisped.......................................................................................................................E. tuberosa 2. Tuber often short and turnip-shaped and usually unbranched, leaves gradually merging from blade into petiole, margins neither undulating nor crisped................................................................................................E. silenifolia

Euphorbia ecklonii (Klotzsch & Garcke) Baill., Adansonia 3: 144 (1863). Tithymalus ecklonii Klotzsch & Garcke, Abh. Königl. Akad. Wiss. Berlin 1859: 68 (1860). Type: South Africa, Cape of Good Hope, Swellendam district, Breede River at Swellendam (70.10), hills under 1000’, Aug., Ecklon & Zeyher Euphorb. 16 (W, holo.; MEL, P, S, TCD, iso.). Euphorbia pistiifolia Boiss. in DC., Prodr. 15 (2): 93 (1862). Lectotype (Bruyns 2012): South Africa, Cape of Good Hope, Swellendam district, Breede River at Swellendam (70.10), hills under 1000’, Aug., Ecklon & Zeyher Euphorb. 16 (S; MEL, P, TCD, W, iso.). Unisexual spineless usually unbranched glabrous geophyte with subterranean stem and with leaves spreading on

surface of ground in rosette (rosettes very rarely clustered together) 30–150 mm broad, stem swollen fusiform to turnip-shaped, smooth, 30–100 × 10–60 mm, rapidly tapering below into slender tap root from which fibrous roots arise. Branches usually none; leaves in dense cluster at apex of stem, blade 15–50 × 4–40 mm, spreading and tightly pressed to surface of ground in flat rosette, deciduous, narrowly linear to elliptic to nearly circular, acute to obtuse or apically notched, dark green and slightly shiny above with impressed main and side veins (paler green and reticulately veined below, usually initially suffused with red), glabrous to slightly puberulous, tapering towards base to abruptly narrowing into glabrous petiole 8–20 mm long, margins spreading to slightly recurved so that upper surface of leaf often slightly convex, rarely slightly ciliate. Synflorescences many

200

per stem towards apex, usually solitary in axil of leaf on spreading red-green peduncle 10–40 mm long with 2–4 redgreen bracts beneath terminal cyathium and sometimes with 2 more bracts on rays 5–15 mm long developing from axils of bracts beneath terminal cyathium, puberulous; bracts sessile, ovate-lanceolate to obovate, acute to obtuse, puberulous, 2.5–5 × 1.5–5 mm, margins straight to slightly wavy and suffused with red; cyathia cupular, puberulous, 3–5 mm broad (often broader in male than in female), with 5 lobes with deeply toothed margins, dark green; glands 5, transversely elliptic, 1–2.5  mm broad, spreading, ±  contiguous, dark green to brown, outer margins spreading to slightly descending and usually entire, stamens with finely and

2  Euphorbia subg. Athymalus

sparsely pubescent pedicels, bracteoles filiform and pubescent; ovary ellipsoidal, pubescent, raised on short pubescent pedicel; styles 1.5–2  mm long, branched and spreading in upper half and recurved above, puberulous near base. Capsule 4.5–5  mm diam., obtusely 3-angled, pubescent, sessile. Distribution & Habitat Euphorbia ecklonii is only known on the southern coastal plain of the former Cape Province from Caledon to Albertinia and northwards to Cloete’s Pass near Herbertsdale. Along the Breede River it has been recorded from Swellendam to near the mouth, near Infanta.

Fig. 2.247.  Distribution of Euphorbia ecklonii (© PVB).

Plants of E. ecklonii typically grow on gravelly, alluvial deposits (sometimes overlaying shales) on gentle slopes or in flat areas in heavy, fine-grained, loam. It mainly occurs in renosterveld, usually among low bushes and tends to be most common where the bushes are not too dense. There it will often be found with several other succulents such as Acrodon, Glottiphyllum and various species of Crassula and Haworthia as well as E. silenifolia and E. tuberosa. Diagnostic Features & Relationships Euphorbia ecklonii is a distinctive, small geophyte. The plant consists of a single, swollen, subterranean stem that is usually 10–20 mm beneath the surface of the soil, tapering gradually below into a slender tap root and ending more abruptly above in an obtuse apex, from where the leaves

arise. After rains (usually in the months of April to June and often as early as March) leaves develop in a rosette around the apex of the stem. Each leaf develops from a short, erect petiole, from which it spreads out, to become closely pressed to the ground and to any surrounding stones. The leaves in the rosette are placed quite densely around the top of the tuber and cover the surface of the soil completely immediately around the plant. When they first emerge, the leaves and their petioles often have a reddish hue. Once mature the leaves are dark green and slightly glossy, usually with the veins indented into their upper surface (and prominently reticulate on the underside). In the western part of its distribution (from Caledon to Swellendam), the leaves of E. ecklonii are often almost nearly circular, narrowing very abruptly into their petioles. Here, their margins are usually

2.1 Sect. Anthacanthae

201

Fig. 2.248.  Euphorbia ecklonii, typical plant with broad leaves, ± 6 cm broad, PVB 11162, Swellendam, South Africa, 22 Aug. 2008 (© PVB).

Fig. 2.249.  Euphorbia ecklonii, plant with slightly narrower leaves, south of Riversdale, South Africa, 18 May 2012 (© PVB).

slightly recurved so that the upper surface of each leaf is slightly convex. Around Albertinia, the leaves are not so broad, they are usually acute-tipped and are also variable in breadth. On some plants they are subulate while in others they are narrowly linear. Flowering takes place as soon as the leaves are mature which, in years of good rains is usually between April and June. The peduncles emerge and spread somewhat around the apex of the plant and among the upper leaves. They are relatively short, with erect apices and each bears a small unisexual cyathium surrounded by two or more bracts. Further peduncles may develop in the axils of these bracts, again each bearing a single cyathium. The bracts may be

slightly reddish green and this may also extend to the stamens and styles, while the glands are dark green or somewhat brownish green. The cyathia are very sweetly scented. Euphorbia ecklonii grows with E. tuberosa along the banks of the Breede River and the two seem to maintain their distinctness by flowering at different times, with E. ecklonii finished flowering and in fruit by July and August, when E. tuberosa is still in flower. It also occurs together with E. silenifolia between Swellendam and Albertinia and, here, those plants of E. ecklonii with narrow leaves are hard to distinguish from E. silenifolia. Euphorbia silenifolia usually differs by being more grey-green, in having obtuse apices to the

202

2  Euphorbia subg. Athymalus

Fig. 2.250.  Euphorbia ecklonii, plant with still narrower leaves, PVB 12136, west of Albertinia, South Africa, 18 May 2012 (© PVB).

Fig. 2.251.  Euphorbia ecklonii, plant with still narrower leaves, PVB 12136, west of Albertinia, South Africa, 18 May 2012 (© PVB).

Fig. 2.252.  Euphorbia ecklonii, plant with yet narrower leaves, PVB 12136, west of Albertinia, South Africa, 18 May 2012 (© PVB).

leaves and lacking the indented veins on the leaves that are typical of E. ecklonii. Euphorbia ecklonii is closely related to both E. silenifolia and E. tuberosa and the differences between all three of them are discussed further under E. tuberosa.

mens Burchell 6704 and 6678 at Kew). Burchell made three specimens of it but the third (Burchell 8499, also at Kew) has no information about either the date or the place of collection. The material on which the names E. ecklonii and E. pistiifolia (the Euphorbia with pestle-shaped leaves) were based, was collected on hills along the Breede River at Swellendam by Ecklon & Zeyher in about 1825. Another specimen (Drège 8195) was collected in the same area by J.F. Drège and his brother Carl, probably in February 1830.

History Euphorbia ecklonii was first recorded by William Burchell close to the Soetmelks River near Riversdale, on 17 November 1814 (according to annotations on the speci-

2.1 Sect. Anthacanthae

203

Fig. 2.253.  Euphorbia ecklonii, plant with very narrow leaves, PVB 12136, west of Albertinia, South Africa, 18 May 2012 (© PVB).

Fig. 2.254.  Euphorbia ecklonii. A, male cyathium from side (scale 1 mm, as for rest). B, C, female cyathium from side. D) anthers and bracteoles. E, female floret. Drawn from: PVB 11162, Swellendam, South Africa (© PVB).

Euphorbia silenifolia (Haw.) Sweet, Hort. Brit., ed. 1, 2: 356 (1826). Tithymalus silenifolius Haw., Rev. Pl. Succ.: 61 (1821). Lectotype (Bruyns 2012): South Africa, Cape of Good Hope, received 1823, Bowie (no specimen preserved, painting number 810/147 by T. Duncanson at K).

Euphorbia elliptica Thunb., Prodr. Fl. Cap. 2: 86 (1800), nom. illegit., non Lam. (1788). Tithymalus ellipticus (Thunb.) Klotzsch & Garcke, Abh. Königl. Akad. Wiss. Berlin 1859: 69 (1860). Type: South Africa, Cape, Thunberg (UPS-THUNB, holo.).

204

Tithymalus bergii Klotzsch & Garcke, Abh. Königl. Akad. Wiss. Berlin 1859: 68 (1860). Type: South Africa, Cape, Bergius (missing). Tithymalus longipetiolatus Klotzsch & Garcke, Abh. Königl. Akad. Wiss. Berlin 1859: 68 (1860). Type: South Africa, Cape, Bergius (missing). Tithymalus attenuatus Klotzsch & Garcke, Abh. Königl. Akad. Wiss. Berlin 1859: 69 (1860). Type: South Africa, Cape, Bergius (missing). Euphorbia elliptica var. undulata Boiss. in DC., Prodr. 15 (2): 93 (1862). Neotype (Bruyns 2012): South Africa, Cape of Good Hope, received 1823, Bowie (no specimen preserved, painting number 810/147 by T. Duncanson at K). Euphorbia mira L.C.Leach, S. African J. Bot. 52: 10 (1986). Type: South Africa, Cape, near Tulbagh, Bayer sub Leach 17175 (NBG, holo.; K, PRE, iso.). Unisexual spineless usually unbranched glabrous geophyte with subterranean stem and with leaves above surface of ground usually in single rosette 30–150 mm broad, with swollen fusiform smooth stem 30–100 × 10–60 mm rapidly tapering below into slender tap root from which fibrous roots arise. Branches usually none; leaves in cluster at apex of stem, blade 25–100 × 2–12 mm, spreading to erect, deciduous, linear to elliptic-lanceolate, acute, greygreen and slightly glaucous, initially sometimes faintly suffused with red, glabrous, gradually narrowing into glabrous petiole 12–100  mm long, margins usually slightly folded upwards, rarely slightly ciliate. Synflorescences

Fig. 2.255.  Distribution of Euphorbia silenifolia (© PVB).

2  Euphorbia subg. Athymalus

many per stem near apex, solitary in axil of leaf on greygreen peduncle 25–125 mm long, ascending to nearly prostrate, with whorl of 2–4 bracts beneath terminal cyathium and sometimes with 2 more developing on rays to 5–30 mm long from axils of bracts beneath terminal cyathium, puberulous; bracts sessile, obovate, acute to obtuse, glabrous to puberulous, 2.5–10 × 2–6 mm, margins often undulating; cyathia cupular, puberulous, 3.5–6 mm broad (broader in male than in female), with 5 lobes with deeply toothed margins, dark green; glands 5, transversely elliptic, 1.5–3 (4) mm broad, spreading, ±  contiguous, dark green to yellow-green, outer margins spreading and entire to obscurely toothed, stamens with pubescent pedicels, bracteoles filiform and pubescent; ovary globose, pubescent, raised on short pubescent pedicel ±  0.5  mm long; styles 1–1.5 mm long, branched and spreading from near base and strongly recurved above, puberulous at base. Capsule 4.5–6  mm diam., obtusely 3-angled, pubescent, grey-green to red, sessile. Distribution & Habitat Euphorbia silenifolia is known from around Darling, north of Cape Town, to the Cape Peninsula (where it is only known in the northern part from Devil’s Peak to Signal Hill and Camps Bay) and then eastwards to Mossel Bay, Port Elizabeth and East London. Most collections have been made on the southern coastal plain between Hermanus and Mossel Bay. Euphorbia silenifolia usually occurs in gravelly to rocky ground on granites, sandstones and shales (or alluvial depos-

2.1 Sect. Anthacanthae

205

Fig. 2.256.  Euphorbia silenifolia, female plant in growing season after fire, Signal Hill, Cape Town, South Africa, 13 Jul. 2015 (© PVB).

its overlaying shales), on the lower slopes of larger mountains (then often in the zone where sandstones and shales are both present) or on the slopes of hills and sometimes in flat areas. In the west around Cape Town it is often associated with dry, short fynbos or shale-based, karroid vegetation

mixed with renoster. On the southern coastal plain it is mainly associated with renosterveld, while between Port Elizabeth and Grahamstown it occurs in stony grassland.

Fig. 2.257.  Euphorbia silenifolia, male plant (with two younger ones next to it) with very slender leaves, north of Albertinia, South Africa, 18 May 2012 (© PVB).

Diagnostic Features & Relationships Euphorbia silenifolia is an inconspicuous geophyte in which the plant generally consists of a small tuber 10–30  mm beneath the surface, with a rosette of leaves that appears above the ground in the rainy season. The tuber is generally a single, fusiform structure (often shaped like a slender turnip in the

western Cape, though it may be more slender and longer further east) which tapers at its base into a slender tap-root and is more obtuse at its apex. Branching of the tuber seldom occurs, especially in the Western Cape. In the Western Cape the rosette of leaves appears from February and March onwards, often stimulated by the occa-

206

2  Euphorbia subg. Athymalus

Fig. 2.258.  Euphorbia silenifolia, PVB 11937, Baviaanskloof, NW of Patensie, South Africa, 27 Aug. 2011 (© PVB).

Fig. 2.259.  Euphorbia silenifolia, two plants excavated to show tubers, PVB 13737, Arniston, Bredasdorp, South Africa, 5 Jul. 2019 (© PVB).

sional showers well before the rainy season begins and with the onset of slightly cooler weather. In the southern and eastern Cape, they grow at about the same time, which is often well after the rains have begun, but again after the hottest weather is over. Initially the leaves are slender and filiform and one cannot distinguish blade and petiole. Gradually the blade may expand to become narrowly elliptic or narrowly lanceolate, while in some areas these broader blades are only found on young plants (as shown in Fig. 1.8). In many plants it remains slender (as in Fig. 2.257) or with only a slightly dilated tip and some of these particularly slender-leaved specimens were described by Leach as Euphorbia mira. The peduncles begin to develop when the first leaves appear. In the Western Cape this is between February and May, with flowering sometimes continuing into July. The

peduncles are also slender, ascending in a group in the centre of the rosette (often prostrate in recently burnt areas) and they are variable in length: even on one plant they may range from 25 to 80 mm long. Each peduncle is tipped by a whorl of small bracts surrounding the first cyathiium and it often branches to produce further cyathia. The bracts usually have somewhat wavy margins, which may also be tinged with red. Older male plants can become fairly conspicuous, but females are much less conspicuous until the capsules are fully developed. Initially dark greyish green, matching the colour of the leaves, these turn reddish as the leaves dry off (in the process these often also become reddish) and release the seeds in mid- to late October. In its growth-form E. silenifolia is similar to E. ecklonii and E. tuberosa and the differences between them are dis-

Fig. 2.260.  Euphorbia silenifolia, male plant, PVB 12709, Ruitersbos, north of Mossel Bay, South Africa, 23 May 2014 (© PVB).

2.1 Sect. Anthacanthae

207

Fig. 2.261.  Euphorbia silenifolia, female plant, Signal Hill, Cape Town, South Africa, 13 Jul. 2015 (© PVB).

Fig. 2.262.  Euphorbia silenifolia, capsules, Signal Hill, Cape Town, South Africa, 18 Oct. 2015 (© PVB).

cussed under E. tuberosa. Some slender-leaved forms of E. ecklonii in the vicinity of Albertinia are difficult to distinguish from E. silenifolia (see under E. ecklonii).

rect one for this species, though this was based on plants collected by Bowie. Bowie brought this material to Kew in 1823 and a painting was made of some of it by Thomas Duncanson (Fig. 2.264). It is not known where this was collected but it is likely to have been in the south-eastern Cape, between Port Elizabeth and Grahamstown, where Bowie was most active as a collector. It was also collected at Cape Town by Burchell in December 1810 and by Carl Heinrich Bergius, who arrived at the Cape in May 1815.

History Euphorbia silenifolia was first collected by Thunberg at the Cape around 1774. Thunberg named these plants Euphorbia elliptica, but this name had been used before in Euphorbia. Consequently, the epithet ‘silenifolia’ of Haworth is the cor-

Fig. 2.263.  Euphorbia silenifolia. A, side view of cyme of male cyathia (scale 4 mm, as for B). B, side view of cyme of female cyathia. C, female cyathium from above. D, anthers, bracteole and rudimentary female floret from male cyathium. E, female floret. Drawn from: PVB, Signal Hill, Cape Town, South Africa (© PVB).

208

2  Euphorbia subg. Athymalus

Fig. 2.264.  Euphorbia silenifolia, South Africa, Cape of Good Hope, received 1823, Bowie. Watercolour 810/147 by T. Duncanson, the lectotype of Tithymalus silenifolius (© Royal Botanic Gardens, Kew).

Although the name E. mira was maintained as a distinct species in Bruyns et al. (2006), observations of more populations of E. silenifolia have made it clear how E. silenifolia may begin to produce leaves well before winter. It is now clear how narrow the leaves may be in some populations, often mixed up with plants with considerably broader leaves (see Fig. 1.8). Thus, while Leach (1986b: 11) believed he had found three, possibly even four distinct geophytic species of Euphorbia growing together at the type locality of E. mira (near Tulbagh in the Western Cape), it is clear from the photograph (Leach 1986b: fig. 2) and the specimens made that he found the two species, E. silenifolia and E. tuberosa (with several sympatric forms of the former), at this locality. Euphorbia tuberosa L., Sp. Pl. 1: 456 (1753). Tithymalus tuberosus (L.) Haw., Syn. Pl. Succ.: 137 (1812). Lectotype (Bruyns 2012): J. Burm., Rar. Afric. Pl.: 9, t. 4 (1738). Euphorbia crispa (Haw.) Sweet, Hort. Brit., ed. 1, 2: 356 (1826). Tithymalus crispus Haw., Rev. Pl. Succ.: 61 (1821). Type: none located.

Unisexual spineless few- to many-branched mostly glabrous geophyte with subterranean stem and branches and only leaves protruding 10–20 mm above ground, 50–500 mm broad, with age branching from upper parts of swollen fusiform stem 50–150 × 20–60  mm gradually tapering below into tap root from which fibrous roots arise. Branches 5–50, ascending to erect, subterranean, shortly clavate to cylindrical and often tapering below into narrow join to stem, simple to repeatedly branched, 20–150 × 6–40 mm, rounded and with at most very obscure tubercles (and sometimes with persistent remains of petioles towards apex), glabrous, brown; tubercles absent to low and obscure; leaves in rosettes at apex of each branch, blade 15–80 × 5–25 mm, ascendingspreading and sometimes pressed to ground, deciduous, oblong to lanceolate-oblong, obtuse, grey-green, glabrous to puberulous, subcordate to cuneate into often slightly pubescent petiole 10–40 mm long, margins erect often reddish and often undulating to very strongly crisped and ciliate. Synflorescences many per branch towards apex, usually solitary in axil of leaf on grey-green peduncle 10–50 mm long with 2–4 bracts beneath terminal cyathium and sometimes with 2 more cyathia developing on rays to 18 mm long from axils of bracts beneath terminal cyathium, glabrous to pubes-

2.1 Sect. Anthacanthae

cent; bracts sessile, elliptic to obovate, acute to obtuse glabrous or pubescent, 4–8 × 3–6 mm, margins often reddish and very slightly crisped; cyathia cupular, glabrous or pubescent, 4–6 mm broad (as broad in male as in female), with 5 pubescent lobes with deeply toothed margins, dark green; glands 5, transversely elliptic, 1.5–2 mm broad, spreading, ±  contiguous, dark green to yellow, outer margins slightly descending and entire to obscurely toothed, stamens with pubescent pedicels, bracteoles filiform and pubescent; ovary

209

globose, puberulous to tomentose, raised on short pubescent pedicel; styles 2–3.5 mm long, branched and spreading from middle and slightly recurved near apex, glabrous to puberulous. Capsule 6–9 mm diam., obtusely 3-angled, pubescent, grey-green, ± sessile. Distribution & Habitat Euphorbia tuberosa is a widespread species that is known from the coastal sands of Namaqualand (where it is recorded

Fig. 2.265.  Distribution of Euphorbia tuberosa (© PVB).

from as far north as Hondeklip Bay) to the Cape Peninsula, on the Worcester-Robertson Karoo and along the lower northern slopes of the Witteberge to near Matjiesfontein. It is also known along the southern coastal plain from Hermanus in the west to Still Bay in the east. From Namaqualand to the Cape Peninsula and around Worcester, it is mainly found in flat areas in white to reddish, sandy soils and it frequently grows among fynbos. In the mountains around Cape Town it has been recorded at altitudes of 1000 m or more, also in fynbos. Euphorbia tuberosa may also be associated with karroid vegetation, though not in very arid areas, as on the northern margins of the Knersvlakte, from Nuwerus, around Loeriesfontein and in the higher parts of the Kamiesberg, where it grows in hard, stony ground among short bushes (as in Fig. 2.266). Along

the southern coastal plain, it also occurs in hard, shallow ground with other succulents or on low shale outcrops where it is wedged into crevices among the layers of rock. Diagnostic Features & Relationships In E. tuberosa the underground stem may remain solitary but usually produces swollen branches similar to but shorter than itself below the soil-surface, so that the plant is, in structure, like a small subterranean form of E. heptagona. During autumn (usually in April) the apex of each branch gives rise to a rosette of spreading petiolate leaves and several branched peduncles with fairly small unisexual cyathia. In habitat the plants of E. tuberosa mostly disappear entirely during the dry summer.

210

2  Euphorbia subg. Athymalus

Fig. 2.266.  Euphorbia tuberosa, each rosette ± 8 cm diam., leaves purplish and with finely crisped margins, PVB 5345, north of Leliefontein, Kamiesberg, South Africa, 13 Jul. 2006 (© PVB).

Fig. 2.267.  Euphorbia tuberosa, large male plant ±  20  cm diam., Constantiaberg, Cape Town, South Africa, 3 Aug. 2008 (© PVB). Fig. 2.269.  Euphorbia tuberosa, grey-leaved male plant ± 15 cm diam., Knolfontein, eastern Cedarberg, South Africa, 13 Sep. 2008 (© PVB).

Fig. 2.268.  Euphorbia tuberosa, female plant with ripening capsules, Constantiaberg, Cape Town, South Africa, 8 Oct. 2013 (© PVB).

2.1 Sect. Anthacanthae

In the coastal sandveld of Namaqualand, E. tuberosa forms very diffuse plants with relatively few branches that spread over a patch of up to 0.5 m. The branches are slender and are often 10 cm long, weakly joined to the stem and their tips are

211

mostly around 3 to 5 cm beneath the surface. Small clusters of a few leaves arise at the apex of each branch and these leaves are usually flat on the ground, forming small, widely scattered rosettes over the patch that the plant occupies.

Fig. 2.270.  Euphorbia tuberosa, grey-leaved male plant ± 15 cm diam. sunken among quartz gravel on plateau, Driekuilen, NE of Montagu, South Africa, 13 Aug. 2016 (© PVB).

Euphorbia tuberosa differs from the other two similar species by its branched underground stems. In the other two the stem is unbranched, except occasionally if it is damaged. In E. ecklonii the leaves are always dark green and somewhat shiny, with prominent veins on the upper surface. Generally, once fully developed, they are tightly pressed to the surface of the ground and are a little convex, with the margins straight or slightly reflexed. In the other two species the leaves are grey-green and narrower, varying in attitude from prostrate to erect. The margins are curved upwards (often strongly so) and are often strongly undulating or crisped. In E. tuberosa the leaves usually spread close to the ground in a fairly dense rosette, there is a fairly abrupt change from petiole into the much broader leaf and the margins of the leaves are erect and often crisped. In E. silenifolia the leaves are often erect, the petiole merges gradually into the slender, spathulate to filiform leaf and the margins are flat or slightly raised and are rarely very slightly crisped. Florally Euphorbia ecklonii, E. silenifolia and E. tuberosa differ in only a few, small details. The peduncles are slender and tall in E. silenifolia, usually erect among the leaves. In E. tuberosa they are generally stouter and much shorter, gath-

ered into a dense cluster in the centre of the rosette, often filling up much of the space in the centre. In E. ecklonii, they are also gathered in the centre, close to the leaves, but are not as densely clustered as in E. tuberosa. In E. silenifolia the female floret has short styles which spread horizontally on the top of the ovary, while they are longer and more erect in both E. ecklonii and E. tuberosa. The bracts and styles in E. ecklonii are usually red-green, which is not the case in the other two. Flowering in E. tuberosa takes place from June to August. Since E. silenifolia flowers from February to May and E. ecklonii from April to June, both are usually finished flowering before this gets underway in E. tuberosa. History Euphorbia tuberosa was known to Linneaus, but only apparently from the excellent and clearly identifiable plate published by Johannes Burman in 1738. Burman (1738) mentioned that the plant was collected by Paul Hermann, who visited the colony in 1672 and was the first person to collect plants there for a herbarium. It is probable that the ‘Tithymalus Africanus, humilis, radice tuberosa, foliis oblon-

212

2  Euphorbia subg. Athymalus

Fig. 2.271.  Euphorbia tuberosa, male plant with undulate margins to the leaves, Bantams, at northern base of Witteberg, west of Laingsburg, South Africa, 9 Aug. 2008 (© PVB).

Fig. 2.272.  Euphorbia tuberosa, female plant with capsules, margins of leaves scarcely undulate, Constantiaberg, Cape Town, South Africa, 8 Oct. 2013 (© PVB).

gis, in summo cordatis, pedunculis insidentibus biuncialibus’ listed in the Cat. plantarum Africanarum, a list of plants collected by Hermann at the Cape and forming an Appendix to Burman’s Thesaurus Zeylanicus (Burman 1737), refers to E. tuberosa. Euphorbia tuberosa is also the subject of Folio 189  in the volume of the Codex witsenii known as Icones plantarum et animalium (Macnae & Davidson 1969). Boissier (1862) considered that E. crispa and E. silenifolia were the same species and gave them the name E. ellip-

tica var. undulata. Here, E. silenifolia and E. tuberosa are treated as separate species, with the name E. crispa included under E. tuberosa. The name Tithymalus tuberosus is sometimes attributed to John Hill (Hortus Kewensis: 172.3 (1768)), but it only appeared there in a list of plants cultivated at Kew, without any further details, so the new combination was not validly published there but by Haworth in 1812.

Fig. 2.273.  Euphorbia tuberosa. A, side view of male cyathia (scale 4 mm, as for B). B, side view of female cyathia. C, anther and bracteole (scale 1 mm, as for D). D, female floret. Drawn from: PVB, Olifantsbos, Cape Point Nature Reserve, South Africa (© PVB).

2.1 Sect. Anthacanthae

2.1.2.4 Ser. Treisia Euphorbia ser. Treisia (Haw.) Bruyns, Taxon 62: 1194 (2013). Euphorbia subsect. Treisia (Haw.) Benth. & Hook.f., Gen. Pl. 3 (1): 260 (1880). Euphorbia sect. Treisia (Haw.) Baill., Étude Euphorb.: 284 (1858). Treisia Haw., Syn. Pl. Succ.: 131 (1812). Type (designated by Wheeler 1943): Treisia clava (Jacq.) Haw. (= Euphorbia clava Jacq.). Bisexual, spineless, mainly single-stemmed succulents (rarely branching into a shrub, as in E. pubiglans; in the others branching if stem damaged), stem cylindrical, densely tuberculate, with small slender rootstock, without sterile short-shoots (peduncles occasionally remaining but not

213

becoming spiny). Leaves soft, conspicuous, deciduous, persisting for a season, in tuft around apex of stem on tips of tubercles. Synflorescences solitary on very short to long sometimes branching peduncle, peduncle with 2–3 fairly prominent bracts at the tip immediately subtending and somewhat exceeding terminal bisexual cyathium, occasionally branching from axils of bracts below these. All five species of this section are found in South Africa only. Euphorbia clandestina, E. cylindrica and E. pubiglans are endemic to the Greater Cape Floristic Region, while E. clava occurs from its eastern margin into the Eastern Cape. The most slender-stemmed species, Euphorbia bubalina, is known from the Eastern Cape to northern KwaZulu-Natal and is the most widespread species.

Fig. 2.274.  Distribution of Euphorbia ser. Treisia (showing number of species per half-degree square) (© PVB).

Key to the species of ser. Treisia 1. Stem and branches covered densely with prominent and swollen tubercles that are about as long as their bases are tall (in direction of stem)...............................................................................................................................................2. 1. Stem and branches covered sparsely with low and inconspicuous tubercles that are much shorter than their bases are tall (in direction of stem).............................................................................................................................4. 2. Mature plant often forming small shrub with branches and stem equally tall, fertile short-shoots 15–70 mm long, bracts subtending cyathium 6–12 × 8–16 mm.......................................................................E. pubiglans 2. Mature plant single-stemmed or with short branches around apex of much taller stem (usually only if apex of stem damaged), fertile short-shoots ≤ 10 mm long, bracts subtending cyathium 3–8 × 3–4 mm......................................3. 3. Fertile short-shoots 4–10 mm long, cyathia shallowly urceolate, pedicel of female floret 1–1.5 mm long and tomentose...........................................................................................................................................E. cylindrica

214

2  Euphorbia subg. Athymalus

3. Cyathia sessile (short-shoots 0.5–2 mm long) and deeply urceolate, pedicel of female floret ± 0.5 mm long and sparsely pubescent...........................................................................................................................E. clandestina 4. Cyathia conical-cupular, female floret almost sessile (on pedicel < 0.5 mm long), styles 1.5–2.5 mm long and fused for at most 1 mm, though often branching nearly to base..........................................................E. bubalina 4. Cyathia shallowly bowl-shaped, pedicel of female floret ± 0.5 mm long, styles 4–10 mm long and fused for at least 2 mm........................................................................................................................................................E. clava

Euphorbia bubalina Boiss., Cent. Euphorb.: 26 (1860). Type: South Africa, Cape, among thorn-bushes near Buffelsrivier, Drège 4615 (P, holo.). Euphorbia laxiflora Kuntze, Revis. Gen. Pl. 3 (3): 286 (1898). Type: South Africa, East London, 5 Mar. 1894, Kuntze (NY, holo.; K, iso.). Euphorbia tugelensis N.E.Br., Fl. Cap. 5 (2): 335 (1915). Type: South Africa, Natal, near Tugela River, received July 1865, Gerrard 1626 (K, holo.; BM, TCD, W, iso.). Bisexual spineless single-stemmed when young and later few- to many-branched glabrous succulent 0.15–2.5 m tall, with cluster of fibrous roots just beneath soil. Stem erect, 10–25 (30) mm thick and tapering slightly towards base, covered with widely spaced tubercles, pale grey-green or green becoming covered with thin grey corky bark towards base. Branches ascending and soon erect, similar to stem, 0.15–0.5 m long; tubercles low and cushion-like, 1–2 mm long with very elongate hexagonal base, spreading slightly horizontally near tips, not joined into angles but sometimes arranged into 6–8 gradually spiralling but indistinct rows with irregular shallow grooves between them; leaves on tips of new tubercles towards apex of stem, 40–150 × 6–35 mm (broadest usually slightly nearer apex than base), dark greygreen to green, glabrous, spreading to decurved, only very slightly fleshy, present during growing season and shed during dry periods, broadly linear to linear-lanceolate or linearobovate and tapering towards base, longitudinally slightly folded upwards, with straight or very slightly wavy margins, acute to obtuse, sessile. Synflorescences many per stem towards apex, each of 1 bisexual cyathium (occasionally another 1–2 cyathia arising on short rays 6–20 mm long just below primary cyathium) in axil of tubercle, arising on slender short-shoot (30) 50–180 × 1.5–2 mm, with usually 1–2 alternating bracts below but near apex and 2–3 larger bracts in plate-like whorl immediately under cyathium, bracts along short-shoot (and first of those beneath cyathium) 8–20 × 3–8 mm, narrowly elliptic and obtuse, other two under cyathium 10–20 × 10–15 mm, very broadly cordate or deltate-ovate, acute, glabrous to minutely pubescent and sometimes with reddish margins, persistent with

short-shoot but shed after capsule ripens (short-shoot usually falling off but occasionally persisting for one or more seasons as slender dried stalk), grey-green to reddish green; cyathia conical-cupular, glabrous outside, 6–8  mm broad, with 5 finely pubescent lobes with finely crenulate margins, green; glands 5, spreading, transversely elliptic, not contiguous, 2 × 2.5–3 mm, green, outer margins entire to very finely crenulate and spreading, inner margins slightly raised; stamens with finely pubescent pedicels, bracteoles slender filiform and densely pubescent above; ovary ellipsoidal, glabrous, sessile or very slightly raised on pedicel < 0.5 mm long; styles 1.5–2.5 mm long, branched to near base or to above middle. Capsule 8–9 mm diam., very obtusely 3-angled to nearly spherical, somewhat shiny pale green, glabrous, sessile. Distribution & Habitat Euphorbia bubalina is fairly well known in the Eastern Cape, near the coast in river valleys between Port Alfred and East London. Further north it has been recorded from Oribi Gorge in southern Natal, somewhat further north (though still south of Durban) from ‘Alexandra County’ (Saunders, NH) and nearby at Dumisa. Several collections have also been made in Ngoye Forest near Eshowe in the northern part of Natal. It is somewhat peculiar that plants known as E. tugelensis have never been recorded with any precision from the valley of the Tugela River itself. In the Eastern Cape, E. bubalina occurs mainly in the base of valleys in soft sand among grasses, usually around the edges of clumps of trees. In these situations, they may form colonies, some of which consist of large numbers of plants. In Natal, the species is very poorly known. The locality at Dumisa is unlikely to exist any longer, since the area has been taken over by the cultivation of sugar-cane. In the Oribi Gorge E. bubalina was recorded only once in a shaded situation on a moist cliff. At its northernmost station, in the Ngoye Forest, it occurs in shallow, humus-rich soil on domelike outcrops of granite that project from the forest. Here it is locally plentiful, often in the shade of small trees or shrubs.

2.1 Sect. Anthacanthae

215

Fig. 2.275.  Distribution of Euphorbia bubalina (© PVB).

Fig. 2.276.  Euphorbia bubalina, among trees on edge of denser bush, ± 1 m tall, PVB 12130, Bathurst Forest, Kap River, South Africa, 4 Apr. 2012 (© P. Hanáček).

Fig. 2.277.  Euphorbia bubalina, among trees, PVB 12130, Bathurst Forest, Kap River, South Africa, 4 Apr. 2012 (© P. Hanáček).

216

Diagnostic Features & Relationships Euphorbia bubalina is a slender succulent herb that reaches maturity at heights of between 0.2 and 1 m. A typical plant consists of a slender, grey-green, slightly succulent stem 10 to 20 mm thick with a rubbery consistency which may branch occasionally into a shrub but more usually remains solitary. Usually the more shrubby specimens are in more exposed situations, with those that are well-sheltered remaining single-stemmed or only few-branched towards their apices. During the growing season the stem and branches bear a cluster of conspicuous leaves near their tips, which all fall off during the dry season. Each leaf, which is quite long and narrow, is borne at the tip of a low tubercle and leaves an elliptic scar that remains at the tip of the tubercle. The stem is covered with these low tubercles, which lend it a slightly scaly appearance and it is a faintly greyish green in the upper third to upper half, with the older parts gradually losing the green colour and becoming covered with a thin, greyish bark lower down. Once growth becomes vigorous (i.e. between late spring and the middle of summer, continuing to early autumn) the slender, spreading short-shoots develop from the axils of most of the tubercles near the tip of the stem (and branches). They soon reach their typical length of between 70 and 10 cm long and bear cyathia. Most of the short-shoots have a succession of a few elliptic bracts towards their apex and a cluster of considerably larger bracts right under the terminal cyathium. The cluster of larger bracts usually consists of one elliptic one (the first to arise and the lowest on the shortshoot) and then two much broader, more or less triangular ones that are actually closer to the cyathium. The terminal cyathium is quite inconspicuous against these bracts. Further cyathia may arise on branches developing in the axils of the upper bracts. Clearly very closely related to such species as E. clava, E. clandestina and E. cylindrica, E. bubalina differs from these by its much more slender stem with considerably fewer, lower (i.e. flatter) tubercles. In Bruyns (2012), E. bubalina and E. tugelensis were maintained as separate species. They were said to differ (White et  al. 1941) by their differently shaped bracts subtending the cyathia (abruptly acute or obtuse and apiculate = E. bubalina; gradually acute or acuminate = E. tugelensis). Koutnik (1984a) first pointed out that these two species may be the same and an examination of material from both the Eastern Cape and from Natal has confirmed this. It has been found that plants from Ngoye Forest differ from those in the Eastern Cape by the slightly narrower leaves, the narrower cyathia, less closely contiguous glands, shorter fused base of the style and a very slightly longer pedicel to the female flower.

2  Euphorbia subg. Athymalus

Fig. 2.278.  Euphorbia bubalina, in shallow soil on granitic rocks at edge of forest, ± 1 m tall, PVB 12112, Ngoye Forest, South Africa, 7 Jan. 2012 (© PVB).

History Euphorbia bubalina was discovered relatively early in the exploration of the Cape Flora, with the first collection made by William Burchell on 9 October 1813 near Port Alfred. It was described from a collection of J.F.  Drège. This was likely to have been made around the middle of January 1832, when Drège and his brother Carl crossed the Buffels River (near the present-day East London), having joined an expedition of Andrew Smith to Natal. Another two specimens were collected by Thomas Cooper in 1860 somewhere in ‘British Kaffraria’, but E. bubalina was not described from Cooper’s specimens as Phillips (1928a) stated. Euphorbia tugelensis was described from a specimen collected by William Tyrer Gerrard between 1860 and 1865 somewhere near the Tugela River in Natal. At the time of describing it, N.E. Brown saw another specimen collected by Mrs Saunders in Alexandra County, south of Durban. Euphorbia tugelensis was never a well-known species and

2.1 Sect. Anthacanthae

Fig. 2.279.  Euphorbia bubalina, cyathium with bracts, PVB 12130, Bathurst Forest, Kap River, South Africa (© PVB).

217

Fig. 2.280.  Euphorbia bubalina, cyathium with bracts, PVB 12112, Ngoye Forest, South Africa (© PVB).

Fig. 2.281.  Euphorbia bubalina. A, B, cyathium from side (scale 2 mm, as for B–D). C, D, cyathium from above. E, anthers and bracteoles (scale 1 mm, as for F, G). F, G, female floret. Drawn from: A, C, E, F, PVB 12112, Ngoye Forest, South Africa. B, D, PVB 12130, Bathurst Forest, Kap River, South Africa. G, PVB 12129, Kowie River, South Africa (© PVB).

modern records of it are confined to a collection made in January 1931 by A.W. Bayer near Dumisa (living plants of which were seen by White et  al. (1941)) and one made in 1965 by R.G. Strey in the Oribi Gorge. Several collections have been made also in Ngoye Forest, much further to the north, where it was first recorded by E.J.  Moll in March 1972. Euphorbia clandestina Jacq., Pl. Hort. Schönbr. 4: 43, t. 484 (1804). Lectotype (Bruyns 2012): Jacq., Pl. Hort. Schönbr. 4: t. 484 (1804).

Bisexual spineless usually single-stemmed glabrous succulent 80–600 mm (–1.2 m) tall (rarely with few branches, these usually much shorter than stem and developing mainly if apex of stem damaged), with cluster of fibrous roots just beneath surface of soil. Stem erect, clavate and tapering strongly towards base, 25–70 mm thick, covered with tubercles, pale green mottled with purple or dark green; tubercles prominent and conical, 3–9 mm long with square to hexagonal base, spreading horizontally near tips, not joined into angles but arranged into 8–15 spiralling rows with irregular shallow grooves between them; leaves on tips of new tuber-

218

cles towards apex of stem (and branches), 20–60 × 2–4 mm (broadest in middle), grey-green to purplish green, glabrous, ascending to spreading, slightly fleshy, present during growing season and shed during dry periods, linear to linearoblanceolate, longitudinally folded upwards, with straight sometimes slightly reddish margins, acute to obtuse, sessile. Synflorescences many per stem (and branches) towards apex, each of one bisexual cyathium in axil of tubercle, ± sessile (on short-shoot 0.5–2 mm long), surrounded by 6–9 erect to spreading bracts in 2–3 whorls, bracts 4–8 × 3–8 mm (those near base sometimes much smaller), oblong, very shortly obtuse, apiculate, minutely puberulous, closely surrounding cyathium and closing around capsule till ripe then shed rapidly, green to purple-green later becoming reddish and

2  Euphorbia subg. Athymalus

sometimes with darker margins; cyathia urceolate, minutely puberulous outside, 2.5–3  mm broad and half as long as inner bracts, with 5 lobes with crenulate margins, pale green below becoming darker above; glands 5 (–7), ascending, transversely elliptic, contiguous, 1 × 2 mm, green, outer margins entire and ascending to spreading; stamens with finely pubescent pedicels, bracteoles absent; ovary globose, finely pubescent, raised on pedicel ± 0.5 mm long; styles 2–2.5 mm long, branched to around middle. Capsule 4–5  mm diam., obtusely 3-angled, finely pubescent, grey-green, sessile. Distribution & Habitat Euphorbia clandestina is of sporadic occurrence on the Klein Karoo from Montagu eastwards to around Oudtshoorn

Fig. 2.282.  Distribution of Euphorbia clandestina (© PVB).

and in the foothills of the Kamannassie Mountains eastwards to De Rust and to a little north of Uniondale. On the coastal plain south of the Langeberg, it is known (again very sporadically) in slightly arid patches from Swellendam to Albertinia and on the coastal limestones from Bredasdorp to around Mossel Bay. Plants of E. clandestina are found in low-lying, flat areas to the slopes of low hills. Soils are typically heavy loams, where they grow in leaf-litter under scattered trees, among small shrubs in karroid vegetation or in dry patches of renosterveld and dense stands of Pteronia incana, often with Aloe ferox and Cynanchum viminale. In the coastal limestones they occur in dry, karroid patches or even occasionally among fynbos. They are usually locally quite common, though the populations tend to be widely scattered.

Diagnostic Features & Relationships In E. clandestina the plant usually consists of a single, stout, erect stem which tapers off greatly towards the ground and gives rise to slender, spreading and superficial roots just beneath the surface. If the stem is damaged apically, branches may develop around the damaged area, but most plants consist only of a single, unbranched stem. The stems in E. clandestina are very variable in stoutness, with those in more arid, karroid vegetation on hard clays (as, for example on the Klein Karoo or around Riversdale) much thicker than those occurring on the coastal limestones. The stem is green mottled with purple and is covered quite densely with prominent tubercles. When young, each tubercle bears a prominent, soon deciduous leaf at its tip so that, in the growing season, each plant has a rosette of leaves around the apex of the stem.

2.1 Sect. Anthacanthae

219

Fig. 2.283.  Euphorbia clandestina, the taller plants here ± 30 cm tall, with Aloe ferox, PVB 9972, west of Heidelberg, South Africa, 27 May 2005 (© PVB).

Fig. 2.284.  Euphorbia clandestina, PVB 9972, west of Heidelberg, South Africa, 27 May 2005 (© PVB).

Fig. 2.285.  Euphorbia clandestina, PVB 10571, near Dysselsdorp, South Africa, 14 Aug. 2010 (© PVB).

220

2  Euphorbia subg. Athymalus

Leaves in E. clandestina are long, very narrow and deeply channelled on their upper surface, closely matching the stem in colour. Euphorbia clandestina flowers from April to September. During this period many cyathia are produced in the axils of the tubercles near the top of the stem just below new growth for the season (and on the last season’s growth). Since each fertile short-shoot is reduced to a very short axis bearing a single cyathium at its apex, the cyathia are more or less sessile. Each cyathium is surrounded quite tightly by a whorl of erect or spreading, relatively narrow bracts, of which the upper ones greatly exceed the length of the cyathium. The cyathium is urceolate, has fairly small, ascending glands near the top, only small numbers of male florets and no bracteoles inside at all. The female floret has a very short pedicel so that it and the developing capsule are held close to the stem. After flowering is over and the capsule begins to develop, the bracts close up around the cyathium. They remain in this position until the capsule is ripe, falling off soon after it dehisces. Plants appear to be quite short-lived (though older specimens up to 1 m tall will be seen in some populations) and produce lots of seed, by means of which they spread and keep the population going. They mature quickly from seed, often flowering for the first time within 1 to 2 years of germination, though this depends on the amount of rainfall during the warmer months of active growth. Fig. 2.286.  Euphorbia clandestina, plant about 0.5  m tall, east of Oudtshoorn, South Africa, 20 Oct. 2009 (© PVB).

Fig. 2.287.  Euphorbia clandestina, with cyathia enclosed by bracts, PVB 6655, De Hoop, South Africa, Oct. 2007 (© PVB).

Fig. 2.288.  Euphorbia clandestina, with capsules concealed by bracts, PVB 6655, De Hoop, South Africa, Oct. 2007 (© PVB).

2.1 Sect. Anthacanthae

In E. bubalina, E. clava and E. pubiglans the bracts subtending the cyathium are broad, forming a plate-like background in the centre of which the cyathium is situated and the short-shoots may be very long indeed. In E. clandestina and E. cylindrica the bracts surrounding the cyathium closely are narrow and in both species the short-shoot is short (and it is practically absent in E. clandestina). Euphorbia clandes-

221

tina is unique in having particularly narrow bracts which are often quite tightly erect around the cyathium. The markedly urceolate shape of the cyathium is also unique. Further differences between E. clandestina and E. cylindrica are discussed under the latter and between it and E. clava under E. clava.

Fig. 2.289.  Euphorbia clandestina. A, bracts cut away to show cyathium from side (scale 2 mm). B, side view of part of dissected cyathium (scale 1 mm, as for C). C, anther and bracteole. Drawn from: M.B. Bayer, Volmoed turnoff, SW of Oudtshoorn, South Africa (© PVB).

History Euphorbia clandestina was described by Jacquin from plants in cultivation at the Imperial gardens at Schönbrunn Palace in Vienna. These were ostensibly grown from seed collected by the two gardeners from Schönbrunn, Franz Boos and Georg Scholl, who collected plants and other natural history material together at the Cape between May 1786 and February 1788. It is, however, possible that E. clandestina was known earlier. The figure of Commelijn (1703: 58, t. 8) may be of E. clandestina rather than of E. clava, as assumed by White et  al. (1941). Since the specimen figured is sterile, it is impossible to assign it to either of these species with certainty. Nevertheless, E. clandestina occurs closer to Cape Town than E. clava and is more likely to have been encountered very early in the exploration of the region. Linnaeus included Commelijn’s figure under ‘E. caput-medusae var α’ along with some other species.

Euphorbia clava Jacq., Icon. Pl. Rar. 1 (4): 9, t. 85 (1784). Treisia clava (Jacq.) Haw., Syn. Pl. Succ.: 131 (1812). Lectotype (Wijnands 1983): Jacq., Icon. Pl. Rar. 1 (4): 1: t. 85 (1781). Euphorbia canaliculata Lam., Encycl. 2 (2): 417 (1788). Type: South Africa, collector unknown (P-LAM, holo.). Euphorbia clavata Salisb., Prodr. Stirp. Chap. Allerton: 389 (1796), nom. superfl. Euphorbia coronata Thunb., Prodr. Fl. Cap. 2: 86 (1800). Type: South Africa, Cape, Thunberg (UPS-THUNB, holo.). Treisia tuberculata Haw., Suppl. Pl. Succ.: 65 (1819). Type: Introduced by D. Young to Epsom, 1815, fl. Chelsea 1818 (missing). Euphorbia haworthii Sweet, Hort. Brit. 2: 356 (1826), nom. superfl. Bisexual spineless mostly single-stemmed (when young) glabrous succulent 60–800  mm (–1.5 m) tall, later often

222

branching towards apex (especially if stem apically damaged), with cluster of fibrous roots just beneath surface of soil. Stem erect, slightly clavate and abruptly tapering towards base, 25–60  mm thick, covered with tubercles, pale greygreen mottled with purple or dark green, often becoming covered with thin grey bark towards base. Branches erect from shortly spreading base, similar to stem; tubercles low to fairly prominent and broadly conical, 2–5 mm long with rhomboidal to hexagonal base, spreading horizontally near tips, not joined into angles but arranged into 8–15 very obscure and slightly spiralling rows with irregular and slight shallow grooves between them; leaves on tips of new tubercles towards apex of stem and branches, 20–130 × 3–6 mm (broadest usually above middle), grey-green to purplish green, glabrous, ascending to spreading, slightly fleshy, present during growing season and shed during dry periods, linear to linearoblanceolate and tapering towards base, longitudinally folded upwards, with straight margins, usually acute, sessile. Synflorescences many per stem towards apex, each of one bisexual cyathium, arising in axil of tubercle on slender minutely pubescent often shiny purple-red ascending shortshoot (40) 70–180 × 1–1.5 mm that may dry out and persist for several seasons as woody stalk, with 3–6 alternating bracts along them mainly in lower half and 3–5 larger bracts in broadly bowl-like whorl immediately under cyathium, bracts along short-shoot 4–10 × 2–6 mm, oblong to lanceolate and acute, those under cyathium 6–10 × 6–12 mm, broadly ovate or elliptic-rhomboid, abruptly apiculate or acute, glabrous to minutely pubescent, persistent with short-shoot but shed rap-

Fig. 2.290.  Distribution of Euphorbia clava (© PVB).

2  Euphorbia subg. Athymalus

idly after capsule ripens, grey-green to purple-green becoming reddish towards often finely denticulate edges; cyathia shallowly bowl-shaped, glabrous (rarely minutely pubescent) outside, 6–9 mm broad, with 5 lobes with finely toothed margins; glands 5, spreading, semicircular to almost kidneyshaped, contiguous, ± 1.5 × 3–3.5 mm, dark green, glabrous, outer margins often ascending, entire, inner margins not raised; stamens with glabrous pedicels, bracteoles very few and narrowly linear; ovary slightly globose, glabrous, raised on pedicel ±  0.5  mm long; styles 4–10  mm long, branched only in upper third. Capsule 7–10 mm diam., almost spherical, glabrous, grey-green to purplish or reddish, sessile. Distribution & Habitat Euphorbia clava occurs in the Eastern Cape, roughly between Humansdorp and Peddie. It is rarely found more than 50 km from the coast, except for a few records from the vicinity of Grahamstown, which are slightly further inland. Euphorbia clava is often associated with short scrub around the edges of dense thickets, usually on heavy, loamy soils derived from shales. Around Hankey and Port Elizabeth it commonly occurs in open patches (dominated by Pteronia incana) in dry, dense scrub with a large variety of other succulents and is found in similar, open patches in otherwise dense bush further east to near Peddie, occasionally also growing inside these dense bushes as well. Plants inhabit steep slopes as well as flat areas. A few times they have been seen flourishing in disturbed spots alongside roads and seem to have a slight ruderal tendency.

2.1 Sect. Anthacanthae

223

Diagnostic Features & Relationships Plants of Euphorbia clava are quite variable in habit. This appears to be largely dependent on the amount of shelter received, with those growing in denser bush often with the longest and least branched stems. Many specimens consist of a single, erect stem which may reach a metre or more in height, with a few short branches near its apex. However, if

this stem is damaged apically, many branches may develop around the damaged part and the whole plant may remain less than 0.5  m tall. The stem and branches are greyish green and fleshy, covered with short, but very broadly based tubercles, whose bases are usually much longer in the direction of the axis of the stem than in the direction across the stem. Each tubercle has a small apical projecting

Fig. 2.291.  Euphorbia clava, longest stem ± 1 m tall, alongside dense bush just east of Uitenhage, South Africa, 25 Oct. 2012 (© PVB).

Fig. 2.292.  Euphorbia clava, robust branched plant ± 0.8 m tall, just east of Uitenhage, South Africa, 25 Oct. 2012 (© PVB).

point that tapers into the leaf and spreads somewhat from the surface of the stem. They are usually arranged into very obscure, slightly spiralling rows along the stem. After rains, new tubercles develop at the apex of the stem and branches, each of which tapers off into a long, spreading and usually very slender, greyish green leaf. The leaves persist for a season in a cluster around the apex of the stem and each branch and then fall off, leaving a small, almost circular scar at the tip of each tubercle. Somewhat below the tops of the stem and branches, the old, slightly woody remains of peduncles from previous flowering seasons are found.

Flowering appears to be dependent on rains and has been observed between August and October. The cyathia are borne on slender, relatively pliable short-shoots that arise among the uppermost tubercles, usually just below or among the crop of leaves present at the time. The short-shoots vary greatly in length but are frequently at least 80 mm long and always hold the cyathium far away from the stem. Each short-shoot bears several small bracts that are widely spaced along its shaft followed by a whorl of three to five, greatly enlarged, almost petal-like (sometimes reddish but usually greyish green) bracts immediately under the cyathium. These expanded bracts form a broadly bowl-like structure which closely sur-

224

2  Euphorbia subg. Athymalus

Fig. 2.293.  Euphorbia clava, plant about 0.4  m tall in rocky spot alongside dense bush, north of Bathurst, South Africa, 4 Apr. 2012 (© PVB).

Fig. 2.294.  Euphorbia clava, alongside bushes with small Delosperma, north of Bathurst, South Africa, 4 Apr. 2012 (© PVB).

rounds the solitary cyathium in the centre. The cyathium is shallowly cup-shaped, with the closely contiguous, glabrous glands rising above it. Inside it has many inconspicuous glabrous male florets and a small, glabrous ovary topped by a

relatively long, cylindrical style which projects well above the cyathium before dividing and spreading. The ovary develops into an almost spherical, glabrous capsule around which the bracts persist as dried-out husks.

Fig. 2.295.  Euphorbia clava, plant in flower, PVB 12156, west of Uitenhage, South Africa, Oct. 2012 (© PVB).

Fig. 2.296.  Euphorbia clava, cyathium with bracts, PVB 12156, west of Uitenhage, South Africa, Oct. 2012 (© PVB).

2.1 Sect. Anthacanthae

Euphorbia clava differs from E. bubalina by the much stouter stems and branches and the more prominent tubercles. In E. clandestina, E. clava, E. cylindrica and E. pubiglans the tubercles on the surface of the stems and branches stand out more and are also much more densely packed on the stem or branch. Furthermore, in these four species the cyathium is solitary at the tip of the short-shoot unlike in E. bubalina, where the short-shoot branches extensively to pro-

225

duce many cyathia. From E. clandestina and E. cylindrica, both E. clava and E. pubiglans differ most obviously by the much longer short-shoots and by the much broader bracts subtending the cyathia. Less obvious differences lie in the broader and shallower cyathia of E. clava and E. pubiglans. The differences between E. clava and E. pubiglans are discussed under E. pubiglans.

Fig. 2.297.  Euphorbia clava. A, B, bracts cut away to show cyathium from side (scale 2 mm, as for B). C, side view of dissected cyathium (scale 1 mm, as for D). C, anthers and bracteole. Drawn from: PVB 12156, west of Uitenhage, South Africa, South Africa (© PVB).

History Euphorbia clava has been known in Europe for a very long time, where it was introduced first to various Dutch gardens. One of these plants was figured by Burman in 1738, but it is not known who brought these early collections to Europe. White et al. (1941) also reproduced an even earlier figure by Commelijn (1703: 58, t. 8) under their account of E. clava. However, this may represent E. clandestina rather than E. clava, since the leaves are very slightly broader and especially since the plants shown are sterile, when these two species are difficult to separate. Slightly better documented are various collections that were made by early botanical explorers at the Cape. One such was a collection by Francis Masson, who introduced it to England in 1774, according to Aiton (1789). This must have been encountered during his expedition with C.P.  Thunberg as far as Uitenhage and the Sundays River from September 1773 to January 1774 and it is probable that

the plants named E. coronata by Thunberg were collected during the same expedition. Euphorbia cylindrica Marloth ex A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 962 (1941). Type: South Africa, Cape, Kubiskow Mountain, Marloth 12860 (PRE, holo.). Bisexual spineless usually single-stemmed glabrous succulent 0.1–0.5 (1) m tall (rarely with few branches, these usually identical to stem and developing only if apex of stem damaged), with cluster of fibrous roots just beneath surface of soil. Stem erect, clavate and tapering towards base, 35–60  mm thick, covered with tubercles, pale green mottled with purple or dark green; tubercles prominent and conical, 3–10 mm long with broad hexagonal base, spreading horizontally or slightly deflexed near tips, not joined into angles but arranged into 8–15 spiralling rows with

226

irregular shallow grooves surrounding them; leaves on tips of new tubercles towards apex of stem (and branches), 30–80 × 5–12 mm (broadest in middle), usually somewhat folded upwards lengthwise, initially red then becoming dark grey-green above with a paler midrib and undersurface, glabrous, ascending to spreading, present during winter months and shed during summer, narrowly elliptic-oblong usually with slightly undulate and often reddish margins, obtuse, narrowing towards base into ascending petiole 3–10  mm long. Synflorescences many per stem (and branches) towards apex, each with bisexual cyathium in axil of tubercle on finely pubescent short-shoot 4–10 × 1.5–2.5 mm, with 3–7 ascending broadly spathulate-oblong minutely pubescent caducous bracts 3–5  × 3–4  mm; cyathia urceolate, finely pubescent outside, 5–6  mm broad, with 5 whitish lobes (faintly suffused with red) with deeply incised margins;

2  Euphorbia subg. Athymalus

glands 5, cuneate, contiguous, 1.5 × 3–3.5 mm, pale green below, reddish on edges, deep green above, inner margins slightly raised, outer margins entire; stamens with finely tomentose pedicels, bracteoles few, filiform, pubescent in upper half; ovary ellipsoidal, finely tomentose, tapering gradually into glabrous styles, raised on finely tomentose pedicel 1–1.5  mm long; styles 1.5–2  mm long, divided to near middle, reddish pale green. Capsule 5–8  mm diam., obtusely 3-angled, finely pubescent, sessile to borne on pedicel 2–4 mm long. Distribution & Habitat Euphorbia cylindrica is known in the former Cape Province between Loeriesfontein, Nieuwoudville and Calvinia, on the eastern sides of the mountains that form the escarpment and in some of the valleys east of these mountains.

Fig. 2.298.  Distribution of Euphorbia cylindrica (© PVB).

Plants may be found on the slopes of shaly hills among shrubs of Pteronia incana and various species of Eriocephalus, often on the cooler and more densely vegetated, southern aspect, sheltering within or alongside these shrubs. In some localities it was also seen in fairly arid, flat areas at and near the bases of hills growing among Eriocephalus bushes and low shrublets of Leipoldtia schultzei. They are invariably found in small, often very widely scattered colonies of up to 30 individuals of various ages.

Diagnostic Features & Relationships Specimens of E. cylindrica form single, erect stems which occasionally branch near their apex if the growing point is damaged. The stem varies greatly in length, with most of them in a colony between 0.15 and 0.3 m tall. While most plants are probably fairly short-lived, a few manage to reach a ripe old age, to achieve 1  m in height and project well above the bushes surrounding them. The stem is quite stout, green mottled with purple or darker green and covered with prominent tubercles that taper quickly to a small, spreading,

2.1 Sect. Anthacanthae

227

Fig. 2.299.  Euphorbia cylindrica, young plant 0.2  m tall with capsules, among shrubs of Pteronia incana, PVB 1097, Grasberg, north of Nieuwoudville, South Africa, 11 Jul. 2006 (© PVB).

Fig. 2.300.  Euphorbia cylindrica, plant nearly 0.75  m tall, PVB 10864, Loeriesfontein, South Africa, 13 Sep. 2007 (© PVB).

but not at all sharp tip. On new growth each tubercle bears a prominent leaf at its apex and these leaves spread away from the stem in a rosette around its apex. When they appear (which may happen as early as January, if early rains are received) the leaves are bright red and this colour soon changes to dark grey-green above and slightly paler below. They are folded upwards along their length, with obviously wavy, often reddish margins. The leaves are rapidly shed with the onset of the drier months and each one leaves a neat, elliptic scar at the tip of its respective tubercle. Flowers appear any time between January and early June (depending on rains), in the axils of the tubercles around the apex of the stem, usually just below new growth for the season. After this, the stem usually elongates slightly more to produce further tubercles and leaves, so that the fruits may ripen under a canopy of leaves, though the leaves are often eaten off by stock, mostly leaving the stem and capsules unharmed. Most of the peduncle and cyathium is covered

with fine, short, greyish hairs. Each cyathium is borne on a short short-shoot and is surrounded by a whorl of short and broad, grey-green bracts that usually just exceeds it, so that the cyathium just fills up the bowl-like area between these bracts. These bracts soon dry up and fall off, even while the cyathium is still actively flowering. On the cyathium, the glands touch each other laterally and their dark green inner surfaces (matching the colour of the upper surfaces of the leaves) appear to form a more or less continuous ring inside the bracts. This dark colour contrasts with the pale green of the bracts and the whitish of the cyathial lobes and sides of the cyathium but, being small and hidden among the leaves, the cyathia are not conspicuous. Inside the cyathium, the female florets and the pedicels of the males are covered with a dense tomentum of fine hairs. The pedicels of the males nearest to the female are pressed into this tomentum on the female pedicel and on the ovary, leaving indentations in it when they are pulled away.

228

2  Euphorbia subg. Athymalus

Fig. 2.301.  Euphorbia cylindrica, plant with male-only cyathia, PVB 11192, SW of Loeriesfontein, South Africa, 24 Jan. 2012 (© PVB).

Fig. 2.302.  Euphorbia cylindrica, cyathia bisexual, PVB 1097, Grasberg, north of Nieuwoudville, South Africa, 29 Mar. 2009 (© PVB).

Fig. 2.303.  Euphorbia cylindrica, cyathia in early female stage, PVB 1097, Grasberg, north of Nieuwoudville, South Africa, 6 Feb. 2009 (© PVB).

Euphorbia cylindrica is closely related to E. clandestina, sharing with it both the very similar habit and appearance. The two differ in that the leaves of E. cylindrica are broader than those of E. clandestina and are usually distinctly undulate along their margins, the cyathia in E. cylindrica are borne on a short-shoot at least 4 mm long, while in E. clandestina they are more or less sessile. The bracts are somewhat broader and shorter than those of E. clandestina, with the upper few more or less equalling or just exceeding the height of the cyathium. The cyathia in both are similarly shaped, but that in E. clandestina is smaller and more strongly urceolate. The pedicel of the female flower is much longer in E. cylindrica, but the styles are both shorter and are not nearly as slender as those of E. clandestina. The manner

in which the ovary tapers into the styles in E. cylindrica is also distinctive. In E. cylindrica the ovary and pedicel of the female and of the males is covered with a fine, grey-green tomentum, which is much denser than the hairiness on these organs in E. clandestina. The male florets in E. cylindrica are more densely packed into the cyathium and bracteoles are present inside the cyathium. Here the bracts on the peduncle fall off while the cyathium is still flowering and before the capsule begins to develop so that, while it is ripening, the capsule is exposed at the tip of the short-shoot (as in Fig. 2.304). In E. clandestina, on the other hand, the bracts remain wrapped around the capsule until it is ripe and are almost double its length. The capsule is also larger in E. cylindrica than in E. clandestina.

2.1 Sect. Anthacanthae

229

Fig. 2.304.  Euphorbia cylindrica, fruiting plant among Leipoldtia schultzei, PVB 11192, SW of Loeriesfontein, South Africa, 6 Sep. 2008 (© PVB).

Fig. 2.305.  Euphorbia cylindrica. A, cyathium from side with some bracts removed (scale 2 mm). B, part of dissected cyathium with anther and bracteoles (scale 1 mm, as for C–E). C, anthers and bracteoles. D, E, female floret (base of ovary and pedicel tomentose in D but not shown). Drawn from: A–C, E, PVB 1097, Grasberg, north of Nieuwoudville, South Africa. D, PVB 11192, SW of Loeriesfontein, South Africa (© PVB).

History Euphorbia cylindrica was discovered by the land-surveyor E.B. Watermeyer in the Hantam Mountains in 1916 and he also brought a specimen from the ‘Bokkeveldberg’ to Marloth, ostensibly in January 1925. Marloth collected the material later designated as the type on 7 September 1926

and this species still occurs around the base of Kubiskow Mountain, where he found it. R.A. Dyer wrote on a label on the type specimen at PRE that the description and name were ‘taken partly from Marloth’s notes’ and so ‘Marloth ex’ has been added to the authorship of the name here.

230

Euphorbia pubiglans N.E.Br., Fl. Cap. 5 (2): 338 (1915). Type: South Africa, Cape, near Port Elizabeth, Sept. 1912, I.L. Drège (K, holo.). Bisexual spineless single-stemmed (when young) glabrous succulent 60–300 mm tall later often branching from near base (and rarely unbranched in older plants) to form clump of roughly equally tall branches with stem, with cluster of fibrous roots just beneath surface of soil. Stem erect, slightly clavate and abruptly tapering towards base, 20–50 mm thick, covered densely with tubercles, pale greygreen mottled with purple or dark green, often becoming covered with thin grey bark towards base. Branches erect or ascending from shortly spreading base, similar to stem; tubercles prominent and conical, 3–6 mm long, with rhomboidal to hexagonal base, spreading slightly horizontally near tips, not joined into angles but arranged into 9–15 spiralling rows with sometimes quite prominent grooves between them; leaves on tips of new tubercles towards apex of stem and branches, 20–110 × 2–4 mm (broadest usually near apex), grey-green to purplish green, glabrous, ascending to spreading, slightly fleshy, present during growing season and shed during dry periods, linear to linear-oblanceolate and tapering towards base, longitudinally folded upwards, with straight margins, usually obtuse, sessile. Synflorescences many per stem towards apex, each of one bisexual cyathium, arising in axil of tubercle on slender glabrous to minutely pubescent pale green short-shoots 15–70 × 1–2 mm that dry out and persist for several seasons as woody stalks, with 3–6 alternating bracts along them and 3–5 larger bracts in plate-

Fig. 2.306.  Distribution of Euphorbia pubiglans (© PVB).

2  Euphorbia subg. Athymalus

like whorl immediately under cyathium, bracts along shortshoot 2–6 (10) × 1–4 mm, usually lanceolate (ovate) and acute, those under cyathium 6–12 × 8–16 mm, very broadly ovate to rhomboid, obtuse and minutely apiculate, usually glabrous, persistent with short-shoot but shed rapidly after capsule ripens, usually grey-green and occasionally becoming reddish towards edges; cyathia shallowly bowl-shaped, minutely pubescent outside, 6–7  mm broad, with 5 broad and ascending lobes with finely toothed margins; glands 5, spreading, transversely elliptic, contiguous, 1 × 2.5–3.5 mm, dark green, outer margins finely pubescent entire and spreading, inner margins slightly raised and finely pubescent; stamens with finely pubescent pedicels, bracteoles absent; ovary globose, finely pubescent, raised on pedicel 0.7–1 mm long; styles 1–1.5  mm long, branched to near base but remaining close together. Capsule 7–8  mm diam., almost spherical, slightly flattened, finely pubescent, grey-green, sessile. Distribution & Habitat Euphorbia pubiglans occurs between Knysna and Port Elizabeth in the southern and south-eastern parts of the former Cape Province, mainly within 60 km of the coast, with some records in the Long Kloof being the furthest from the sea. Plants of E. pubiglans are especially associated with dry fynbos on acidic soils derived from sandstones or on banks of alluvium (also mainly originating from sandstones and probably also mainly acidic). Around Port Elizabeth it has been recorded in dry grassland with other succulents. In the

2.1 Sect. Anthacanthae

231

Baviaanskloof Mountains E. pubiglans inhabits exposed outcrops of pale sandstone on north-facing slopes, surrounded by short grassland. In crevices among rocks in these spots it grows with many other succulents, belonging to Crassula, Senecio and the succulent Aizoaceae. The same habitat is favoured in the mountains between Knysna and Uniondale. Steep banks of alluvial pebbles, among dry vegetation dominated by Dodonea, Erica, Passerina and renoster and with

other succulents such as Aloe ferox, Crassula ericoides, C. perforata and C. rubricaulis are the favoured habitat between Plettenberg Bay and Port Elizabeth. In some of these it grows within a few hundred metres of E. clava, which appears to be restricted to shale outcrops and is mainly found in the denser ‘valley bushveld’ that develops on those soils.

Fig. 2.307.  Euphorbia pubiglans, ± 20 cm tall, on conglomerates next to road among dry fynbos, south of Hankey, South Africa, 24 Oct. 2012 (© PVB).

Fig. 2.308.  Euphorbia pubiglans, multi-branched plant ± 30 cm tall, south of Hankey, South Africa, 24 Oct. 2012 (© PVB).

Diagnostic Features & Relationships Plants of Euphorbia pubiglans are usually short and rarely exceed 20  cm tall. Young plants are single-stemmed and occasional older ones remain single-stemmed too. Nevertheless, older specimens usually consist of the stem and several very similar branches. These arise near the base of the stem to form a small shrub that may ultimately possess between 15 and 30 branches. The stem and branches are all stout and highly succulent, with a grey-green epidermis that changes gradually to a thin grey bark towards the

base. They are densely covered with crowded, short, conical and somewhat round-tipped tubercles, with very short bases. The tubercles are arranged into spiralling rows (that may spiral to the left or to the right) in the lower part of the stem or branch, but this organization is disturbed by the appearance of peduncles, after which the spirals disappear and the tubercles are randomly arranged. The upper half of the stem and branches are usually adorned with the old, slightly woody, dried-out peduncles from previous flowering seasons. Rains bring on the development of new tubercles at

232

2  Euphorbia subg. Athymalus

Fig. 2.309.  Euphorbia pubiglans, tubercles especially prominent, PVB 12215, near Plettenberg Bay, South Africa, 24 Oct. 2012 (© PVB).

Fig. 2.310.  Euphorbia pubiglans, large multi-branched plant ± 60 cm tall, south of Hankey, South Africa, 24 Oct. 2012 (© PVB).

the top of the stem and branches. Each new tubercle tapers into a fairly long, very slender, usually obtusely-tipped, greyish green leaf, all of them spreading in a rosette around the tip of the stem or branch. The leaves persist for a season and then fall off, leaving a quite prominent, almost U-shaped scar at the tip of each tubercle. Flowering appears to be dependent on rains and has been observed between January and June. The cyathia are borne on slender short-shoots, usually between 30 and 50  mm long, among the uppermost tubercles. Each short-shoot bears small, insignificant bracts widely spaced along its shaft followed by a whorl of much expanded and petal-like bracts that form a plate-like structure immediately under the solitary cyathium. The cyathium is shallowly cup-shaped, with the closely contiguous, puberulous glands forming a spreading, ring-like margin. Inside it has many inconspicuous male florets and a small, pubescent ovary with short styles divided nearly to their bases but hardly spreading at

all. The ovary develops into an almost spherical, finely pubescent capsule around which the bracts persist as driedout husks. Capsules dry out and dehisce in late spring and early summer between October and December. White et al. (1941) maintained E. pubiglans as distinct from E. clava on the basis of its smaller stature, the more densely crowded, blunt tubercles on the stems, the shorter fertile short-shoots with round-tipped bracts under the cyathia and a short pubescence on the spreading involucral glands. Further distinctions are the frequently shrub-forming habit of E. pubiglans, the differently shaped cyathial glands, the much longer styles in E. clava (divided only above a quite long, cylindrical fused part), the more or less glabrous ovary and glabrous male pedicels. The two appear to maintain their distinctness by their different flowering times, with E. pubiglans flowering mainly towards and after the end of summer and E. clava flowering in spring.

2.1 Sect. Anthacanthae

233

Fig. 2.311.  Euphorbia pubiglans, multi-branched but low-growing plant ± 12 cm tall, PVB 11914, Groot River Mountains, SE of Steytlerville, South Africa, 7 Jun. 2011 (© PVB).

Fig. 2.312.  Euphorbia pubiglans, cyathia with bracts, PVB 9986, Prince Alfred’s Pass, south of Uniondale, South Africa, 5 Jul. 2006 (© PVB).

History Euphorbia pubiglans was first collected by Carl Zeyher, who gathered both this species and E. clava in the vicinity of the Swartkops River, between Port Elizabeth and Uitenhage, where he lived for some years in the early 1830’s. The type

of E. pubiglans was collected by Isaac Louis Drège, the son of Carl F. Drège (Gunn and Codd 1981). Euphorbia pubiglans was erroneously subsumed under E. clava in Bruyns (2000, 2012).

234

2  Euphorbia subg. Athymalus

Fig. 2.313.  Euphorbia pubiglans. A, side view of cyathium with bracts (scale 2 mm, as for B). B, side view of cyathium with bracts removed. C, anthers and female floret in dissected cyathium (scale 1 mm, as for D). D, female floret. Drawn from: A–C, PVB 9986, Prince Alfred’s Pass, south of Uniondale, South Africa. D, PVB 11914, Groot River Mountains, SE of Steytlerville, South Africa (© PVB).

2.1.3 Subsect. Medusea Euphorbia subsect. Medusea (Haw.) Pax & K.Hoffm. in Engl., Nat. Pflanzenfam. ed. 2, 19c: 216 (1931). Euphorbia sect. Medusea (Haw.) Baill., Étude Euphorb.: 284 (1858). Medusea Haw., Syn. Pl. Succ.: 133 (1812). Type (designated by Wheeler 1943): Medusea major (Aiton) Haw. (= Euphorbia caput-medusae L.). Euphorbia subsect. Pseudomedusea (A.Berger ex Pax) Pax & K.Hoffm. in Engl., Nat. Pflanzenfam. 19c: 216 (1931). Euphorbia sect. Pseudomedusea A.Berger, Sukkul. Euphorb.: 10, 118 (1906). Type: Euphorbia procumbens Mill. Bisexual succulent shrubs or dwarf succulents often with a thick, mostly subterranean, flat-topped stem (sometimes with several swollen roots well beneath the surface) and more slender gradually deciduous branches radiating from near the apex of the stem (occasionally a single thick cylindrical stem or erect succulent stem and many similar branches), with green photosynthetic bark, stem and branches tuberculate, spineless (except in E. schoenlandii) but sometimes with persistent woody remains of fertile peduncles (rarely also some sterile ones, cf. E. arida) in axils of tubercles (especially long in E. fasciculata and E. restituta). Leaves spiralling, narrow and usually reduced to minute rudiments, entire, sessile, rapidly caducous, arising at apex of conical tubercle (tubercles arranged vertically into many spiralling rows along stem and branches); stipules absent. Synflorescences solitary in axils of tubercles, reduced to a single cyathium terminating a short, finely tuberculate

peduncle (peduncle to 15  cm long and branching among uppermost bracts to bear further cyathia only in E. fasciculata and E. restituta rarely in others). Cyathia all similar and bisexual, 2.5–15 mm diam.; glands 4–5, outer margins crenulate or with 2–many finger-like, sometimes branched projections. Capsule obtusely 3-angled, smooth to somewhat ridged along edges, glabrous to sparsely pubescent, 5–12 mm diam., as broad as tall to slightly broader than tall, subsessile (rarely exserted on pedicel to 5  mm long). Seeds 4-angled (rarely not angled), papillate to coarsely tuberculate. This subsection contains 33 species, all of which are endemic to southern Africa. Most of them are also endemic to South Africa, where the greatest diversity in this subsection is found inside and on the margins of the Greater Cape Floristic Region (with winter rainfall and dry summers) and in the Nama Karoo Region (rainfall mainly in summer). Elsewhere the diversity decreases greatly and there are only a few such widely distributed species as E. clavarioides. A few species (E. braunsii, E. davyi, E. duseimata and E. maleolens) reach the Kalahari ‘Desert’ of Botswana and the highlands of Namibia around Windhoek and into the Waterberg further north in Namibia. This subsection includes all the well known ‘medusoid’ succulents. These are characterized by a thick, flat-topped, coarsely tuberculate stem, from which more finely tuberculate branches spread out from the axils of the tubercles on the stem. In typical ‘medusoid’ plants, the stem varies from projecting partly above the ground (as often in E. esculenta) to wholly subterranean (as in E. gerstneriana). Plants with such

2.1 Sect. Anthacanthae

235

Fig. 2.314.  Distribution of Euphorbia subsect. Medusea (showing number of species per half-degree square) (© PVB).

underground stems can be geophytic, in the sense that the whole plant readily survives the dry season with all the branches removed. While the ‘medusoid’ form is typified most clearly by the widespread and quite variable E. caputmedusae (where individuals may exceed 1 m in diameter), there are also much smaller species such as E. procumbens, where the whole plant may be only 5 cm in diameter. However, other species that are not as clearly ‘medusoid’ also belong to this subsection. This includes species like E.

fasciculata, E. hallii, E. restituta and E. schoenlandii, all restricted to Namaqualand (of the Greater Cape Floristic Region) in western South Africa. Here the plant usually consists of a single, cylindrical stem (though the characteristic tuft of leaves at its apex is inconspicuous). A further deviation from the ‘medusoid’ form is found in E. braunsii and some forms of E. multiceps, where the plant forms a small shrub in which the stem is more or less indistinguishable from the branches.

Key to the species of subsect. Medusea 1. Stem mostly above ground (and usually tapering rapidly into fibrous roots beneath ground, though see large rootstock of E. hallii), erect and unbranched to sparsely branched, covered with comparatively large and broad-based tubercles..........2. 1. Stem mostly subterranean (and tapering off gradually into quite long tap-root from which fibrous or swollen roots arise), freely branched (sometimes very sparsely if wholly subterranean, see E. hypogaea), branches covered with comparatively small-based tubercles.............................................................................................................................5. 2. Sterile and fertile short-shoots present, fertile short-shoots usually arising in axil of tubercle between a sterile short-shoot and the tubercle....................................................................................................................................3. 2. Sterile short-shoots absent, fertile short-shoots arising alone in axils of tubercles...........................................................4.

236

2  Euphorbia subg. Athymalus

3. Tubercles low-conical and becoming indistinct lower on stem, stem 20–40 mm thick, rootstock a quite deeply located cluster of swollen tubers from which fibrous roots arise, male florets with glabrous pedicels, ovary exserted on pedicel 5–8 mm long, styles 7–10 mm long ......................................................................................................E. hallii 3. Tubercles tall-conical and remaining very prominent lower on stem, stem 40–200 mm thick, rootstock a shallow cluster of fibrous roots, male florets with pubescent pedicels, ovary ± included inside cyathium on pedicel ± 2 mm long, styles 0.8–1.5 mm long.........................................................................................................E. schoenlandii 4. Stem usually among several similarly thick branches, tubercles without cavity on upper surface (relative to stem) from which short-shoots arise, outer margins of cyathial glands with quite prominent, conspicuous teeth..............E. restituta 4. Stem almost always solitary and unbranched, tubercles with cavity on upper surface (relative to stem) from which shortshoots arise, outer margins of cyathial glands with tiny, inconspicuous teeth or entire................................E. fasciculata 5. Plant with stem hidden underground (either well below ground or hidden by dense mat of small branches)...................6. 5. Plant with stem visible on surface of ground (not hidden by dense mat of small branches) or protruding from ground (stem then sometimes indistinguishable from branches but usually considerably thicker than branches)............9. 6. Plant with large numbers of small, roughly spherical branches hiding subterranean stem and forming dense mat, each branch joined shortly to next level of branches or to stem................................................................................7. 6. Plant with few ellipsoidal to slender cylindrical branches and not forming dense mat......................................................8. 7. Cyathia and cyathial glands yellow to yellow-green, glands spreading and much more prominent than cyathial lobes.......................................................................................................................................E. clavarioides 7. Cyathia and cyathial glands green to reddish green, glands descending and as prominent as cyathial lobes.........................................................................................................................................E. willowmorensis 8. Plant with few ellipsoidal branches covered with slender prominent tubercles, branches with slender ‘necks’ joining them to underground stem, cyathia deeply cupular and styles 4–6 mm long with slender apices curled back...........E. hypogaea 8. Plant with few to many slender cylindrical branches with low obscure tubercles and joined shortly to hidden stem, cyathia shallowly conical and styles 1.5–2 mm long with short stout apices spreading..................E. gerstneriana 9. Mature plant a dwarf often cushion-shaped shrub with few to many branches ± indistinguishable from stem...............10. 9. Plant with thick ± club-shaped stem (apex flush with ground to protruding well above ground) tapering gradually beneath ground into slender tap-root, stem with erect, spreading to ascending much more slender (often cylindrical) branches over most of above-ground surface to branches (also much more slender than stem) radiating in rosette from around (often bare) apex of stem..................................................................................................................................12. 10. Styles 2–4.5 mm long, cyathia 5–12 mm broad (2–2.5 mm long below insertion of glands), pedicels of male florets not exceeding 5 mm long..............................................................................................................................11 10. Styles 7–9 mm long, cyathia 10–15 mm broad (4–7 mm long below insertion of glands), pedicels of male florets 7–9 mm long.............................................................................................................................................E. filiflora 11. Margins of leaves undulating, cyathia ± 2.5 mm long below insertion of glands, cyathial glands 0.7–2 mm broad, outer margins deeply divided into finger-like teeth not exceeding 1 mm long......................................E. braunsii 11. Margins of leaves not undulating, cyathia 2 mm long below insertion of glands, cyathial glands 1.5–3 mm broad, outer margins deeply divided into finger-like teeth often over 1 mm long.........................E. multiceps subsp. tanquana 12. Plant a typical medusoid ‘vingerpol’ with swollen clavate-truncate and densely tuberculate stem (often with slightly depressed apex) sunken into ground and with only its apex visible on surface or projecting from ground, with many more finely tuberculate branches ascending to spreading or in rosette on surface of ground, branches short to long and cylindrical to club-shaped, each arising in axil of tubercle on surface of stem, branches sometimes with persistent rapidly drying and soon spike-like simple fertile or sterile short-shoots.............................................................................13. 12. Plant not medusoid, with somewhat swollen tapering stem projecting from ground (often with several similar somewhat swollen tapering stem-like branches rather than one stem) bearing many spreading to erect branchlets (sometimes drying off at tips) each arising in axil of tubercle on surface of stem, plant often with several woody, often branched persistent spikes around apex...............................................................................................................................................................33. 13. Plant a ‘small vingerpol’ with mature specimens frequently only 50–100 (150) mm diam. (though see E. flanaganii), stem normally sunken into ground and only its apex visible on surface, branches cylindrical, often short and spreading in rosette on surface of ground, sometimes ascending or erect.......................................................................................14. 13. Plant a ‘large vingerpol’ with mature specimens usually over 150 mm diam., with prominent and thick stem (often with slightly depressed apex) usually projecting from ground and with many ascending to spreading, short to long and cylindrical to club-shaped branches..................................................................................................................................21.

2.1 Sect. Anthacanthae

237

14. Processes on outer margins of cyathial glands finger-like and about as long as lower entire part of gland, upper surface of gland green in lower half (to around base only) and above this cream or pinkish cream....................E. brakdamensis 14. Processes on outer margins of cyathial glands short and stubby and much shorter than lower entire part of gland to absent, upper surface of gland uniformly coloured, apart from finger-like processes..................................15. 15. Cyathia deeply conical-cupular and longer below glands than breadth just below glands, branches usually with many persistent fertile and sterile short-shoots often becoming spine-like ...............................................16. 15. Cyathia shallowly conical or shallowly cupular and much shorter below cyathial glands than breadth just below glands, branches without (or with only few) persistent fertile and sterile short-shoots becoming spine-like.............................17. 16. Plant with complex rootstock consisting of several swollen roots from base of stem, stem sometimes divided beneath ground, leaf-rudiments 4–10 mm long, ovary glabrous................................................................E. arida 16. Plant with simple rootstock where stem tapers off below the ground into fibrous roots (without swollen roots), stem not divided beneath ground, leaf-rudiments 1–3 mm long, ovary usually pubescent.................................E. decepta 17. Cyathial glands roughly circular with outer margin erect to inwardly curved and raised around depressed central area..................................................................................................................E. albipollinifera 17. Cyathial glands somewhat rectangular to wedge-shaped (not circular) with outer margins spreading to recurved...........18. 18. Cyathial glands strongly reflexed and usually with deep to shallow groove in inner margin....................E. procumbens 18. Cyathial glands horizontally spreading and with entire inner margins............................................................................19. 19. Cyathial glands bright yellow, often reaching 4 mm broad and closely contiguous, finely and regularly toothed on outer margin with teeth the same colour as gland.........................................................................E. flanaganii 19. Cyathial glands yellow-green to dark green or red to purple-brown, not exceeding 2.5 mm broad and distant around cyathium, outer margins entire or with teeth differently coloured to gland.............................................20. 20. Stem tapering gradually beneath surface into deep tap-root...............................................E. arida subsp. camdeboensis 20. Stem tapering rapidly beneath surface into short tap-root..........................................................................E. procumbens 21. Leaf-rudiments (6) 10–30 mm long, fairly conspicuous and present over most of summer growing-period.................22. 21. Leaf-rudiments mostly < 10 mm long, inconspicuous and rapidly caducous, though sometimes longer in E. caput-medusae but then present during winter-months only.......................................................................................25. 22. Many persistent dried-out spine-like short-shoots present on branches, female floret raised on pedicel 2.5–3 mm long and projecting from cyathium as males mature, styles 1.5–2.5 mm long...........................E. namibensis 22. Persistent dried-out short-shoots absent or very few on branches and not sharp-tipped, female floret sessile or only slightly raised on pedicel (< 0.5 mm long) and remaining included in cyathium as males mature, styles > 3 mm long........23. 23. Processes on outer margins of cyathial glands slender-tipped and considerably longer than solid part of gland............................................................................................................................................E. maleolens 23. Processes on outer margins of cyathial glands shorter than solid part of gland, often stubby.....................................….24 24. Peduncles 2–5 mm long and cyathia usually surrounded by and partly hidden by several persistent bracts, leaves to 4 mm broad, branches usually 12–20 mm thick......................................................................................E. davyi 24. Peduncles (5) 6–32 mm long with bracts very soon deciduous and not hiding cyathium, leaves to 2 mm broad, branches to 10 mm thick......................................................................................................................E. duseimata 25. Cyathium not densely furry or woolly above around female floret, bracteoles sparsely hairy towards tips only….....….26. 25. Cyathium densely furry or woolly above around female floret, from longer hairs concentrated near tips of bracteoles and on margins of lobes…..........................................................................................................................30. 26. Cyathial glands with conspicuous white to cream or pinkish cream often irregularly warty, ± truncate processes along outer margins, contrasting with and usually longer than solid portion of glands (green to red)….......................27. 26. Cyathial glands with relatively small finger-like to acute processes (shorter than solid portions) to irregularly finely toothed on outer margins, similarly coloured to paler than rest of gland but not strongly contrasting to it…..........28 27. Dark green basal part of cyathial gland wedge- or fan-shaped, processes on outer margins of glands often finely divided, with pustulate upper surfaces................................................................................E. caput-medusae 27. Dark green basal part of cyathial gland circular, processes on outer margins of glands short and stubby, with ± smooth upper surfaces.............................................................................................................................E. pentops 28. Cyathial glands uniformly yellow-green with finely crenulate outer margins, ovary ± sessile......................E. colliculina 28. Cyathial glands and processes usually differently coloured, processes finger-like, ovary raised usually around 1 mm on short pedicel……..................................................................................................................................29. 29. Plants often developing several rosettes of branches each on stout stem-like branch, branches usually with many persistent fleshy grey-green sometimes branched peduncles, tubercles arranged into 5 rows along branches, cyathial glands usually slightly longer than broad (excluding processes on outer margins)..................................…...E. audissoui

238

2  Euphorbia subg. Athymalus

29. Plants always with single rosette of branches on stem usually protruding substantially from ground, branches usually with many dried-out unbranched persistent peduncles, tubercles not organized into rows along branches, cyathial glands usually somewhat broader than long (excluding processes on outer margins)........................................E. crassipes 30. Processes on outer margins of cyathial glands white, extensively branched and antler-like, cyathial lobes spreading and not pressed onto tops of immature male florets...........................................................................E. inermis 30. Glands without white, antler-like processes, cyathial lobes inwardly inclined and pressed onto tops of immature male florets......................................................................................................................................................31. 31. Glands small (usually only ± 1 mm broad) and ± circular, inconspicuous and without processes on outer margins, often partly hidden by furriness of cyathium, bracteoles densely hairy only at apices....…...….....E. esculenta 31. Glands conspicuous, 2–3 mm broad and not circular, bracteoles hairy at apices and below..........................................32. 32. Glands red-green to purple, outer margins with fine whitish to reddish teeth....................................................E. fortuita 32. Glands bright yellow (turning orange with age), often radially grooved and sometimes with few coarse bright yellow teeth on outer margins....................................................................................................E. huttoniae 33. Young branches not tapering significantly towards tips, ovary glabrous, plants often with several woody, often branched persistent spikes around apex.................................................................................................................34. 33. Young branches distinctly tapering towards tips, ovary pubescent.................................................................................35. 34. Plant forming dense conical mound around solitary stem with many tightly packed branches decreasing in length towards apex of stem........................................................................................................................E. multiceps 34. Plant forming low, irregular mound around 1 stem or several stem-like branches................................E. melanohydrata 35. Short-shoots bearing cyathia sparingly tuberculate (like branches), arising at or near the tips of the branches mainly around apex of plant, often with tip of branch elongating into peduncle........................................E. friedrichiae 35. Short-shoots bearing cyathia densely tuberculate (more densely than branch) and arising in axils of tubercles on branches mainly in lower half of plant..................................................................................E. namaquensis

Euphorbia albipollinifera L.C.Leach, S. African J. Bot. 51: 281 (1985). Type: South Africa, Cape, Springbokvlakte, Dec. 1978, Bruyns 1826 (NBG, holo.; K, PRE, iso.). Bisexual spineless and glabrous dwarf succulent 20–50  mm tall with inverted top-shaped tuberculate stem 20–50 mm thick projecting from the ground for 10–40 mm, tapering beneath ground into tap-root bearing fibrous roots, with few to many branches forming rosette 30–90 mm diam. around its somewhat flat to slightly depressed apex, apex without branches. Branches ascending to spreading, simple, cylindrical and slightly thickening towards bases, 5–40 (75) × 4–8 (–10) mm, covered with tubercles, glabrous and smooth, brownish green; tubercles low and not prominent, conical, 1–4 mm long (much larger on stem than on branches) with elongated rhomboidal to hexagonal base, vertically arranged into 5–8 obscure spiralling rows, with fine grooves around them, rarely with persistent peduncles drying off into white spike-like remains (5–20 mm long) in some axils; leaf-rudiments on tips of new tubercles towards apices of branches and of stem, 1–4 × 1 mm, erect, fleeting, ovate to ovate-lanceolate, grey-green with reddish margins folded upwards and eciliate, acute, sessile. Synflorescences 1–7 per branch mainly towards apex, each of one bisexual cyathium (sometimes more developing if peduncle persists) in axil of tubercle on stout erect peduncle 3–20 mm long, with 3–6 slightly spathu-

late caducous bracts 0.5–2 mm long and glabrous except for cilia along margins near apex; cyathia shallowly cupular, glabrous, 5–7.5  mm broad (±  2  mm long below insertion of glands), with 5 reddish lobes with finely divided margins, pale green; glands 5, ± circular, 1.5–2 mm broad, spreading, dark green with paler margins, with shallow to deep depression in middle above, inner margins somewhat notched in middle, outer margins bent upwards, entire to very slightly divided into 3–5 short stubby teeth; stamens with pubescent pedicels, bracteoles filiform and pubescent; ovary globose, sparsely pubescent above, raised on erect pedicel ±  1  mm long; styles 1–2.5  mm long, branched nearly to base or in upper half, spreading. Capsule 6  mm diam., obtusely 3-angled, sparsely pubescent to glabrous, grey to dull green, sessile. Distribution & Habitat Euphorbia albipollinifera is a rarely collected species which has been recorded scantily in the drier parts of the Eastern Cape from near Kirkwood via Steytlerville to a little southeast of Willowmore. Plants usually occur in flattish areas or on gently sloping, low hills among pebbles and other small stones, often protected by the stones and short bushes. They are usually quite scattered in occurrence and may be very cryptic.

2.1 Sect. Anthacanthae

239

Fig. 2.315.  Distribution of Euphorbia albipollinifera (© PVB).

Fig. 2.316.  Euphorbia albipollinifera, next to small shrub, with Haworthia bruynsii, PVB 11516, west of Steytlerville, South Africa, 22 Oct. 2009 (© PVB).

Diagnostic Features & Relationships Plants of E. albipollinifera are small, with a small rosette of short branches radiating from an almost spherical or flattopped stem. The stem is usually sunken into the ground, often with only the apex projecting from the soil. A little above the surface of the ground, several short, mostly spread-

ing branches are produced between the tubercles, usually leaving the centre of the stem bare. Flowering takes place during the months of spring and summer, though in cultivation cyathia develop from August through most of the summer. The fairly small, faintly sweetly scented cyathia are produced at the tips of obscurely tubercu-

240

2  Euphorbia subg. Athymalus

Fig. 2.317.  Euphorbia albipollinifera, among rocks, PVB 12719, SE of Willowmore, South Africa, 26 May 2014 (© PVB).

late peduncles. Around their margins there are five almost circular to faintly toothed glands, whose centre is usually depressed. Euphorbia albipollinifera is surrounded by two other, similar-looking species, E. decepta and E. procumbens, between which it is in some senses intermediate. While plants of E. decepta may be of a similar size when growing in shallow, very stony ground, in deeper ground the stem may become considerably larger and project from the ground so that its apex and many of the branches are well above the

surface. In E. decepta there are also more branches and they are more crowded right to the apex of the stem. The branches may become longer and more erect and have many persistent woody peduncles along them, especially towards their tips. The cyathia are deeper (more cup-shaped) in E. decepta, with correspondingly longer fused parts of the styles (and a somewhat mushroom-shaped apex to the styles), the cyathial glands may lack the depression in their middle and generally have more conspicuous, more widely spreading teeth around their margins.

Fig. 2.318.  Euphorbia albipollinifera, shallow cyathia with yellow pollen, PVB 11516, west of Steytlerville, South Africa (© PVB).

Fig. 2.319.  Euphorbia albipollinifera, shallow cyathia with some white pollen, PVB 9795, east of Steytlerville, South Africa (© PVB).

Euphorbia procumbens shares the relatively shallow cyathium of E. albipollinifera with short styles that are generally deeply divided. In E. procumbens the cyathial glands are spreading to deflexed, usually ornamented with much more

conspicuous, differently-coloured, finger-like teeth around their outer margins, while their upper surface is convex to almost flat and lacks the depression above of E. albipollinifera that is sometimes present in E. decepta too.

2.1 Sect. Anthacanthae

241

Fig. 2.320.  Euphorbia albipollinifera, cyathia, PVB 9795, east of Steytlerville, South Africa (© PVB).

Leach (1985) thought that the white pollen in the type collection of E. albipollinifera was unique. In the interim, it has been found (first reported by Marx 1987, 1994) that yellow pollen is produced in some populations of what is otherwise typical E. albipollinifera. White pollen is known on plants from near Uitenhage (with glands not circular and with considerably more toothed outer margins and there-

fore here included under E. procumbens (e.g. PVB 10624, near Addo). Deep yellow pollen is known in material from west of Steytlerville (PVB 11516) that is included here under E. albipollinifera. White pollen has also been found in some collections from west and east of Jansenville (PVB 12987, 12994) that are included here under E. decepta.

Fig. 2.321.  Euphorbia albipollinifera. A, B, cyathium from side (scale 2 mm, as for B, C). C, side of cyathium and glands from above. D, side view of dissected cyathium (scale 1 mm, as for E, F). E, F, female floret. Drawn from: A, C, E, PVB 11516, west of Steytlerville, South Africa; B, D, F, PVB 9795, east of Steytlerville, South Africa (© PVB).

2  Euphorbia subg. Athymalus

242

History Euphorbia albipollinifera was described in 1985 from material collected on 9 December 1978. However, similar plants had been seen long before this. Some were found by G.G. Smith in May 1946 (near Springbokvlakte, Smith 6352 at PRE, recorded only in a photograph taken by R.A. Dyer in June 1946) and others were observed by R.A.  Dyer in July 1947 at a spot ‘Between Steytlerville and Uitenhage near Twowaters on very stony rise frequent with Haworthia retusa and other dwarf succulents 5/7/47 RAD’ (photos at K and PRE). No herbarium material was made from either collection, but Dyer annotated these photographs as E. brevirama. Euphorbia arida N.E.Br., Fl. Cap. 5 (2): 319 (1915). Type: South Africa, Cape, Britstown div., near De Aar, Schönland (K, holo.). Bisexual spineless glabrous succulent 50–150 (–200) × 50–100  mm with globose to slenderly clavate tuberculate greyish to brownish green stem 20–100 mm thick (sometimes several stem-like branches together originating from stem well beneath ground) projecting from the ground for 30–200 mm, tapering beneath ground into usually several somewhat swollen roots bearing fibrous roots, with many ascending branches forming rosette 40–150  mm diam. around apex, apex without branches. Branches ascending to spreading, simple, cylindrical and slightly thickened towards bases, 5–50 × 4–12 mm, covered with tubercles, smooth, greyish to brownish green; tubercles low and not prominent, conical,

1–4 mm long (much broader on stem than on branches) with elongated rhomboidal to hexagonal base, vertically arranged into 5–8 obscure spiralling rows, with fine grooves around them, usually with many spreading slender but firm and slightly woody persistent sterile and fertile short-shoots in axils of tubercles (sterile short-shoots more numerous than fertile, drying out white and often sharp and spike-like before gradually becoming grey and wearing off); leaf-rudiments on tips of new tubercles towards apices of branches and of stem, (2) 4–10 × 1–1.5 mm, erect, fleeting, linear-lanceolate, greygreen, reddish margins folded upwards and eciliate, obtuse to acute, sessile. Synflorescences 1–8 per branch mainly towards apex, each of one bisexual cyathium arising in axil of tubercle on short-shoot 6–25  mm long, with 3–5 slightly spathulate caducous bracts 0.5–2.5  mm long and glabrous except for cilia along margins near apex; cyathia deeply conical-cupular to rounded-cupular, glabrous, 4–8 mm broad (± 3.5 mm long below insertion of glands), with 5 lobes with finely divided cilate margins, pale green finely dotted with red; glands 5, ±  cuneate to transversely rectangular, 1.5–2.5  mm broad, spreading and widely separated, dark green to purple-red, concave above, inner margins somewhat raised, outer margins usually divided into 2–6 short stubby dark green to cream teeth; stamens with pubescent pedicels, bracteoles filiform and pubescent; ovary globose-ellipsoidal, glabrous or pubescent near apex, sessile to raised on very short pedicel to 0.5 mm long; styles 1.5–5 mm long, branched near apex or to near middle into broad spreading slightly bifid lobules. Capsule 6–8  mm diam., obtusely 3-angled, glabrous, greygreen, sessile.

Fig. 2.322.  Distribution of Euphorbia arida (● = ssp. arida; ▲ = ssp. camdebooensis) (© PVB).

2.1 Sect. Anthacanthae

243

Two subspecies are recognised: 1. Branches with many persistent spiky often white remains of short-shoots, styles 3–5 mm long and branched only near apex……………..............................................................................…………………subsp. arida 1. Branches mostly without spiky remains of short-shoots, styles 1.5–2 mm long and branched to middle.....................................................................................................................subsp. camdebooensis

Euphorbia arida subsp. arida Branches usually with many persistent usually white spike-like dried-out remains of short-shoots. Cyathia deeply

conical-cupular to rounded-cupular, glands with outer margins usually divided into 2–6 short stubby dark green to cream teeth, ovary sessile, styles 3–5 mm long, branched only near apex into broad spreading slightly bifid lobules.

Fig. 2.323.  Euphorbia arida ssp. arida, without persistent sterile peduncles on the branches, PVB 6627, SW of Bloemfontein, Free State, South Africa, 10 Apr. 2014 (© PVB).

Distribution & Habitat Euphorbia arida subsp. arida has been recorded on the north-eastern parts of the Great Karoo from Biesiespoort near Victoria West (in the south-west) and Murraysburg (in the south-east) to De Aar, Hopetown and Colesberg in the former Cape Province. In the adjoining part of the Free State it has been recorded frequently around Lückhoff and Fauresmith and occurs to about 60  km south-west of Bloemfontein. In habitat, plants of subsp. arida are generally inconspicuous and are usually rather scattered. They have been observed on flat to gently sloping, stony banks of calcrete or shale or even on baked shales near low doleritic dykes but have also been seen in deeper, loamy ground in flat, some-

times almost pan-like areas. Usually they grow among a sparse cover of short shrubs with loose pieces of calcrete or shale on the ground between the bushes. The shrubs may be Ruschia spinosa, species of Eriocephalus, Felicia and Pentzia or Salsola, depending on how calcareous the terrain is. Specimens of subsp. arida are frequently found sheltering near the base of these shrubs or alongside a stone and only the larger and more venerable plants occasionally grow fully in the open. Diagnostic Features & Relationships In subsp. arida the whole plant is usually not more than 10 cm in diameter, with a stem (usually not more than 60 mm thick) from which several short branches ascend in an often

244

2  Euphorbia subg. Athymalus

Fig. 2.324.  Euphorbia arida ssp. arida, with many persistent white sterile peduncles on the branches, PVB 11022, NW of Colesberg, South Africa, 26 Apr. 2008 (© PVB).

Fig. 2.325.  Euphorbia arida ssp. arida, 3 km south of Lückhoff, Free State, South Africa, 27 Apr. 2008 (© PVB).

very neat rosette. The stem descends for 150  mm or more into the ground and, what appears to be the stem, may itself prove to be one of a few branches (rarely more than five and mostly one to three) arising from an even deeper-seated stem. This branching is very likely to be caused by damage to the apex of the original stem by animals like porcupines. Such damage gives rise to what appear to be several plants clustered closely together that actually constitute a single specimen and are connected below the surface to the same

stem. Beneath the stem there is usually a thickened roostock (which, together with the stem may sometimes be up to 20 cm long) spreading out or descending vertically further into the ground and, from these, the finer, fibrous roots arise. The branches rarely exceed 40  mm long and often have a distinctive grey-green colour. As in several others, in subsp. arida the branches usually bear many slender persistent spike-like structures a little behind their apices. Some of these are derived from fertile

2.1 Sect. Anthacanthae

Fig. 2.326.  Euphorbia arida ssp. arida, showing long underground stem and taproot, PVB 11022, NW of Colesberg, South Africa, 26 Apr. 2008 (© PVB).

Fig. 2.327.  Euphorbia arida ssp. arida, cyathia, PVB 11022, NW of Colesberg, South Africa (© PVB).

245

short-shoots and others from sterile ones. These may be sufficiently dense to form a protective covering over the plant and are often fairly rigid, somewhat sharp and spike-like, though they soon wear off and are absent lower down on the branches. At the stage when they are sharp and rigid, they are often white and may be the most conspicuous feature of the plant, contrasting with the dull colour of the rest of the plantbody. The comparatively long, but short-lived leaf-rudiments are borne near the apices of the branches. Flowering takes place during summer, more or less whenever rain is received. The cyathia are usually not produced in large numbers and their dark green to purple glands with short and stubby teeth on the margins add to their inconspicuousness. They are quite variable in shape, from conical to rather swollen and more cupular. One of several ‘dwarf medusiod’ species, both subspecies of E. arida are similar in appearance to E. albipollinifera, E. decepta and E. procumbens, in that the plants are usually small and neat, with a stout stem and many short, slender, ascending branches. In E. arida the stem neither protudes as far from the ground, nor does it become as thick towards its apex as in E. decepta, where it is often hemispherical above the soil. Below the surface there is a more complex rootstock in E. arida than in the others. This consists of a thickened roostock below the stem (and the stem itself may be divided, as mentioned above), whereas in E. decepta the stem tapers off below the ground into fibrous roots and never forms any swollen roots. The respective stems and branches are different in colour, with those of E. decepta often a distinctly brownish green and those of E. arida more greyish. The leafrudiments in E. arida are usually much longer than those of

246

2  Euphorbia subg. Athymalus

Fig. 2.328.  Euphorbia arida ssp. arida, with a few cyathia and capsules, PVB 12459, near Britstown, South Africa, 25 Mar. 2013 (© PVB).

E. decepta, though they are similar in shape. Florally there is little to distinguish them. Both have small, often conical and comparatively deep cyathia (relative to the much shallower cyathia of, say, E. procumbens) with dark green to purple, flat to almost convex glands with short stubby teeth around

their outer edges. In both the female floret (usually pubescent on the ovary in E. decepta and often glabrous in E. arida) is sessile or nearly sessile, bearing a fairly long, cylindrical style that is divided only near its apex.

Fig. 2.329.  Euphorbia arida ssp. arida. A, tip of branch with leaf-rudiments (scale 1 mm, as for F–H). B, C, cyathium from side (scale 2 mm, as for C–E). D, E, cyathium from above. F, side view of dissected cyathium. G, anther and bracteole. H, female floret. Drawn from: A, B, D, F, PVB 4234, Biesiespoort, near Victoria West, South Africa; C, E, G, H, PVB 11022, NW of Colesberg, South Africa (© PVB).

2.1 Sect. Anthacanthae

History Euphorbia arida subsp. arida was described from a plant collected near De Aar by Selmar Schönland and sent to N.E. Brown at Kew in 1911. This plant grew there and flowered in the garden in June 1913. White et al. (1941) recorded what they considered to be other forms of it, gathered by Dyer near Philippolis. These differed in the length for which the styles were fused, but perhaps more markedly in being much spikier than the type plant, with white, persistent short-

247

shoots. Although the figure made from the original plant sent by Schönland and cultivated at Kew (Fig. 2.330) bears a certain resemblance to E. crassipes (except for the fact that, according to N.E.  Brown’s measurements, it was much smaller than a mature specimen of E. crassipes), the comparatively deep cyathia and relatively long style shown makes it clear that this plant was not E. crassipes (as Brown also pointed out). The comparative lack of white, spiny, persistent short-shoots on this plant is probably due to its having

Fig. 2.330.  The type-plant of Euphorbia arida ssp. arida, Schönland, watercolour by Matilda Smith, with annotations by N.E. Brown (© Royal Botanic Gardens, Kew).

been cultivated for at least two years in England before it flowered. However, this is not the earliest collection of subsp. arida. Plants collected early in March 1842 (Gunn and Codd 1981) ‘on the southern side of the Snowy Mountains, Richmond Div.’ by Joseph Burke and preserved in the Kew Herbarium belong to subsp. arida, so that Burke was actually the discoverer of this species. N.E.  Brown identified this as E. decepta, having matched it with another specimen in the herbarium at Lübeck, collected by Drège in 1827 at the Camdeboo Mountains, Aberdeen Division. Drège’s specimen has not been located, so its identity remains uncertain. A plant of subsp. arida, collected by C.A.  Smith near Lückhoff, Free State, was figured as ‘E. fusca’ in Phillips

(1929c). Its relatively deeply embedded stem, small cluster of branches and deeply cupular cyathia make it clear that this was not E. crassipes (= E. fusca). Euphorbia arida subsp. camdebooensis Bruyns, Haseltonia 25: 31 (2018). Type: South Africa, Eastern Cape, Cradock, James sub NBG 615/37 (BOL, holo.). Branches usually without dried-out spike-like remains of short-shoots. Cyathia conical, glands with outer margins usually divided into 2–6 very short stubby dark green to cream teeth, ovary raised on very short pedicel ±  0.5  mm long, styles 1.5–2  mm long, branched to near middle into broad spreading slightly bifid lobules.

248

2  Euphorbia subg. Athymalus

Fig. 2.331.  Euphorbia arida ssp. camdebooensis, among low shrubs, PVB 11649a, SE of Cradock, South Africa, 5 Aug. 2011 (© PVB).

Distribution & Habitat Subsp. camdebooensis is known south-east of Cradock and west of Graaff-Reinet. Here it is found in relatively deep loamy ground, sometimes with submerged rocks, in flat areas at the foot of hills, often with many other small succulents including various small succulent Aizoaceae, several species of Crassula, Senecio radicans and forms of E. patula around Cradock. Diagnostic Features & Relationships As in subsp. arida, in subsp. cambebooensis the stem may be eaten off to below the ground and then give rise to several

Fig. 2.332.  Euphorbia arida ssp. camdebooensis, PVB 11649a, SE of Cradock, South Africa (© PVB).

stem-like branches, each with a rosette of radiating branchlets. In subsp. cambebooensis the plant is smaller than in subsp. arida and the branchlets are often (but not always) prostrate. The short-shoots bearing cyathia are mostly not persistent (some do sometimes remain, especially in cultivated plants), so the plant never has the spiky appearance that is typical of subsp. arida. In subsp. cambebooensis the cyathia are shallower than in subsp. arida, but have very similar, spreading, toothed glands. The female floret is generally shorter and slightly raised above the base on a short pedicel.

2.1 Sect. Anthacanthae

History Euphorbia arida subsp. cambebooensis was discovered around Cradock in 1937 by H.R.  James, but his collection seems to have remained unknown to White et  al. (1941).

249

Drège’s specimen of 1827 from the Camdeboo Mountains, Aberdeen Division may have belonged to this subspecies, but this remains uncertain.

Fig. 2.333.  Euphorbia arida ssp. camdebooensis. A, side view of cyathium (scale 2 mm). B, side view of female floret (scale 1 mm). Drawn from: Bruyns 11649a, SE of Cradock, South Africa (© PVB).

Euphorbia audissoui J.G.Marx, Euphorbia World 10: 23 (2014). Type: South Africa, Western Cape, east of Albertinia, ± 60 m, Marx 798 (GRA, holo.). Bisexual spineless glabrous succulent 50–150  mm tall with very stout clavate tuberculate stem 50–160  mm thick sometimes projecting from the ground for up to 50 mm, tapering gradually beneath ground into tap-root bearing fibrous roots, with many branches forming rosette 100–250 (400) mm diam. around obtuse apex, often stem branched and up to 15 rosettes arising from one stem and densely clustered into mound to 0.5  m diam. Branches ascending to spreading, simple, obscurely 5-angled, 30–120 (200) × 8–12 mm, densely covered with small tubercles, smooth, greyish green; tubercles low and not prominent, conical, 1 mm long with ± hexagonal base, arranged into 5 low slightly spiralling rows, with fine grooves around them, with blunt clavate persistent green sometimes rebranching peduncles 5–40  mm

long in their axils towards apices of branches (gradually drying up and falling off lower down); leaf-rudiments on tips of new tubercles towards apices of branches, 2–5 × 2 mm, canaliculate above and recurved, fleeting, subulate, margins reddish and slightly folded upwards and eciliate, shortly acute, sessile. Synflorescences 1–30 per branch around apex, each of one bisexual cyathium in axil of tubercle on slightly tuberculate peduncle 3–5 mm long, with 3–5 slightly spathulate glabrous caducous bracts 2–3  mm long and pilose along margins towards apex, sometimes additional cyathia developing from axils of tubercles at base of first; cyathia conical, glabrous, 4.5–7 mm broad (± 2 mm long below insertion of glands), with 5 glabrous reddish lobes with fimbriate margins, reddish green to green; glands 5, ± elliptic to circular, 2–3  mm broad, widely separated, slightly to strongly reflexed, maroon to brownish green to yellow, convex and smooth above, outer margins with 3–5 spreading irregularly finger-like paler sometimes very short processes; stamens

250

with pubescent pedicels, bracteoles filiform and pubescent towards apex; ovary globose, sparsely pubescent, raised on pedicel 1.5–2 mm long; styles ± 1.5 mm long, branched only near apex. Capsule 5–7 mm diam., obtusely 3-angled, glabrous to sparsely pubescent, shiny bright green, raised on pedicel ± 2 mm long. Distribution & Habitat Euphorbia audissoui is only known in the dry and deeply incised country along the Gouritz River, east of the small

Fig. 2.334.  Distribution of Euphorbia audissoui (© PVB).

Fig. 2.335.  Euphorbia audissoui, among sparse shrublets on gentle north-facing slope in hard loam, two fairly young plants each ± 15 cm diam., PVB 13747, east of Albertinia, South Africa, 17 Sep. 2019 (© PVB).

2  Euphorbia subg. Athymalus

town of Albertinia and south of the Langeberg. Here it occurs at altitudes of 50 to 100  m on gently sloping, north-facing ground in heavy loamy soil. Higher on these slopes the vegetation may be dense and up to 2 m tall, but where this species occurs the bushes are low, mostly 0.3 m tall or less. Along with Aloe ferox, the vegetation contains many succulents, with Euphorbia burmanni, E. clandestina and E. mauritanica nearby and E. heptagona and E. mammillaris not far away in steeper, more rocky places.

2.1 Sect. Anthacanthae

Diagnostic Features & Relationships This species is a typical large medusoid Euphorbia, with rosettes of branches often around 150 mm in diameter. In this species, many plants develop up to 15 rosettes grouped tightly together to form a somewhat heaped clump, all arising from the same rootstock. With age the stem and rosette-

251

bearing branches develop into stout, nearly spherical bodies (as in Fig. 2.337). Also unusual in this species is the manner in which the tubercles on the branches are arranged vertically into five rows. The branches are then somewhat 5-angled, though these angles become progressively more obscure towards the bases of the branches.

Fig. 2.336.  Euphorbia audissoui, older plant nearly 50 cm diam. consisting of more than 13 densely packed rosettes of branches, other plants in background, PVB 13804, east of Albertinia, South Africa, 25 Mar. 2020 (© PVB).

Fig. 2.337.  Euphorbia audissoui, swollen branches (largest ± 15 cm thick) of dead plant, PVB 13804, east of Albertinia, South Africa, 25 Mar. 2020 (© PVB).

Flowering was observed after rain in mid-September 2019 and also in March 2020. The cyathia give off a sweet scent that is noticeable over a considerable distance when many plants are in flower. The cyathia are borne on short, stout tuberculate peduncles. These peduncles often persist for several seasons and further cyathia may develop on sec-

ondary or tertiary peduncles that arise around the bases of the first and later cyathia, so that peduncles may be repeatedly branched (as in Fig. 2.343). Peduncles remain on the upper 50 to 80 mm of the branches, later drying out and falling off, so that none are present lower on the branches. In the cyathia, the colour changes somewhat with age, starting off

252

2  Euphorbia subg. Athymalus

Fig. 2.338.  Euphorbia audissoui, cyathia with variably dark glands, the darkest being the youngest, PVB 13804, east of Albertinia, South Africa, 25 Mar. 2020 (© PVB).

as dark maroon on the cyathial glands and red on the lobes, fading to paler maroon or even green or yellow on the glands and reddish green on the lobes. The edges of the glands are ornamented with small, irregularly finger-like processes

which are usually paler than the rest of the gland, though they may have dark patches on them as well. Euphorbia audissoui is similar to E. colliculina and E. fortuita from the Klein Karoo and E. inermis from further

Fig. 2.339.  Euphorbia audissoui, old cyathia with yellow glands, PVB 13804, east of Albertinia, South Africa, 25 Mar. 2020 (© PVB).

east, with the same grey-green colour of the branches as E. fortuita and E. inermis. Persistent fleshy peduncles are also found occasionally in E. colliculina and in E. inermis. However, in both of them the female floret is sessile or raised on a much shorter pedicel and the styles are longer than in E. audissoui. The finger-like projections on the outer margins of the cyathial glands in E. audissou are most similar to those in E. fortuita, but they are mostly shorter in E. fortuita.

History Euphorbia audissoui was first noticed by a French succulent-collector, Jean-André Audissou in 2007, alongside the road just east of Albertinia. There seem to be no previous records of this species, despite the work of Fourcade in the vicinity of George and Humansdorp and the investigations of Muir in the area around Riversdale.

2.1 Sect. Anthacanthae

Fig. 2.340.  Euphorbia audissoui, cyathia with variably coloured glands and shiny capsules beginning to develop, PVB 13804, east of Albertinia, South Africa, 25 Mar. 2020 (© PVB).

253

Fig. 2.341.  Euphorbia audissoui, slightly dull and sparsely pubescent capsules, PVB 13804, east of Albertinia, South Africa, 25 Mar. 2020 (© PVB).

Fig. 2.342.  Euphorbia audissoui. A, tip of branch with leaf-rudiments (scale 3 mm). B, cyathium from side (scale 2 mm, as for C). C, cyathium from above. D, anthers, bracteoles and female floret (scale 1 mm, as for E). E, female floret. Drawn from: A, E, PVB 13804, east bank of Gouritz River, east of Albertinia, South Africa. B, C, D, PVB 13747, west of Gouritz River, east of Albertinia, South Africa (© PVB).

254

2  Euphorbia subg. Athymalus

their axils; leaves on tips of new tubercles towards apices of branches and of stem, 3–15 × 1–2 mm, ascending and often distinctly bent towards stem or branch near base, fleeting, linear, longitudinally folded upwards, very shortly acute, sessile. Synflorescences 1–10 per branch towards apex, each of one bisexual cyathium in axil of tubercle terminating peduncle 10–30 mm long, with 3–6 oblong caducous bracts 2–3 mm long glabrous except for cilia along margins; cyathia conical, glabrous, 8–12  mm broad (5–7  mm long below insertion of glands), with 5 glabrous lobes with deeply incised margins, pale green; glands 5, ±  cuneate, 2–3  mm long, not contiguous, glabrous, green in lower half to around base and above this cream or pinkish cream, slightly convex and rugose above, outer margins divided into (0) 1–4 (–5) spreading to slightly recurved finger-like processes; stamens with pubescent pedicels, bracteoles filiform and pubescent; ovary ellipsoidal, glabrous, sessile; styles 4–6  mm long, branched in upper third. Capsule 6–7  mm diam., obtusely 3-angled, glabrous, grey-green, sessile.

Fig. 2.343.  Euphorbia audissoui, tip of branch with persistent peduncles, some also branched, PVB 13747, east of Albertinia, South Africa, 17 Sep. 2019 (© PVB).

Euphorbia brakdamensis N.E.Br., Fl. Cap. 5 (2): 324 (1915). Type: South Africa, Cape, Namaqualand, Brakdam, 1600’, Schlechter 11123 (K, holo.; BOL, BR, COI, GRA, HBG, L, PRE, S, WAG, iso.). Bisexual spineless and glabrous succulent 50–200  mm tall with clavate tuberculate stem 30–80  mm thick sunken into ground (usually rising 10–50  mm above surface) and tapering beneath into tap-root bearing fibrous roots, with many branches towards apex. Branches ascending, simple to occasionally rebranched, cylindrical to slightly clavate, 8–20 mm thick, covered with tubercles, glabrous and smooth, grey-green; tubercles prominent and conical, 3–5 mm long with elongated hexagonal base, vertically arranged into 5–7 obscure spiralling rows, with fine grooves between them, occasionally with solitary simple persistent peduncles in

Distribution & Habitat Euphorbia brakdamensis is restricted to Namaqualand, where it is found in both the drier parts and in some of the wetter areas. It is widely, though very sparsely distributed in the hills of the coastal plain from south of the Groen River to near Wallekraal, but it is not found in the sandveld at all. Records also exist from the Kamiesberg around Kamieskroon and it is widespread on the northern margins of the Knersvlakte, from east of Garies to Kliprand. Habitats vary from gentle gneissic slopes dominated by shrubs of Pteronia incana to hard loam among low succulent Aizoaceae (especially Cephalophyllum and various species of Ruschia), sometimes with some scattered quartz pebbles. On the northern Knersvlakte it grows in hard loam in depressions between shale hills among shrublets of Eriocephalus and Pteronia. Diagnostic Features & Relationships In Euphorbia brakdamensis the swollen tuberculate stem is sunken into the ground, tapering off gradually into a slender tap-root, with only its upper surface projecting above the surface and with a rosette of ascending, much more slender branches surrounding its apex. These branches persist for a few seasons after which they dry up and fall off. Very slender leaves are produced on the young tubercles during the growing season in the months of winter. Euphorbia brakdamensis is similar to E. caput-medusae (see also E. pentops). In Namaqualand, E. caput-medusae mainly occurs in deep sands, while E. brakdamensis is always found in hard, loamy ground among karroid vegeta-

2.1 Sect. Anthacanthae

255

Fig. 2.344.  Distribution of Euphorbia brakdamensis (© PVB).

Fig. 2.345.  Euphorbia brakdamensis, on gentle slope in hard loam with Zygophyllum, ± 15 cm diam., PVB 10819, west of Garies, Namaqualand, South Africa, 5 Jul. 2007 (© PVB).

tion. The two differ vegetatively in that E. brakdamensis is a much smaller plant with generally much shorter branches. The central stem also never reaches the diameter that may be achieved in E. caput-medusae. The slender leaf-rudiments are usually longer in E. brakdamensis than in E. caput-medusae and are always noticeably bent towards the branch at

their bases, though this may occasionally be found in E. caput-medusae too. In E. brakdamensis the cyathia are borne on long and slender peduncles. In some localities these peduncles persist after drying out, without becoming spiny, while in most they disappear from one season to the next. The peduncles are

256

2  Euphorbia subg. Athymalus

Fig. 2.346.  Euphorbia brakdamensis, in full flower, with Ruschia and Crassula expansa ssp. pyrifolia growing among the branches, PVB 10819, west of Garies, Namaqualand, South Africa, 12 Sep. 2007 (© PVB).

Fig. 2.347.  Euphorbia brakdamensis, among clumps of Ruschia, PVB 10819, west of Garies, Namaqualand, South Africa, 12 Sep. 2007 (© PVB).

generally much shorter and stouter in E. caput-medusae, though there are areas (such as in the red sands north of Vanrhynsdorp) where the peduncles in E. caput-medusae may reach 6 cm long and even persist on some plants for a few seasons. In E. caput-medusae the cyathium is considerably shorter below the glands than it is in E. brakdamensis (so that, in this respect, E. brakdamensis is more similar to E. filiflora). In both species the upper surfaces of the glands are generally green in the lower half and cream above. A further difference lies in the finger-like processes on the outer mar-

gins of the glands. In E. brakdamensis these processes are fewer (sometimes they are entirely lacking) and straighter than in E. caput-medusae and they are also not finely tuberculate and subdivided as they often are in E. caput-medusae. Euphorbia brakdamensis flowers later than E. caputmedusae, with most plants flowering from early in September until late October, though specimens will be found occasionally flowering as early as April to June. The cyathia are faintly sweetly scented.

2.1 Sect. Anthacanthae

257

Fig. 2.348.  Euphorbia brakdamensis, with very variably coloured appendages on the cyathial glands, PVB 10819, west of Garies, Namaqualand, South Africa, 12 Sep. 2007 (© PVB).

Fig. 2.349.  Euphorbia brakdamensis, in fruit in gravelly gneissic ground, ± 15 cm diam., PVB 10480, west of Khamieskroon, Namaqualand, South Africa, 28 Oct. 2006 (© PVB).

In Bruyns et  al. (2006), E. brakdamensis was included under E. filiflora. Careful examination of Schlechter’s many pressings of the type collection of E. brakdamensis, shows, however, that this is not correct. In E. filiflora the stem and branches are very similar in shape and thickness, with the stem usually slightly longer than the branches, if it can be located at all. In E. brakdamensis, on the other hand, the branches are very much more slender than the stem, which is largely buried in the ground and is then exceeded in height by the branches. Euphorbia filiflora has unusually long cyathia (often around 8 mm long), with especially long styles

(7–9 mm long), long male pedicels and also unusually prominent cyathial lobes which form a relatively tight, cone-like structure around the florets and rises well above the level of the glands. The cyathia in E. brakdamensis do not exceed 5 mm long and the styles are not longer than 6 mm, the pedicels of the male florets are shorter and the cyathial lobes do not rise above the level of the glands. The marginal processes on the glands in E. brakdamensis are much more brightly coloured than in E. filiflora, but they are considerably longer, more slender and far more numerous in E. filiflora.

258

2  Euphorbia subg. Athymalus

Fig. 2.350.  Euphorbia brakdamensis. A, tip of branch with leaf-rudiments (scale 4  mm). B, cyathium from side (scale 3  mm, as for C). C, cyathium from above. D, anther and bracteole (scale 1 mm, as for E). E, female floret. Drawn from: PVB 3841, SE of Wallekraal, Namaqualand, South Africa (© PVB).

History Euphorbia brakdamensis was first recorded by Rudolf Schlechter on 7 September 1897 on the farm Brakdam, which lies along the present-day national road (N7) between Garies and Kamieskroon in Namaqualand. He made at least ten specimens from this collection, picking off many of the branches on some plants and pressing these, but he did not press any of the stems. There has been some uncertainty about the application of this name. This was particularly the case in White et  al. (1941) and their figures 494 and 496 are of E. filiflora. These photographs were taken of plants at Ratelpoort. Ratelpoort also lies along the N7, but it is over 150 km north of Brakdam, between Springbok and Steinkopf and is therefore not ‘near the type locality of E. brakdamensis’ as they believed (p. 483). Leach (1986a) pointed out this confusion but did not clarify the identity of E. brakdamensis. N.E.  Brown (1915) also suggested that a collection of Drège from between Mamre and Saldanha Bay (Drège 8202a at K) could represent ‘a very closely related species’ but this is a collection of some sterile branches of E. caput-medusae. Euphorbia braunsii N.E.Br., Fl. Cap. 5 (2): 326 (1915). Type: South Africa, Cape, Aberdeen distr., without precise locality, Brauns (K, holo.). Euphorbia rudis N.E.Br., Fl. Cap. 5 (2): 322 (1915). Lectotype (Bruyns 2012): Namibia, sandy plains NE of

Narudas Süd, 28 Dec. 1912, Pearson 8141 (BOL; K, SAM, iso.). Euphorbia marientalii Dinter, Neue Pflanzen DeutschSWA’s: 31 (1914). Type: Namibia, Mariental, Dinter 3164 (SAM, holo.). Euphorbia rangeana Dinter, Neue Pflanzen Deutsch-SWA’s: 31 (1914). Type: none cited. Bisexual spineless densely to loosely branched nearly hemispherical glabrous succulent 50–150 (−200) × 50–300 mm, consisting of ±  clavate tuberculate stem 50–150 × 20–40  mm with very similar branches arising mainly in lower half, stem gradually tapering below into slender tap root bearing fibrous roots. Branches many, ascending, cylindrical to clavate, mainly simple (though often with some branchlets near apices), 30–150 × 10–40 mm, densely covered with tubercles, smooth, dark green above becoming grey-green below; tubercles low to prominent, conical, spreading to slightly reflexed near tip, 1–5  mm long with ± hexagonal base, often spirally arranged into 10–15 rows, with shallow grooves separating tubercles and a circular white apical scar from leaf; spines absent but few to several grey to white woody persistent fertile or sterile occasionally forked peduncles present on some to most branches, 6–30 × 1–3 mm, spreading and not sharp-pointed once dried out, with prominent outstanding scars of bracts; leaves on tips of new tubercles at apices of branches and stem, 4–30 × 1–3

2.1 Sect. Anthacanthae

mm, ascending-spreading, often persistent for some months then deciduous, narrowly linear-oblanceolate, obtuse and apiculate, sessile and tapering gradually into tubercle, longitudinally folded, grey-green sometimes with faint paler midrib visible, margins usually wavy and pinkish. Synflorescences 1–many per branch towards apex, solitary (rarely in cymes of 2–3) in axil of tubercle on green to brown-green sometimes persistent peduncle 6–30 × 2–3 mm with 1 terminal cyathium (rarely with 2 lateral cyathia and occasionally with several short branches), glabrous, with 4–8 oblanceolate to spathulate glabrous caducous bracts (with marginal cilia) 1.5–2.5  mm long along peduncle on short ascending canaliculate tubercles; cyathia urceolate to cupular, glabrous, 5–8 mm broad (2.5 mm long below insertion of glands), with 5 reddish green lobes with deeply incised ciliate margins, green often finely streaked with red or brown; glands (4) 5, elliptic to cuneate, 0.7–2 mm broad, spreading and widely separated, green to brown, slightly concave above, with raised inner margin, outer margins deeply divided into 2–5 spreading and slightly deflexed sometimes reddish or yellowish finger-like teeth 0.5–1 mm

Fig. 2.351.  Distribution of Euphorbia braunsii (© PVB).

259

long; stamens with pubescent pedicels, bracteoles filiform and pubescent; ovary ellipsoidal and slightly 3-angled, glabrous, sessile to raised on short pedicel up to 1 mm long; styles 3–4.5 mm long, branched and spreading to recurved in upper third. Capsule 5–7 mm diam., obtusely 3-angled, glabrous, sessile. Distribution & Habitat Euphorbia braunsii is very widely distributed in the arid parts of southern Africa from north of Mariental to Klein Karas and south-east of Warmbad in Namibia, from where it continues southwards via Goodhouse and Kakamas and then more sporadically across the Great Karoo to Beaufort West, Prince Albert and Willowmore. Only near Jakkalswater, north-east of Steinkopf does it grow near areas receiving winter-rainfall, otherwise always occurring in areas receiving rainfall in summer. Mostly, Euphorbia braunsii grows in gravelly to loamy (rarely coarsely sandy), flat areas among scattered tufts of grass and occasional shrubs (often among colonies of Rhigozum trichotomum), rarely sheltering in them and usu-

260

ally growing fully in the open, though starting off sheltered by a shrub or a stone. It often occurs in calcareous patches with gravel of calcrete, gneiss or jasper on the surface.

2  Euphorbia subg. Athymalus

Specimens are generally scattered and dense colonies have never been observed, though it may be fairly common in some areas.

Fig. 2.352.  Euphorbia braunsii, plant ± 15 cm diam., in vigorous growth and with plenty of the narrow leaves, Prutkraal near Rietbron, South Africa, 9 Apr. 1988 (© PVB).

Diagnostic Features & Relationships Euphorbia braunsii forms characteristically low, mound-like plants in which the branches are laxly grouped together into a roughly hemispherical shape close to the ground. In such clumps the stem is usually as long as or very slightly longer than the branches (often slightly thicker than they are too) but is otherwise indistinguishable from them. Occasional very densely branched specimens occur, but this is excep-

Fig. 2.353.  Euphorbia braunsii, ± 20 cm diam., without leaves and with brownish green cyathia, PVB 10085, west of Grünau, Namibia, 18 Jul. 2005 (© PVB).

tional. The branches and stem are covered with low to quite prominent tubercles and are usually slightly clavate in shape. Their surface is grey-green, often becoming whitish or pale brown with age from deposits of wax. Many of the branches (and the stem) are ornamented in their upper half by a few, woody, persistent peduncles. These peduncles are initially a chalky white on drying out and later become grey before wearing off towards the base of the branch. Leaves are only

2.1 Sect. Anthacanthae

present around the tips of the branches when new tubercles develop after rains. They vary greatly in length, becoming quite prominent in some areas (or after particularly good rains) and are generally slightly broader towards their tips, narrowing gradually towards the base into the tubercle. Those at the tip of the branch often spread out horizontally to

261

form a rosette and their pinkish margins may lend a faint, reddish hue to the otherwise grey colour of the branch. Their margins are slightly thickened and are nearly always finely undulating. Flowering takes place between August and October (or at other times depending on rain), with the cyathia produced

Fig. 2.354.  Euphorbia braunsii, densely branched plant ± 40 cm diam., in fruit, PVB 11245, east of Springbok, South Africa, 19 Oct. 2008 (© PVB).

around the tips of the branches, usually in small numbers. The peduncles are variable in length. In some plants they are almost as long as the cyathia while in others they are many times the length of the cyathia and up to 30 mm long. In E. braunsii the cyathia are inconspicuous, since their Fig. 2.355.  Euphorbia braunsii, leaves with slightly more undulating margins, PVB 11363, SE of Warmbad, Namibia, 8 Apr. 2009 (© PVB).

green colour and dark green glands match the colour of the branches. Their glands bear quite prominent teeth around their outer edges, but these are few and are not brightly coloured. The colour of the glands varies considerably, with many of those on the Great Karoo a brown colour while

262

2  Euphorbia subg. Athymalus

those on the eastern edge of Namaqualand are green. The male florets may be distinctly reflexed around the styles, which is particularly obvious in those with white pedicels.

Euphorbia braunsii and E. multiceps subsp. tanquana are very similar and the two are separable only with difficulty. In E. multiceps subsp. tanquana the plant is altogether a brighter, shinier green colour often slightly tinged with red

Fig. 2.356.  Euphorbia braunsii, with brownish green cyathial glands, Prince Albert, South Africa, 8 Aug. 2011 (© PVB).

Fig. 2.357.  Euphorbia braunsii, with green cyathial glands, Prince Albert, South Africa, 8 Aug. 2011 (© PVB).

(rather than the duller, darker grey-green of E. braunsii) and the older stems do not become pale grey from the deposition of wax, as they do in E. braunsii. Furthermore, the leaves in E. multiceps are less obviously glaucous and have straight rather than wavy margins. Florally they are hard to separate. The cyathia are somewhat deeper and more cupular with narrower and widely separated glands in E. braunsii, while in E.

multiceps they are shallower and the glands are broader and often closely pressed to one another laterally. The teeth on the outer margins of the glands are more numerous, longer and more obviously differently coloured from the rest of the gland in E. multiceps than in E. braunsii.

Fig. 2.358.  Euphorbia braunsii, small plant ± 12 cm diam., in fruit, PVB 11114, east of Jakkalswater, South Africa, 18 Oct. 2008 (© PVB).

2.1 Sect. Anthacanthae

History The type of Euphorbia braunsii was collected near Aberdeen by J.C.E.H.J. Brauns in 1911 and sent to Kew by Schönland from Grahamstown in September of that year. Slightly earlier than this, in January and February of 1909, several collections from southern Namibia were made by H.H.W.  Pearson and these were used by N.E.  Brown to describe another ‘species’, E. rudis.

263

Although E. rudis was maintained as distinct from E. braunsii in Bruyns et  al. (2006), White et  al. (1941: 474) experienced problems separating them and mentioned that ‘there is really no sharp line of distinction between the two plants, but rather a gradation. The typical forms of the two are fairly clearly distinguishable, while many of the intermediate forms are very confusing indeed and difficult to classify satisfactorily.’ The distinctions that they mentioned

Fig. 2.359.  Euphorbia braunsii. A, B, cyathium from side (scale 2 mm, as for B, C). C, cyathium from above. D, side view of dissected cyathium, with bracteole (scale 1 mm, as for E–G). E, anthers. F, G, female floret. Drawn from: A, C, D, PVB 11363, SE of Warmbad, Namibia. B, E, G, PVB 11014, just east of Beaufort West, South Africa. F, PVB 11114, east of Jakkalswater, South Africa (© PVB).

between the two included: the smaller ‘average size of the stem’, the ‘more slender’ branches with the tubercles ‘somewhat more recurved at the apex’ and ‘rather smaller’ cyathia and ‘more completely united styles’ in E. rudis. These features are all subject to considerable variation and there are many collections that could be placed under either of these names. With the many new collections that are now available it becomes clear that one cannot separate these into two species and all these collections are placed now under E. braunsii.

Euphorbia caput-medusae L., Sp. Pl. 1: 452 (1753). Lectotype (Wijnands 1983: 98): J. Burm., Rar. Afric. Pl.: 17, t. 8 (1738). Euphorbia fructus-pini Mill., Gard. Dict., ed. 8: Euphorbia no. 10 (1768). Medusea fructus-pini (Mill.) Haw., Syn. Pl. Succ.: 134 (1812). Neotype (Bruyns 2012): J. Burm., Rar. Afric. Pl.: t. 8 (1738). Euphorbia caput-medusae var. geminata Aiton, Hort. Kew., ed. 1, 2: 136 (1789). Lectotype (Bruyns 2012): J. Burm., Rar. Afric. Pl.: t. 9, fig. 1 (1738).

264

Euphorbia caput-medusae var. major Aiton, Hort. Kew., ed. 1, 2: 135 (1789). Lectotype (Bruyns 2012): Commelijn, Praeludia Bot.: t. 7 (1703). Euphorbia caput-medusae var. minor Aiton, Hort. Kew., ed. 1, 2: 135 (1789). Lectotype (Bruyns 2012): Breyne, Prodr. rar. pl. sec.: t. 19 (1739). Euphorbia tuberculata Jacq., Pl. Hort. Schönbr. 2: 43, t. 208 (1797). Dactylanthes tuberculata (Jacq.) Haw., Syn. Pl. Succ.: 133 (1812). Treisia tuberculata (Jacq.) Haw., Suppl. Pl. Succ.: 65 (1819). Medusea tuberculata (Jacq.) Klotzsch & Garcke, Abh. Königl. Akad. Wiss. Berlin 1859: 61 (1860). Lectotype (Bruyns 2012): Jacq., Pl. Hort. Schönbr. 2: t. 208 (1797). Euphorbia medusae Thunb., Prodr. Fl. Cap. 2: 86 (1800). Lectotype (Bruyns 2012): South Africa, Cape, Thunberg (UPS-THUNB 11494). Medusea major Haw., Syn. Pl. Succ.: 134 (1812). Lectotype (Bruyns 2012): Commelin, Praeludia Bot.: t. 7 (1703). Medusea tessellata Haw., Syn. Pl. Succ.: 135 (1812). Euphorbia tessellata (Haw.) Sweet, Hort. Suburb. Lond.: 107 (1818). Type: none cited. Euphorbia commelinii DC., Cat. Pl. Horti. Monsp.: 110 (1813). Lectotype (Wijnands 1983: 98): Commelijn, Praeludia Bot.: t. 7 (1703). Euphorbia fructus-pini var. geminata Sweet, Hort. Brit., ed. 1, 2: 356 (1826). Lectotype (Bruyns 2012): J.Burm., Rar. Afric. Pl.: 18, t. 9, f. 1 (1738). Euphorbia bolusii N.E.Br., Fl. Cap. 5 (2): 333 (1915). Type: South Africa, ‘Transvaal, near Middelburg’, Sept. 1886, H.Bolus 9767 (BOL, holo.; K, iso.).

Fig. 2.360.  Distribution of Euphorbia caput-medusae (© PVB).

2  Euphorbia subg. Athymalus

Euphorbia ramiglans N.E.Br., Fl. Cap. 5 (2): 306 (1915). Type: South Africa, Namaqualand, 1883, H.  Bolus sub BOL 9448 (BOL, holo.; K, iso.). Euphorbia marlothiana N.E.Br., Fl. Cap. 5 (2): 331 (1915). Type: South Africa, Cape, near Neu Eisleben, fl. Oct.Nov. 1914, Marloth 5733 (PRE, holo.; BOL, NBG, K, iso.). Euphorbia muirii N.E.Br., Fl. Cap. 5 (2): 331 (1915). Lectotype (Bruyns 2012): South Africa, Cape, Platbos, Still Bay, Muir 174 (BOL; PRE, SAM, iso.). Euphorbia tuberculatoides N.E.Br., Fl. Cap. 5 (2): 332 (1915). Lectotype (Bruyns 2012): South Africa, Cape, near Theefontein, Malmesbury div., Bachmann 1042 (K). Euphorbia macowanii N.E.Br., Fl. Cap. 5 (2): 334 (1915). Euphorbia tuberculata var. macowanii (N.E.Br.) A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 1: 372 (1941). Lectotype (Bruyns 2012): South Africa, Clanwilliam (wrongly labelled as Cannon Hill, Uitenhage), MacOwan 3286 (K). Euphorbia confluens Nel, Kakteenk.: 193 (1933). Type: South Africa, Cape, open flats, Kliphoogte, Sept. 1929, Herre sub SUG 5549 (missing). Lectotype (Bruyns 2012): figure in protologue. Bisexual spineless and glabrous succulent 0.1–0.75 m tall with clavate tuberculate stem 50–200 mm thick usually sunken into ground and rising 0.01–0.1 (0.2) m above surface, tapering beneath into tap-root bearing fibrous roots, with many branches towards apex. Branches ascending to spreading, simple to rebranching and sometimes up

2.1 Sect. Anthacanthae

to 0.5  m long, cylindrical to slightly clavate, (8–) 10–30  mm thick, covered with tubercles, glabrous and smooth, grey-green; tubercles prominent and conical, 3–8  mm long with elongated hexagonal base, vertically arranged into 8–14 obscure spiralling rows, with fine grooves between them, occasionally with solitary simple persistent woody peduncles in their axils; leaves on tips of new tubercles towards apices of branches and stem, (3) 4–12 (20) × 1–2 mm, ascending and slightly spreading near often reddish apex, fleeting, linear, longitudinally folded upwards near apex, tapering to acute apex and narrow base, sessile. Synflorescences 10–40 per branch towards apex, each of one bisexual cyathium in axil of tubercle terminating stout fleshy green to reddish peduncle 4–30 (–40) mm long, with 3–6 oblong-ovate caducous often reddish bracts 2–4  mm long (longest subtending cyathium) glabrous except for ciliate margins; cyathia shallowly conical, glabrous, 10–15 mm broad (1.5–4 mm long below insertion of glands), with 5 glabrous reddish lobes with deeply incised margins, pale green; glands 5, ±  cuneate, not contiguous, glabrous, 3–4  mm long, dark green to olive green in lower united portion, spreading, slightly convex and rugose above, inner margins sometimes reddish green and slightly raised against lobes, outer margins cream then divided into 3–7 ascending to slightly recurved finger-like irregularly warty cream or pinkish cream processes often further finely subdivided towards their apices; stamens with pubescent or glabrous pedicels,

Fig. 2.361.  Euphorbia caput-medusae, with Ceropegia mixta (Orbea variegata) and other succulents on steep outcrops of shale, Lower Cable Station, Table Mountain, Cape Town, South Africa, Jan. 1985 (© PVB).

265

bracteoles filiform and pubescent; ovary ellipsoidal, glabrous, sessile to slightly raised on pedicel < 0.2 mm long; styles 3–4.5  mm long, branched in upper third. Capsule 6–7  mm diam., obtusely 3-angled, grey-green, glabrous, sessile. Distribution & Habitat Euphorbia caput-medusae is found in the winter-rainfall region of western South Africa and the south-western corner of Namibia. It occurs in the coastal areas only. Here it is widespread in Namaqualand, from just north of Alexander Bay (in Namibia) to the mouth of the Olifants River, southwards to the Cape Peninsula and on the surrounding sands of the Cape Flats. Eastwards from Cape Town it is rather sporadically recorded to Still Bay and Mossel Bay and there is a single record from the Worcester-Robertson Karoo. Around Cape Town plants sometimes grow in stony situations on shale among many other succulents, as on the lower slopes of Table Mountain, on Signal Hill and on some other spots on the southern parts of the Cape Peninsula and north of it on the Blouberg. Mostly, however, they are found in sand, where they occur together with typical ‘arid fynbos’ of the sandveld. Around Still Bay they are found on sandy soils overlaying limestones amongst fynbos and the species also occurs on limestones south-east of Cape Town (towards Strandfontein).

Fig. 2.362.  Euphorbia caput-medusae, on granitic outcrops and again with Ceropegia mixta, Llandudno, Cape Town, South Africa, Aug. 2002 (© PVB).

Fig. 2.363.  Euphorbia caput-medusae, Signal Hill, Cape Town, South Africa, 13 Jul. 2015 (© PVB).

Fig. 2.364.  Euphorbia caput-medusae (‘muirii’), Still Bay, South Africa, 3 Jul. 2013 (© PVB).

2.1 Sect. Anthacanthae

Diagnostic Features & Relationships Euphorbia caput-medusae is the largest of the many ‘medusoid Euphorbias’. In the form formerly known as E. tuberculata, from the Olifants River Valley southwards to around

267

Hopefield, the plants may be up to 0.75 m tall and may even exceed 1 m in diameter. While such large forms are prevalent in the deep sands along the Olifants River around Clanwilliam (for example, but also westwards to near Lambert’s Bay),

Fig. 2.365.  Euphorbia caput-medusae (‘tuberculata’), large plant ± 1 m diam. in deep sand with restios and Apatesia helianthoides, West of Graafwater, South Africa, 7 Sep. 2012 (© PVB).

Fig. 2.366.  Euphorbia caput-medusae, on calcrete outcrops partially submerged in sand, Yserfontein, South Africa, Oct. 1994 (© PVB).

they are not typical of the species. On the Cape Peninsula, they are considerably smaller but with particularly stout stems and stout branches. In general, apart from the exceptionally large size of the ‘tuberculata’ form, the plants decrease steadily in size as one moves into the drier areas. Thus, in the dry sands of coastal Namaqualand, they may be

as small as 10 cm in diameter, with a small stem almost completely sunken into the ground and short branches, often only 5 cm long, forming a lax rosette around its apex. Here occasional specimens may be found where the stem projects for up to 20 cm from the ground, but this is unusual.

268

2  Euphorbia subg. Athymalus

Fig. 2.367.  Euphorbia caput-medusae (‘tuberculata’), west of Graafwater, South Africa, 7 Sep. 2012 (© PVB).

The former E. marlothiana is a very slender-stemmed, slender-branched, spreading form with frequent secondary branching that is found among restios and other sandveld vegetation on the Cape Flats and northwards for some 50 km

from Cape Town to near Yserfontein. It has been observed that these slender-branched forms prevail in the sandy flats here, but the branches become shorter when plants occur on rocky ground (as on the neighbouring Blouberg) so that this

Fig. 2.368.  Euphorbia caput-medusae (‘marlothiana’), among restios in sand, Koeberg, South Africa, Oct. 1994 (© PVB).

Fig. 2.369.  Euphorbia caput-medusae (‘marlothiana’), Koeberg, South Africa, Oct. 1994 (© PVB).

is regarded as a local ecotype of E. caput-medusae. Such slender-stemmed and slender-branched forms are also found among plants referred to E. muirii from the Riversdale area. Flowering in E. caput-medusae occurs from April until September, rarely continuing into October. In years of adequate rainfall, the upper part of each branch becomes covered with the creamy cyathia, which develop in the axils of the tubercles just behind those tubercles that bear the leaves. They emit a sweetish, honey-like odour. Each cyathium is

terminal on a stout, obscurely tuberculate peduncle which is of very variable length (though generally longer in the moister south than in the drier north-west). The peduncle has a few small bracts along it and larger ones at its apex that wrap around the cyathium while it is still small, though parts of the cyathium soon begin to project from between these bracts (unlike in E. pentops, where the bracts are larger and cover the cyathium more effectively until much later in its development). Mostly the peduncles drop off after flow-

2.1 Sect. Anthacanthae

269

Fig. 2.370.  Euphorbia caput-medusae, in deep red sand (screwdriver 30 cm long), PVB 6164, SE of Soutpan, west of Garies, South Africa, Oct. 1994 (© PVB).

Fig. 2.371.  Euphorbia caput-medusae, PVB 6164, SE of Soutpan, west of Garies, South Africa, Oct. 1994 (© PVB).

Fig. 2.372.  Euphorbia caput-medusae, very finely divided appendages on cyathial glands, west of Driekop, near Wallekraal, South Africa, 14 Jul. 1995 (© PVB).

Fig. 2.373.  Euphorbia caput-medusae, ±  15  cm diam., in pale sand close to coast, PVB 11407, north of Groen River Mouth, west of Garies, South Africa, 10 Jul. 2009 (© PVB).

Fig. 2.374.  Euphorbia caput-medusae, PVB 11407, north of Groen River Mouth, west of Garies, South Africa, 10 Jul. 2009 (© PVB).

270

2  Euphorbia subg. Athymalus

ering but occasionally (as in the patch of red sand north of Vanrhynsdorp, Namaqualand) plants with persistent woody peduncles are found. The most prominent feature of the comparatively large cyathia is the series of five glands around its perimeter. Each gland has a green lower part (sometimes very short and broad) with a slight cream outer margin. From this margin a series of up to seven, finger-like growths project. These can be further subdivided finely towards their tips giving (in such cases) an almost crocheted lace-like appearance around the cyathium. As they age, these projections may change to pale pink or become flecked faintly with red.

Apart from its size and the length of the branches and the prominence of the tubercles on them, E. caput-medusae is distinguished from other allied species by the large cyathia with their especially conspicuous and much divided creamy processes on the glands. The closest relative is probably E. brakdamensis and these two are distinguished by the more erect, slender branches of E. brakdamensis with longer leafrudiments, its deeper cyathia (and longer styles) with relatively small glands with at most three simple processes on their outer margins.

Fig. 2.375.  Euphorbia caput-medusae, almost completely submerged in pale windblown sand and only the tips of the branches visible, Kortdoringberg, Alexander Bay, South Africa, 13 Jul. 2014 (© PVB).

Fig. 2.376.  Euphorbia caput-medusae, much larger and more exposed plant, Kortdoringberg, Alexander Bay, South Africa, 13 Jul. 2014 (© PVB).

History It seems likely that E. caput-medusae was in cultivation in the Netherlands and possibly elsewhere in Europe well before 1700. One of the first known illustrations of E. caputmedusae is one painted by Maria Moninckx during 1702 from a plant in cultivation in the Hortus Medicus in Amsterdam. This figure is part of the Moninckx Atlas (Volume 5, t. 46, according to Wijnands 1983). It shows a branch planted upright in the soil and covered with cyathia around its tip and this is clearly identifiable with E. caputmedusae as we know it. Two figures of E. caput-medusae (Folios 120 and 190) are present in the volume of the Codex witsenii known as Icones plantarum et animalium (Macnae and Davidson 1969) and these may have been made even earlier. Euphorbia caput-medusae was known to Linnaeus and was the eighth of the 56 species that he listed in the Species Plantarum of 1753. The ‘varieties’ α, β, γ, δ, ε and ζ. that

Linnaeus included under E. caput-medusae, are today regarded as several distinct species. For var. α, Linnaeus cited ‘Comm. hort. 1: 33, t. 17’, (i.e. Commelijn 1697: 33, t. 17). This figure looks rather like an etiolated specimen of E. procumbens that has suffered for a season or two under conditions of inadequate light in cultivation. He also cited ‘Comm. Prael. 58, t. 8’ (i.e. Commelijn 1703: 58, t. 8). This illustration is either of E. clandestina or E. clava and it is unclear which of the two is shown, since the plant is sterile. Lastly, he cited ‘p. 28, 57, t. 7’ (i.e. Commelijn 1703: 57, t. 7). This figure shows two sterile branches of E. caput-medusae. Consequently, Linneaus’ ‘var. α’ is a mixture of three species that are not considered to be conspecific today. For var. β, Linnaeus cited ‘Isnard t. 11’, which represents E. patula subsp. patula. For var. γ, he cited ‘Burm. afr. t. 7, f. 2’, which is E. patula subsp. anacantha. For var. δ, Linnaeus cited ‘Burm. afr. t. 10. f. 1’, which is E. procumbens. For var. ε, he cited ‘Burm. afr. t. 8’, which is

2.1 Sect. Anthacanthae

271

Fig. 2.377.  Euphorbia caput-medusae. A, B, tip of branch with leaf-rudiments (scale 2 mm, as for B). C, tip of branch and cyathium from side (scale 3 mm). D, E, cyathium from above (scale 2 mm, as for E). F, side view of part of dissected cyathium (scale 1 mm, as for G, H). G, anthers and bracteole. H, female floret. Drawn from: A, Seekoeivlei, Cape Town, South Africa. B, D, G, H, Signal Hill, Cape Town, South Africa. C, E, F, PVB 3819a, Nariep, west of Garies, Namaqualand, South Africa (© PVB).

E. caput-medusae with the stout branches usually seen on plants from the lower slopes of Table Mountain. For var. ζ, he cited ‘Burm. afr. t. 7’, which is also E. caput-medusae. The plant described by N.E. Brown as E. bolusii was initially identified (on the specimen in the Bolus Herbarium) as E. tuberculata and, as White et al. (1941: 373) mentioned, there was little to separate it from E. tuberculata. They mentioned also that the locality given for the type collection, namely ‘Transvaal, near Middelburg’ appears to have been an error. Euphorbia clavarioides Boiss., Cent. Euphorb.: 25 (1860). Lectotype (Bruyns 2012): South Africa, Cape, Sneeuberge at Poortjie, Drège 8200 (P; BM, K, S, TCD, W, iso.). Euphorbia basutica Marloth, Trans. Roy. Soc. South Africa 1: 408 (1910). Type: Lesotho, Leribe, Dieterlin 415 (cult. Phillips, fl. Cape Town in Mar. 1909) sub Marloth 4671 (K, holo.; MPU, NH, PRE, SAM, iso.). Euphorbia clavarioides var. truncata (N.E.Br.) A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 1: 309 (1941). Euphorbia truncata N.E.Br., Fl. Cap. 5 (2): 309 (1915). Lectotype (Bruyns 2012): South Africa, Standerton, Burtt-Davy 1953 (K).

Bisexual spineless and glabrous densely mat-forming or cushion-like to mound-forming succulent 20–250 × 60–500 mm with clavate tuberculate stem 20–100 mm thick usually entirely sunken into ground (occasionally rising 20–50  mm above surface), tapering gradually below and bearing fibrous roots, with many branches arising towards apex and generally covering and hiding stem completely. Branches mostly arising below ground on stem and ascending towards surface of soil or spreading, simple or rebranching, 10–100  mm long, initially almost spherical and later becoming cylindrical to very slightly clavate, 8–20  mm thick, covered with low tubercles, glabrous and smooth, yellow-green to bright green or purplish green; tubercles low and rounded-conical to acutely conical, 1–4 mm long with slightly elongated hexagonal base, vertically arranged into 8–10 obscure slightly spiralling rows, with fine grooves between them, without persistent peduncles in their axils; leaf-rudiments on tips of new tubercles towards apices of branches and stem, 1–2 × ±  1 mm, ascending to slightly reflexed, fleeting, ovate, channelled above, very shortly acute, grey-green sometimes with reddish margins, sessile. Synflorescences 1–10 per branch near apex, each of one bisexual cyathium in axil of tubercle on stout fleshy pedun-

272

cle 2–6 mm long, with 2–10 oblong-ovate caducous bracts 2–4  mm long glabrous except for prominently ciliate margins (margins often glabrous at apex); cyathia broadly conical, glabrous outside, 5–8 mm broad (1.5–3 mm long below insertion of glands), with 5 glabrous inwardly pressed pale yellow lobes (sometimes suffused with red or white with faint red stripes) with ciliate margins, pale green-yellow outside; glands 5 (6), broadly cuneate, spreading to slightly descending, 1.5–2 × 2–4 mm, well separated to contiguous, green (sometimes faintly suffused with red and then brownish) to bright yellow and slightly glandular-pitted towards base, outer margins almost entire to divided into 2–10 or more spreading to slightly recurved stubby rounded teeth; stamens with pubescent pedicels, bracteoles filiform and

2  Euphorbia subg. Athymalus

ciliate for most of length; ovary ellipsoidal, glabrous to sparsely pubescent, sessile to shortly pedicellate; styles 1.5–6 mm long, branched to between middle and near base. Capsule 6–8  mm diam., obtusely 3-angled, glabrous, sessile. Distribution & Habitat Euphorbia clavarioides occurs widely over South Africa, where it is found at elevations above 1000 m and is known to occur in areas as high as 2700  m. The south-westernmost records are on the western part of the Great Karoo from above the escarpment between Fraserburg and Sutherland and near Beaufort West. From here it occurs along the edges of the escarpment to the Eastern Cape and northwards to the

Fig. 2.378.  Distribution of Euphorbia clavarioides (© PVB).

Drakensberg, both in Lesotho (where it is common, Hargreaves 1992b) and in its foothills in Kwazulu-Natal. Euphorbia clavarioides is also recorded further north in various parts of the former Transvaal. There are also a few sporadic records further west in the Northern Cape near Kuruman and Vryburg and in the Free State.

Mostly Euphorbia clavarioides grows in grassland in relatively small, locally dry, quite barren-looking patchess caused by shallow soils over slabs of rock or excessive stoniness in the oil and where the grass remains short. It is often associated with outcrops of basalt or dolerite, which it especially favours in Lesotho (Hargreaves 1992b), but it is also

2.1 Sect. Anthacanthae

273

Fig. 2.379.  Euphorbia clavarioides, two extremely dense mounds ± 0.5 m diam., PVB 11034, NE of Colesberg, South Africa, 27 Apr. 2008 (© PVB).

known on substrates derived from various kinds of shales and sandstones, generally associated with a scanty cover of short shrublets. Individuals are often, but not always, tightly wedged between rocks. Plants growing at high altitudes frequently experience heavy frosts or generous falls of snow in

the winter. Many of them regularly experience burning by fire but, apart from killing off a few of the shortest branchlets and scalding some of the older and longer branches, this usually has little effect on them.

Fig. 2.380.  Euphorbia clavarioides, very dense mound ± 0.3 m diam. with Ruschia spinosa in flower around it, PVB 10588, SW of Fraserburg, South Africa, 14 Oct. 2006 (© PVB).

Diagnostic Features & Relationships A plant of Euphorbia clavarioides begins as a single, fairly slender stem between 10 and 20 mm thick that usually projects a little from the ground, which is covered with low tubercles and tapers below into a slender taproot. This stem gradually swells, developing beneath the ground into a long

tap-root, while branches develop from the axils of some of the tubercles a bit below its apex, to form a small rosette more or less flush with the surface of the soil. By extensive rebranching this rosette gradually expands to form a roughly circular mat or a raised mound which, in due course, hides the stem completely. Eventually, older specimens may

274

2  Euphorbia subg. Athymalus

Fig. 2.381.  Euphorbia clavarioides, in flower, PVB 11340, Winburg, South Africa, 14 Jan. 2009 (© PVB).

develop into dense, flat to hemispherical cushions which can reach 0.5 m in diameter and may rise to a height of 150 mm. These larger and older specimens are made up of extraordinary numbers of very tightly packed branchlets, each of which rarely exceeds 15 mm in diameter. In many areas the branches are all almost spherical and have very low and inconspicuous tubercles, but this is not always the case and the branches may become quite elongated and much more roughly tuberculate. Young tubercles bear very short, thick,

Fig. 2.382.  Euphorbia clavarioides, cyathia with bright yellow glands, PVB 11340, Winburg, South Africa, 14 Jan. 2009 (© PVB).

rounded leaf-rudiments that are rapidly shed to leave a small, round scar at the apex of the tubercle. Large numbers of the sweetly scented cyathia are usually produced between late spring (September– October) and the height of summer (December to January), though plants may begin to flower as early as August. When a plant is fully in flower, many of the branches are adorned near their tips with cyathia, so that the upper surface of the plant becomes covered with cyathia. In some areas (especially in the Fraserburg

2.1 Sect. Anthacanthae

district) cyathia mature first on the warmer northern side of the plant and gradually this progresses over the plant to the southern side. Capsules are already then well-developed on the northern side, as the flowering progresses over the plant. Over most of its distribution the cyathia are bright yellow or yellow-green and, when present in large numbers, they make the plant much more conspicuous than usual. In the Fraserburg district the cyathia are green suffused with red and are much less conspicuous against the green colour of the branches. Each cyathium arises on a short, slightly tuberculate peduncle and has particularly conspicuous, spreading,

Fig. 2.383.  Euphorbia clavarioides, cyathia with greenish glands, PVB 11340, Winburg, South Africa, 14 Jan. 2009 (© PVB).

Fig. 2.384.  Euphorbia clavarioides, mound ± 0.5 m diam. among doleritic rocks, cyathia with dull dark greenish glands, PVB 10594, SW of Fraserburg, South Africa, 14 Oct. 2006 (© PVB).

275

bright yellow or greenish yellow (more rarely green) glands which more or less dominate it both in colour and size. On their outer margin the glands are ornamented with short and stubby, spreading, finger-like processes that are usually the same colour as the glands themselves. Considerable variation in the numbers and the shapes of these processes has been observed within some populations. There is also considerable variation in how closely the glands are held to the sides of the cyathium. Euphorbia braunsii and E. hypogaea sometimes occur near E. clavarioides in some of its westernmost localities.

276

These species are readily distinguished from it by their differently shaped branches, with longer tubercles and considerably longer leaf-rudiments. The comparatively broad cyathial glands in E. clavarioides are also clearly different both in shape and colour from those in E. braunsii and E. hypogaea. They are much more similar to those of E. flanaganii, to which E. clavarioides may well be closely allied and with which it sometimes occurs as well. Euphorbia

Fig. 2.385.  Euphorbia clavarioides, plant in fruit on the warmer, north-facing side of the mound and still in flower on the cooler south-facing side, PVB 10588, SW of Fraserburg, South Africa, 29 Sep. 2012 (© PVB).

Fig. 2.386.  Euphorbia clavarioides, dull greenish cyathia, PVB 10588, SW of Fraserburg, South Africa, 29 Sep. 2012 (© PVB).

2  Euphorbia subg. Athymalus

clavarioides differs from E. flanaganii in the manner in which the branches repeatedly rebranch to form mats which eventually hide the stem and only young specimens of E. clavarioides may be mistaken occasionally for E. flanaganii (in E. flanaganii the branches rarely rebranch but are gradually deciduous and always form a rosette around the stem which remains much thicker than any of the branches.

2.1 Sect. Anthacanthae

History Euphorbia clavarioides was discovered by J.F. Drège in the Sneeuberge of the Eastern Cape, north of Graaff-Reinet, probably in February 1833, in the company of his brother Carl. Boissier’s unusual name ‘like Clavaria’, refers to its

277

possible resemblance to a certain genus of lichen. White et al. (1941) explained in detail why the names E. basutica and E. truncata need not be regarded as species distinct from E. clavarioides and they took a broad view of the species, which seems to be fully justified.

Fig. 2.387.  Euphorbia clavarioides. A, cyathium from side (scale 2 mm, as for B–E). B–E, cyathium from above showing the considerable variation in the size and shapes of the glands. F, anthers and bracteoles (scale 1 mm, as for G–I). G–I, female floret. Drawn from: A, B, D, F, G, I, 3 km south of Trompsburg, South Africa. C, PVB 12214, Lanseria airport, west of Johannesburg, South Africa. E, H, PVB 10588, SW of Fraserburg, South Africa (© PVB).

Euphorbia colliculina A.C.White, R.A.Dyer & B.Sloane, Succ. Euphorb. 2: 963 (1941). Lectotype (Bruyns 2012): South Africa, Cape, 2.5 miles north of Oudtshoorn, Aug. 1939, Dyer 4053 (PRE; BOL, K, PRE, iso.). Bisexual spineless glabrous succulent 50–150  mm tall with stout clavate tuberculate stem 50–150 mm thick sometimes projecting from the ground for up to 50 mm, tapering gradually beneath ground into tap-root bearing fibrous roots, with many branches forming rosette 0.1–0.25 (0.4) m diam. around slightly depressed apex. Branches ascending to spreading, simple, cylindrical, 20–150 × 4–22 mm, densely covered with small tubercles, smooth, pale green; tubercles low and not prominent, conical, 1 mm long with rhomboidal base, densely arranged into 10–15 obscure spiralling rows, with fine grooves around them, very rarely with blunt persistent peduncles 5–10  mm long in their axils; leaf-rudiments on tips of new tubercles towards apices of branches, 1–1.5 × 1 mm, recurved, fleeting, subulate, margins slightly folded upwards and eciliate, shortly acute, sessile. Synflorescences 1–30 per branch around apex, each

of one bisexual cyathium in axil of tubercle on slightly tuberculate peduncle 1–2 mm long, with 3–5 slightly spathulate glabrous caducous bracts 2–3  mm long and pilose along margins towards apex; cyathia conical, glabrous, 4.5–7 mm broad (± 2 mm long below insertion of glands), with 5 glabrous lobes with fimbriate margins, green; glands 5, ±  elliptic, 2–3  mm broad, widely separated to closely contiguous, slightly to strongly reflexed towards edges, green to yellow, convex and smooth above, outer margins entire to finely notched; stamens with pubescent pedicels, bracteoles filiform and pubescent except towards apex; ovary slenderly globose, glabrous, raised on pedicel 0.2– 0.5 mm long; styles 2–3 mm long, branched near apex or to near middle. Capsule 6–7  mm diam., obtusely 3-angled, glabrous, sessile. Distribution & Habitat Euphorbia colliculina is best known around the town of Oudtshoorn on the Klein Karoo, where White et al. (1941) mentioned that it grew on the north side of the town. However, today the town has expanded greatly and E. colliculina is

278

2  Euphorbia subg. Athymalus

Fig. 2.388.  Distribution of Euphorbia colliculina (© PVB).

mainly found south and west of the town, where it is often still quite plentiful, despite the destruction of much of the low-lying habitat in the area by ostrich-farming. It has since proved to be considerably more widespread on the Klein Karoo to the east of Oudtshoorn, where it has been recorded from the hamlet of De Rust to Barandas and Toorwater. Euphorbia colliculina usually grows on low, gently sloping alluvial deposits at the bases of shale ridges or on gently

Fig. 2.389.  Euphorbia colliculina, PVB 10615, south of Oudtshoorn, South Africa, in dry conditions, 2 Dec. 2006 (© PVB).

sloping, low shale hills. Occasionally plants will be found in deeper, loamy soil, but most of them grow in hard, stony ground, where there is little soil, often with many white pebbles lying about on the surface. Although each plant starts off protected by a low shrub or rock, mature specimens are often in the open and are completely exposed to the elements. They usually occur with many other small succulents.

2.1 Sect. Anthacanthae

279

Diagnostic Features & Relationships Euphorbia colliculina is a typical ‘vingerpol’, with a thick stem that is mainly sunken into the ground (in this case with usually at most the upper 10 to 30 mm projecting from the soil) and with many slender branches radiating from near its apex. Here the stem is usually not more than 10 cm thick and is covered with a corky greyish bark except around its tuberculate apex, with the branches spreading from around its apex usually quite close to the ground. The branches form a rosette which is very variable in size, depending on the aridity of the habitat. Thus, plants on stony patches with very little soil may be only 10 to 15 cm in diameter when mature (though even here sheltered specimens can attain 30 cm in diameter), while those in deeper ground may even reach 40 cm across occasionally. Both the branches and the apex of the stem are a pale green and the branches are mostly less than 15 mm thick, densely covered with small, low tubercles that scarcely rise from the surface and give the surface a scaly appearance. Tiny, recurved leaf-rudiments develop on the tips of new tubercles, but are rapidly lost, leaving a small, almost circular, white scar at the tip of each tubercle. The wonderfully sweet-smelling, bright yellow-green cyathia are produced between August and September but, in cultivation, may appear throughout the summer until as late as April. As is usual in these species, they arise around the tips of the branches. In habitat in good years (which are often those with fairly low but evenly spaced out showers of rain) the tips of the branches are densely covered with cyathia and this makes the flowering plants visible for some distance. In E. colliculina, the cyathia are borne on very short, tuberculate peduncles which mostly do not persist Fig. 2.390.  Euphorbia colliculina, PVB 10615, south of Oudtshoorn, South Africa, plant excavated to show rootstock (© PVB).

Fig. 2.391.  Euphorbia colliculina, south-west of Oudtshoorn, South Africa, in full flower, 14 Aug. 2010 (© PVB).

280

2  Euphorbia subg. Athymalus

Fig. 2.392.  Euphorbia colliculina, PVB 11012, Buffelsklip, east of Oudtshoorn, South Africa, in full flower, 14 Aug. 2010 (© PVB).

after flowering is over. Each fairly obviously conical cyathium is relatively small and has quite prominent, spreading, convex, yellow or green glands which are slightly elliptic and have entire or finely notched outer margins. In habitat the glands are generally very tightly pressed laterally to one another, though in cultivation they may be quite widely spaced around the circumference of the cyathium.

Fig. 2.393.  Euphorbia colliculina, PVB 11513, Toorwater, South Africa, in full flower, 15 Aug. 2010 (© PVB).

Inside, the cyathium has relatively few male florets with finely pubescent pedicels and a few filiform and sparsely pubescent bracteoles. The female floret is particularly variable in length, in the extent to which the styles are divided and in the length of its pedicel, but it is always glabrous. The branches in this species resemble those of E. esculenta, under which it was erroneously included in Bruyns

2.1 Sect. Anthacanthae

281

(2000) and Bruyns et al. (2006). However, mature specimens of E. colliculina are altogether more delicate and neither the stem nor the branches reach the thickness that are quite typical for E. esculenta. Euphorbia esculenta also produces several swollen roots from the base of the tap-root (as does E. inermis) and this phenomenon is unknown in E. colliculina, where the thick taproot tapers off quite abruptly into slender, fibrous roots. Florally E. colliculina is easily separated from E. esculenta by the different shape of the cyathia (conical as opposed to nearly spherical), also in that both the cyathial lobes and the bracteoles within the cyathium lack the dense hairiness at their apices that make the cyathium of E. esculenta distinctively furry or woolly. The cyathial glands are also much

larger than those of E. esculenta and are considerably more prominent around the top of the cyathium. From E. fortuita, E. colliculina is separated by its often more slender branches, which are more erect and have a brighter green colour. The cyathia are of a similar diameter in both but the most striking feature of E. fortuita is the dense, woolly appearance of the top of the cyathium. A further difference is in the glands, which are usually red to purple in E. fortuita, with many short and fine, spreading, finger-like, pale projections on their outer margins. Both the male and female florets are suffused with red in E. fortuita and this is unknown in E. colliculina.

Fig. 2.394.  Euphorbia colliculina, south-west of Oudtshoorn, South Africa, in full flower, 14 Aug. 2010 (© PVB).

Fig. 2.395.  Euphorbia colliculina, south-west of Oudtshoorn, South Africa, covered with shiny capsules, with Malephora still flowering alongside it, 4 Oct. 2008 (© PVB).

History Euphorbia colliculina was first recorded by Miss L. Britten at Oudtshoorn and this collection was forwarded to Kew by S. Schӧnland (under his number 2002), where it arrived in

August 1913. Rudolf Marloth collected it again at Oudtshoorn in October 1921. For some reason he did not publish a name for it and it was only described by White et al. (1941), after R.A. Dyer recollected it in August 1939.

282

2  Euphorbia subg. Athymalus

Fig. 2.396.  Euphorbia colliculina. A, tip of branch with leaf-rudiments (scale 2 mm, as for B, C). B, C, cyathium from side. D, part of dissected cyathium from side with anthers and bracteole (scale 1 mm, as for E, F). E, F, female floret. Drawn from: A, C, D, E, PVB 10615, south of Oudtshoorn, South Africa. B, F, PVB 11513, Toorwater, South Africa (© PVB).

Euphorbia crassipes Marloth, Trans. Roy. Soc. South Africa 1: 318 (1909). Type: South Africa, Cape, Biesiespoort, Marloth 4399 (4397) (PRE, holo.; K, iso.). Euphorbia fusca Marloth, Trans. Roy. Soc. South Africa 2: 38 (1910). Type: South Africa, Cape, Britstown, Sept. 1909, Marloth 4682 (PRE, holo.; K, iso.). Euphorbia baliola N.E.Br., Fl. Cap. 5 (2): 327 (1915). Type: Namibia, Great Karas Mountains, between 1st & 2nd outspan between Kraikluft and Narudas Süd, 5400’, 26 Dec. 1912, Pearson 8095 (K, holo.; BOL, GRA, SAM, iso.). Euphorbia inornata N.E.Br., Fl. Cap. 5 (2): 586 (1925). Euphorbia inelegans N.E.Br., Fl. Cap. 5 (2): 322 (1915), nom. illegit., non N.E.Br. (1911). Type: South Africa, Cape, near Kimberley, Moran (sub Schönland) 1718 (K, holo.; GRA, K, iso.). Euphorbia eendornensis Dinter, Feddes Repert. Spec. Nov. Regni Veg. 30: 196 (1932). Type: Namibia, between Witsand and Eendorn, 26 Mar. 1924, Dinter (missing). Neotype (Bruyns 2012): Namibia, Vrede, Bruyns 11362 (NBG). Euphorbia hopetownensis Nel, Kakteenk.: 192 (1933). Lectotype (Bruyns 2012): South Africa, Cape, Hopetown, 1930, E. Markoetter sub SUG 5529 (missing), photograph in Kakteenk.: 192 (1933).

Bisexual spineless (but usually with many somewhat spiky persistent peduncles) glabrous succulent 80–400 mm tall with broadly clavate tuberculate stem 50–200 mm thick projecting from the ground for 30–300 mm, densely covered around apex with many branches (apex usually free of branches), tapering rapidly beneath ground into tap-root bearing fibrous roots. Branches ascending to spreading, simple, cylindrical and slightly thickened towards bases, 20–150 × 6–16 mm, covered with tubercles, smooth, dark green to brown-green when young, later grey-green from waxy covering; tubercles low and not prominent, conical, 1–3  mm long with elongated hexagonal base, vertically arranged into 8–10 obscure spiralling rows, with fine grooves between them, occasionally with solitary simple persistent peduncles in their axils; leaf-rudiments on tips of new tubercles towards apices of branches and of stem, 1–3 × 1–2 mm, spreading, fleeting, ovate-deltate, margins slightly folded upwards and finely ciliate, shortly acute, sessile. Synflorescences 1–10 per branch towards apex, each of one bisexual cyathium arising in axil of tubercle on brownish green peduncle (5) 10–28  mm long, with 3–6 spathulate caducous red-brown bracts 1–2 mm long and glabrous except for ciliate margins; cyathia very shallowly conical, glabrous, 5–8  mm broad (1–2 mm long below insertion of glands), with 5 red-green

2.1 Sect. Anthacanthae

to dark red-brown lobes with deeply incised white-ciliate margins, dark green to brown-green; glands 5–6, ± elliptic, 2–3  mm broad, spreading and usually somewhat deflexed, dark green (rarely olive-green) or purple-green to dark brown, convex and finely papillate above, outer margins divided into 3–7 spreading to reflexed finger-like cream to pale green processes to 2 mm long sometimes further finely subdivided near tips; stamens sparsely pubescent on pedicel (anthers often dark purple); bracteoles filiform and pubescent, ovary globose, sparsely to densely pubescent above,

283

raised on pedicel 0.5–1  mm long; styles 1.5–3  mm long, branched in upper third to half. Capsule 6–8  mm diam., obtusely 3-angled, sparsely pubescent, brown-green, sessile. Distribution & Habitat Euphorbia crassipes is a widely distributed species on the Great Karoo, where it is known from north-east of Kuruman (towards the border with Botswana) to around Postmasburg in Griqualand West and southwards to Williston, Britstown

Fig. 2.397.  Distribution of Euphorbia crassipes (© PVB).

and Phillipstown. It is also found, though it remains very scantily recorded, in Bushmanland from near Kliprand (very close to the edge of the winter-rainfall region) northwards to Pofadder and eastwards towards Kenhardt. Another area where it is known is the south-eastern portion of Namibia on the upper plateau of the Great Karas Mountains and in the area to the south-east of Warmbad. Plants generally grow in flat or gently sloping areas among low karroid bushes (with a few larger shrubs of Lycium and Rhigozum trichotomum) and tufts of grass on hard, loamy ground, often with scattered jasper-like pebbles

and chips of limestone on the surface. South-east of Warmbad, E. crassipes grows in finely gravelly, gneissic ground. In the slightly moister parts of its distribution, older plants are generally found growing fully in the open. In places with a harsher climate (such as south-east of Warmbad, Namibia, where the average rainfall is only 75 mm per year or west of Kenhardt, where about 125  mm is received per year), most specimens are found sheltering next to or inside a bush. Here they are also few and far between, which is in strong contrast to the quite large colonies that develop in some areas (such as near Britstown).

284

2  Euphorbia subg. Athymalus

Fig. 2.398.  Euphorbia crassipes, PVB 11352, plateau of the Great Karas Mountains, NE of Grünau, Namibia, 5 Apr. 2009 (© PVB).

Fig. 2.399.  Euphorbia crassipes, large plant close to 30 cm diam., PVB 11362, SE of Warmbad, Namibia, 7 Apr. 2009 (© PVB).

Diagnostic Features & Relationships In E. crassipes each plant consists of a swollen stem that may protrude up to 0.3 m from the soil and tapers rapidly beneath the surface into a slender but substantial taproot. The stem often has a furrow-like groove at the apex (mentioned by White et  al. 1941 for E. fusca, but not by N.E. Brown for either E. crassipes or E. fusca), but this is not always present and seems to develop only in older plants. The flat to slightly depressed apex of the stem usually remains bare and is surrounded by the typically short, stubby, cylindrical, finger-like branches that become slightly thicker towards their bases and have a rounded end. The branches

are covered with low tubercles and often bear the blunt, persistent remains of dried-out peduncles. Altogether (taking both the stem and branches into account) an undamaged, mature specimen may form a nearly perfect sphere just above the ground. In habitat the plants may have a brownish green colour and young growth is often particularly dark brown in colour, though this is usually obscured later by a thick covering of wax, giving them a greyish appearance. The brownish colour of the stem and branches is created by a dense arrangement of fine red-brown dots on a grey-green surface. When cut, the branches give off a strong, acrid odour.

2.1 Sect. Anthacanthae

285

Fig. 2.400.  Euphorbia crassipes, around 25 cm diam., PVB 11678, north of Groblershoop, South Africa, 14 May 2010 (© PVB).

Fig. 2.401.   Euphorbia crassipes, very large, old plant with stem exposed by erosion, just south of Britstown, South Africa, 25 Mar. 2013 (© PVB).

Flowers are produced from late September into November (or later, depending on rain), with few to many of the particularly shallow cyathia developing at the tips of each branch in the axils of the newer tubercles. Each cyathium is borne well away from the branch on a peduncle that is also usually darkly coloured. This dark colouring is brought about by fine, red-brown or red, longitudinal streaks on a green background and is also present on the cyathial lobes, the outside of the cyathium, the backs of the cyathial glands and on the capsules (and is similar to the fine dots that give the brownish colour to the stem and branches). Several small bracts are present along the peduncle, but these are usually shrivelled

by the time the cyathium matures. These peduncles dry out after the cyathium falls off and persist as hard, but not sharptipped spikes along the branch, usually wearing off by the time they reach the middle of the branch (though they may persist longer in the plants south-east of Warmbad, Namibia and in those from north-east of Kuruman). Each cyathium bears five often slightly reflexed glands, whose upper surface is uniformly dark green to brown and whose outer margin is ornamented by slender, cream to pale green slightly descending, finger-like processes, often finely subdivided near their tips. The glands and lobes usually surround a contrastingly bright, white, woolly centre surrounding the anthers and the

286

2  Euphorbia subg. Athymalus

Fig. 2.402.  Euphorbia crassipes, flowering, cyathia becoming greenish with age, PVB 10095, south of Pofadder, South Africa, Aug. 2006 (© PVB).

green to brown styles. The anthers start off red-green and ripen to red-black or black, which contrasts strongly against the bright yellow pollen. In most areas the cyathia emit a faint, slightly foetid odour but those on plants from the Great

Karas Mountains and from south-east of Warmbad, Namibia emit a sweetish odour. The very thick stem and relatively short, nearly cylindrical branches with low tubercles that scarcely protrude from the surface, separate E. crassipes from most other similar

Fig. 2.403.  Euphorbia crassipes, flowering and fruiting, Wigton, north of Hopetown, South Africa, 2 Jan. 2010 (© PVB).

species. This distinctiveness continues into the shallow, dark-coloured cyathia with their dark glands and slender, finger-like processes. Euphorbia crassipes often occurs with E. braunsii, but these two are readily distinguished, especially by the fact that in E. braunsii the stem is not differ-

ently-shaped from the branches. In size, plants of E. crassipes may rival E. esculenta, but these two are not close and it is to E. decepta that one must compare E. crassipes most carefully (see below).

2.1 Sect. Anthacanthae

287

Fig. 2.404.  Euphorbia crassipes, flowering and with pale, slightly shiny capsules, PVB 4702, north of Kliprand, South Africa, 20 Oct. 2008 (© PVB).

History Marloth described both E. crassipes and E. fusca. Euphorbia crassipes was described from material he collected at Biesiespoort, north of Beaufort West and no date was given for this collection. In the original publication it was cited as Marloth 4397. The specimen in PRE, which is assumed to be the type, is labelled ‘4399 (4397)’, with the annotation ‘originally 4397, to avoid confusion with E. albert. changed to 4399’, in Marloth’s own hand. This information and the locality is lacking on the isotype at Kew, which consists only of four detached cyathia. Another specimen, also with the number Marloth 4397, this time from Prince Albert and preserved at Kew, was described by Brown as E. albertensis and is included here under E. decepta. Euphorbia fusca was described by Marloth from material that he had gathered in September 1909 near Britstown. His description of E. fusca is particularly brief and somewhat casual (in it he even referred to the cyathia as ‘flowers’) and he mentioned other collections from Kimberley, Steynsburg and Britstown. The first recorded gathering of E. crassipes was made by H.H.W.  Pearson in December 1908 between Kliprand and Aalwynsfontein on the south-western edge of Bushmanland. Euphorbia crassipes is not well-known from Namibia. They are uncommon in the scanty but surprisingly diverse vegetation around Eendoorn, to the south-east of Warmbad. In December of 1912, Pearson collected E. baliola in the

Great Karas Mountains (well before Dinter collected E. crassipes further south around Eendoorn). This was recollected in April 2009 and found also to belong to E. crassipes, with the same growth-habit and the typically small, darkish, shallow cyathia with quite prominent, pale finger-like processes on the glands. Marloth (1910) compared E. fusca to E. caput-medusae and E. crassipes, saying that it differed from E. crassipes by the non-persistent peduncles (some peduncles being persistent in E. crassipes: in fact the type in PRE is a remarkably stout plant about 10 cm tall, 11 cm broad arising from a taproot about 15 mm thick with a dense head of stout branches 30–50 × 6–8 mm, most of which have 8–15 stout, blunt, tapering, whitish spikes ± 10 × 1.5 mm near their tips). In most populations of E. crassipes one finds plants with persistent peduncles and others without them so this character cannot be used to distinguish between them. In addition, N.E. Brown (1915) observed that Marloth had been wrong about the lack of persistent peduncles in E. fusca, where Brown found that they were present. White et  al. (1941) maintained that the main differences between E. crassipes and E. fusca were the slightly more cylindrical stem, thicker branches, the deeper involucres and the green glands. However, the type of E. crassipes at PRE is a typical specimen of what is usually referred to as ‘E. fusca’ and the various differences are insignificant in this widely distributed and quite variable species where populations with green glands and others with purple glands are known. Since the

288

2  Euphorbia subg. Athymalus

Fig. 2.405.  Euphorbia crassipes. A, tip of branch with leaf-rudiments (scale 2 mm). B, cyathium from side (scale 2 mm, as for C, D). C, differently shaped processes on cyathial glands. D, part of dissected cyathium from side with anthers and bracteoles. E, anther and rudimentary female (scale 1 mm, as for F–H). F, part of dissected cyathium with female floret. G, H, female floret. Drawn from: A–E, H, PVB 4702, north of Kliprand, South Africa. F, PVB 10095, south of Pofadder, South Africa. G, PVB 13115, near Williston, South Africa (© PVB).

name E. crassipes was published before ‘fusca’, it must replace the well-known name Euphorbia fusca. Nel’s E. hopetownensis was described from a small plant (only 5 cm broad) with ascending, relatively stout branches which bore unusually short peduncles at 5–7 mm long and ‘pink-purple’ glands with five teeth. The small figure in the text and these few details are strongly suggestive of E. crassipes, under which this name was subsumed (Bruyns 2012). In the case of E. inornata, the glands were given as olive-green on their upper surface, which is paler that usual but not unknown for E. crassipes. Nevertheless, the shape of the plant, the relative thickness of the branches and the shape of the cyathia and glands all fit E. crassipes and so this name was also included under E. crassipes. Euphorbia davyi Pax ex N.E.Br., Fl. Cap. 5 (2): 305 (1915). Lectotype (Bruyns 2012): South Africa, Transvaal, near Pretoria, received at Kew 19 Nov. 1901, J.W.C. Kirk 48 (K; PRE, iso.). Bisexual spineless densely to loosely branched nearly hemispherical glabrous succulent 50–150 (−200) × 50–400

mm, consisting of mainly subterranean clavate to truncate prominently tuberculate stem 50–150 × 30–60  mm (25– 50 mm tall and pale green above ground) bearing crown of branches around bare apex, stem gradually tapering below into slender tap root bearing fibrous roots. Branches many, ascending, cylindrical to clavate, mainly simple, 30–150 × 10–30 mm, densely covered with tubercles, smooth, greygreen sometimes suffused with red; tubercles prominent and conical, spreading to slightly reflexed near tip, 3–5 mm long with ±  hexagonal base, often spirally arranged into 8–10 rows, with shallow grooves separating tubercles and a white apical scar from leaf; spines absent, but few grey to white woody persistent fertile or sterile peduncles present on some branches, 8–20 × 2–3 mm, spreading and not sharp-pointed once dried out; leaves on tips of new tubercles at apices of branches and stem, 10–30 × 1.5–4 mm, ascending and then spreading, deciduous, linear to linear-oblanceolate, acute, sessile and tapering slightly into tubercle, longitudinally folded, grey-green sometimes with faint paler midrib, margins straight and reddish. Synflorescences 1–many per branch towards apex, solitary in axil of tubercle on green to browngreen sometimes persistent peduncle 2–5 × 1–3 mm with 1

2.1 Sect. Anthacanthae

terminal cyathium, glabrous, with 4–8 oblanceolate to spathulate glabrous caducous bracts (with marginal cilia in upper half) 1.5–3.5  mm long along peduncle; cyathia cupular to slightly conical, glabrous, 5–8  mm broad (2.5  mm long below insertion of glands), with 5 reddish green lobes with finely incised ciliate margins, green often finely streaked with red or brown; glands 5, broadly cuneate, not contiguous, 1–2 mm broad, spreading, green to brown, flat to slightly convex above, with raised inner margin, outer margins divided into 2–5 spreading and slightly deflexed sometimes

289

reddish irregular short finger-like teeth 0.5–1 mm long; stamens with pubescent pedicels, bracteoles filiform and pubescent; ovary ellipsoidal and slightly 3-angled, glabrous, raised on short pedicel up to 0.5 mm long; styles 2.5–3.5 mm long, branched and slightly spreading in upper third. Capsule 5–7 mm diam., obtusely 3-angled, glabrous, almost sessile. Distribution & Habitat Euphorbia davyi is mainly found between Pretoria and Nylstroom, extending to about 20 km east of Pretoria.

Fig. 2.406.  Distribution of Euphorbia davyi (© PVB).

Habitats range from flat areas in ‘black turf soils’ among tufts of grass and trees, to stony patches on the Magaliesberg, between tufts of grass and outcrops of sandstone. Plants may be quite common in the latter situation, growing both in crevices between rocks and in small patches of deeper ground between outcrops of rock. Here other succulents such as Adromischus umbraticola, Aloe davyi, A. pretoriensis and Kalanchoe lucieae are plentiful, but trees are mostly absent. Diagnostic Features & Relationships Plants of E. davyi are very variably shaped, from small to large rosettes of ascending branches and sometimes also

dense hemispherical mounds of branches around the central stem. The stem generally does not rise more than about 10 cm above the surface of the ground, so that the branches are the most visible part of the plant. The branches vary in length but are mainly ascending and cylindrical and are covered with prominent, conical tubercles. Leaves are produced in clusters at the apex of the branches as the branches grow, each arising on a new tubercle as it develops. The leaf persists for most of the growing season and falls off near its end, leaving an often quite obvious, white scar at the tip of the tubercle. The leaves are particularly prominent and are usually bright grey-green.

Fig. 2.407.  Euphorbia davyi, among pieces of sandstone with a fine Selaginella, PVB 12213, east of Pretoria, South Africa, 20 Sep. 2012 (© PVB).

Fig. 2.408.  Euphorbia davyi, ± 25 cm diam., PVB 12213, east of Pretoria, South Africa, 20 Sep. 2012 (© PVB).

Fig. 2.409.  Euphorbia davyi, stout branches, leaves not yet developed fully, PVB 12213, east of Pretoria, South Africa, 20 Sep. 2012 (© PVB).

2.1 Sect. Anthacanthae

291

Flowering takes place in E. davyi early in the growing season, from September to November and possibly later. Many cyathia are produced near the apex of each branch, each arising in the axil of a tubercle. They are fairly insignificant, on short peduncles that scarcely raise them above the surface of the branch and are green or suffused with red (giv-

ing them a brownish hue). The glands of the cyathia are green, with short teeth on their outer margins that are variable in length and number. Euphorbia davyi is very similar to E. duseimata and E. maleolens and the differences between them are discussed under E. maleolens.

Fig. 2.410.  Euphorbia davyi, cyathia on short peduncles, PVB 12213, east of Pretoria, South Africa, 20 Sep. 2012 (© PVB).

Fig. 2.411.  Euphorbia davyi, with capsules, PVB 12213, east of Pretoria, South Africa, 20 Sep. 2012 (© PVB).

History Euphorbia davyi was described from a specimen collected by J.W.C.  Kirk and two specimens of J.  Burtt-Davy. Kirk was the first to collect it, in 1899 or 1900 while stationed at Pienaarspoort near Pretoria and his specimen was forwarded

to Kew by his father, Sir John Kirk, in November 1901. Burtt-Davy’s number 5562 was collected in December 1905, while the date of collection of the other number cited is unknown.

Fig. 2.412.  Euphorbia davyi. A, cyathium from side (scale 2 mm, as for B, C). B, C, cyathium from above showing variation in the shape of the glands. D, anthers and bracteole (scale 1 mm, as for E). E, female floret. Drawn from: PVB 12213, east of Pretoria, South Africa (© PVB).

2  Euphorbia subg. Athymalus

292

Usually the author of this species is cited as ‘N.E.Br.’. However, in the Kew Herbarium there is a letter written by Ferdinand Pax from the Royal Botanic Garden in Breslau (then in Prussia, today known as Wroclav and situated in Poland) on 17 Feb. 1906 to Burtt-Davy (and presumably forwarded by Burtt-Davy to Brown) that requested Burtt-Davy’s permission to name this species after him. Consequently, Pax first suggested the name Euphorbia davyi and so it is cited here as ‘Pax ex N.E.Br.’. Euphorbia decepta N.E.Br., Fl. Cap. 5 (2): 320 (1915). Type: South Africa, Cape, Beaufort West district, Willowmore side, 11 Sept. 1911, Brauns (sub Schönland) 1712 (K, holo.). Bisexual spineless glabrous succulent 50–150 (200) mm tall with globose and nearly spherical to clavate tuberculate brownish green stem 20–60 (100) mm thick projecting from the ground for 30–200 mm, tapering beneath ground rapidly into tap-root bearing fibrous roots, with many branches forming rosette 50–150 mm diam. around apex, apex sometimes without branches. Branches ascending to spreading, simple, cylindrical and slightly thickening towards bases, 5–50 × 4–8  mm thick, covered with tubercles, smooth, brownish green; tubercles low and not prominent, conical, 1–4 mm long (much broader on stem than on branches) with elongated rhomboidal to hexagonal base, vertically arranged into 5–8 obscure spiralling rows, with fine grooves around them, with slender but firm and slightly woody persistent

sterile and fertile peduncles in some axils; leaf-rudiments on tips of new tubercles towards apices of branches and of stem, 0.5–3 × 0.5–1 mm broad, erect, fleeting, ovate to linear-lanceolate, margins folded upwards and eciliate, obtuse, sessile. Synflorescences 1–8 per branch mainly towards apex, each of one bisexual cyathium in axil of tubercle on peduncle 4–15 mm long, with 3–5 slightly spathulate caducous bracts 0.5–1.5 mm long and glabrous except for cilia along margins near apex; cyathia deeply conical-cupular, glabrous, 4–7 mm broad (± 2 mm long below insertion of glands), with 5 lobes with finely divided margins, pale green; glands 5, ± cuneate to transversely rectangular, 1.5–2.5  mm broad, spreading and widely separated, dark green, convex to concave above, inner margins somewhat raised, outer margins usually divided into 3–6 short stubby slightly yellowish to dark green teeth; stamens with pubescent pedicels, bracteoles filiform and pubescent; ovary globose, pilose usually mainly towards top, sessile to raised on pedicel to 0.7  mm long; styles 2–4 mm long, branched only near apex into broad spreading slightly bifid lobules. Capsule 6–7  mm diam., obtusely 3-angled, grey-green, glabrous, sessile. Euphorbia decepta is one of the smaller species of ‘vingerpol’, along with such species as E. arida and E. procumbens. Its stem never reaches the massive thickness that will often be observed in mature plants of E. crassipes or E. esculenta and the numerous branches are also more slender and shorter than in these larger species. Two subspecies are recognised that may be separated as follows:

Fig. 2.413.  Distribution of Euphorbia decepta (● = ssp. decepta; ▲ = ssp. gamkaensis) (© PVB).

2.1 Sect. Anthacanthae

293

1. Styles 2.5–4 mm long, branched near top into slightly spreading lobules...................................................subsp. decepta 1. Styles 1.5–2 mm long, branched in upper half into spreading lobules.................................................subsp. gamkaensis

Euphorbia decepta subsp. decepta Euphorbia albertensis N.E.Br., Fl. Cap. 5 (2): 323 (1915). Type: South Africa, Cape, near Prince Albert, Marloth 4397 (K, holo.; PRE, iso.). Euphorbia suppressa J.G.Marx, Cact. Succ. J. (US) 71: 33 (1999). Type: South Africa, Cape, near Seekoeigat, ± 950 m, Marx 227 (GRA, holo.). Euphorbia brevirama N.E.Br., Fl. Cap. 5 (2): 317 (1915). Type: South Africa, Cape, Jansenville div., near Klipplaat, Schönland 1716 (K, holo.; GRA, iso.).

Euphorbia astrophora J.G.Marx, Cact. Succ. J. (US) 68: 311 (1996). Euphorbia brevirama var. astrophora (J.G.Marx) J.G.Marx & Van Veldh., Euphorbia World 11: 8 (2015). Type: South Africa, Cape, north of Klipplaat, 10 Jan. 1995, Marx 204 (GRA, holo.). Euphorbia brevirama var. supraterra J.G.Marx & Van Veldh., Euphorbia World 11: 9 (2015). Type: South Africa, near Letjesbos siding, Apr. 1999, Marx 559 (GRA, holo.).

Fig. 2.414.  Euphorbia decepta ssp. decepta, among shrubs of Ruschia spinosa, PVB 12965, SE of Beaufort West, South Africa, 2 Sep. 2015 (© PVB).

Stem 20–60 (100) mm thick in mature plants, usually projecting from ground for 30–200  mm. Branches ascending, 5–50 × 4–8 mm thick, usually with many persistent dried-out whitish peduncles. Cyathia deeply conical-cupular, 4–7 mm broad (±  2.5  mm long below insertion of glands); glands cuneate to transversely rectangular, above convex to concave or with deep depression in middle, 1.5–2  mm broad, dark green, outer margins crenate to divided into 2–4 short stubby dark green to paler teeth; ovary globose, pilose, raised on pedicel