Paleoethnobotany of the Kameda Peninsula Jomon 9780932206954, 9781949098983

Citation preview

ANTHROPOLOGICAL PAPERS

MUSEUM OF ANTHROPOLOGY, UNIVERSITY OF MICHIGAN NO. 73

PALEOETHNOBOTANY OF THE KAMEDA PENINSULA JOMON

by GARY W. CRAWFORD

ANN ARBOR, MICHIGAN

1983

© 1983 Regents of The University of Michigan The Museum of Anthropology All rights reserved Printed in the United States of America ISBN 978-0-932206-95-4 (paper) ISBN 978-1-949098-98-3 (ebook)

CONTENTS

Figures ................................................... v Tables .................................................... VI Plates .................................................... VB Acknowledgments .......................................... ix Abstract in Japanese, translated by Masao Nishimura ......... xi I. INTRODUCTION .................................... . Research Objectives .................................. 4 Physical Setting ..................................... 5 Vegetation History ................................... 15 Archaeology ........................................ 16 Jomon Subsistence ................................... 23 2. PLANT REMAINS IDENTIFICATIONS AND ECOLOGICAL DATA ................................. 29 Sasa ............................................... 30 Nuts ............................................... 30 Grain Seeds ......................................... 3 I Fleshy Fruit Seeds ................................... 43 Other Seeds ......................................... 45 3. METHODOLOGY ..................................... 49 Laboratory Analysis ................................. 59 4. HAKODATE AIRPORT AND YAGI SITE PLANT REMAINS .................................... 61 Nakano B and Locality 4 ............................ 61 Yagi Site ........................................... 65 5. HAMANASUNO SITE PLANT REMAINS .............. 85 Plant Remains Analysis .............................. 87 Middle Jomon Component ........................... 107 Plant Remains Distributions .......................... 109

iii

6.

USUJIRI B SITE PLANT REMAINS ................... Saibesawa VII Houses ............................... Daigi Houses ....................................... Nodappu II and Unspecified Assemblages .............. Plant Remains Distributions .......................... 7. INTERPRETATIONS AND CONCLUSIONS ............ Utilization of Represented Plants ...................... Initial Through Middle Jomon: Intersite Comparisons ... North America: A Search for Analogies ............... Ecological Interpretations ............................. Implications ......................................... APPENDIX 1. The Vegetation of Minamikayabe ............ A. Beech Forests .......................... B. Hill Vegetation ......................... C. Coastal Vegetation ...................... APPENDIX 2. Hamanasuno Site Flotation Samples: Number of Carbonized Seeds ................ APPENDIX 3. Usujiri B Site Flotation Samples: Number of Carbonized Seeds ................ APPENDIX 4. Plant Remains from the Seizan Site .......... BIBLIOGRAPHY .........................................

iv

113 115 129 133 136 141 142 146 151 153 157 165 165 172 180 184 187 189 191

FIGURES

I. 2. 3. 4. 5. 6. 7. 8. 9. 10. II. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

Oshima Peninsula ..................................................... Locations of some Hokkaido Initial, Early, and Middle lomon sites ....... Location of the Usujiri B, Hamanasuno, and Yagi sites ................... Location of the Nakano B site and Locality 4 of the Hakodate Airport site ................................................. Frequency distribution of barnyard grass ................................ Froth flotation apparatus .............................................. Nakano B site, Grid 3 ................................................ Nakano B site, House II .............................................. Yagi site ............................................................. Yagi site flotation samples ............................................. House 3, Yagi site .................................................... Yagi site, House 5 .................................................... Hamanasuno site ..................................................... Hamanasuno site, House 60 ........................................... Hamanasuno site flotation samples ..................................... Hamanasuno site, Houses 62 and 63 .................................... Hamanasuno site, Houses 70, 71, and 74 ................................ Hamanasuno site, Houses 72 and 73 .................................... Usujiri B site ......................................................... Usujiri B site, Houses 6 and 12 ........................................ Usujiri B site, Saibesawa VII component flotation samples ................ Usujiri B site, House 21 ............................................... Usujiri B site, Daigi VIIIb component flotation samples .................. Usujiri B site, Nodappu II component and unspecified assemblages, flotation samples ...................................................... Usujiri B site, Houses 9b and 20, and burial ............................ Summary of seed types as percentage of total number of seeds and nuts as percentage of total sample weight ............................... Summary of seed classes as percentage of total number of seeds and nuts as percentage of total sample weight ...............................

v

6 8 10 12 33 53 63 64 67 75 79 81 86 92 94 98 101 104 116 118 119 126 131 134 137 147 148

TABLES

I. 2.

3.

4. 5.

6. 7. 8. 9. 10. I I.

12.

Initial through Middle Jomon chronology in southwestern Hokkaido ....... Flotation method comparison for Usujiri B, House 20, level X3: light fraction and seed densities and seed types as percentage of total number of seeds ...................................................... Flotation method comparison for Usujiri Band Hamanasuno: light fraction and seed densities and seed types as percentage of total number of seeds ...................................................... Nakano B, House II, flotation samples: contents as percentage of total sample weight and light fraction and seed densities ....................... Nakano B, House II, flotation samples: carbonized plant remains components as percentage of total weight of carbonized plant remains and number of seeds .................................................. Vagi site flotation samples: number of carbonized seeds .................. Hamanasuno site, 1974 flotation samples: carbonized seeds as percentage of total number of carbonized seeds .................................... Hamanasuno site flotation samples (1976-77): carbonized seeds as percentage of total number of carbonized seeds .......................... Usujiri B site, Saibesawa VII component, flotation samples: carbonized seeds as percentage of total number of carbonized seeds .................. Usujiri B site, Daigi VIIIb component, flotation samples: carbonized seeds as percentage of total number of carbonized seeds .................. Usujiri B site, Nodappu II component and unspecified assemblages, flotation samples: carbonized seeds as percentage of total number of carbonized seeds ...................................................... Presence of five taxa in flotation samples as a percentage of the total number of samples per subphase .......................................

vi

20

56

57 65

66 70 90 91 120 132

135 150

PLATES

I. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

Echinochloa type seeds (barnyard grass) ................................. Type I grass ......................................................... Type 2 grass ......................................................... Grass caryopses (probably Paniceae tribe) from Hamanasuno .............. HNS type knotweed seeds and Type A knotweed seeds ................... {jinutade seeds and Type B knotweed seeds ............................. Carbonized buckwheat (Fagopyrum esculentum Moench) .................. Dock seeds and chenopod seeds ........................................ Selected seeds: I. Dogwood, 2. OSlrya, 3. Malalabi (Actinidia), 4. Sumac, 5. Amur corktree, 6. Udo (Aralia) ...................................... Hamanasuno site, House 70 ........................................... Hamanasuno site, House 71 ........................................... Hamanasuno site, House 72 ........................................... Hamanasuno site, House 73 ........................................... Hamanasuno site, House 74 ........................................... Hamanasuno site, House 74 ........................................... Hamanasuno site, House 78 ........................................... Usujiri B site, 1977 excavation. View to the north ....................... Usujiri B site, House 10 ............................................... Usujiri B site, House 10 ............................................... Usujiri B site, House 16 ............................................... Usujiri B site, House 21 ............................................... Usujiri B site, House 21, Pits I and 2 .................................. Usujiri B site, burial ..................................................

vii

32 36 37 38 39 40 41 42 46 100 102 105 106 108 110 III 114 122 124 125 128 130 138

ACKNOWLEDGMENTS

Three generous Canada Council graduate fellowships (numbers 45-762, W762513, and 453-784807) made this research possible. The Yagi project was supported by grants from the National Geographic Society (2173), National Science Foundation (BNS 7803855), and the Social Sciences and Humanities Research Council of Canada (410-79-0105) to Peter Bleed and William Hurley. This book has been published with the help of a grant from the Social Science Federation of Canada using funds pro­ vided by the Social Sciences and Humanities Research Council of Can­ ada. To all of these agencies I express my heartfelt gratitude. The Minamikayabe Board of Education was, and is continuing to be, supportive in every respect. I thank the Board members for their gracious help. Tadahisa Ogasawara deserves special thanks for his assistance throughout the period of my research. His help was instrumental to any success that this project has attained. Masakazu Yoshizaki was responsible for all arrangements in Japan necessary for conducting field work there. Furthermore, he acted and still acts as a patient guide to the pleasures and problems of working in Japan and has taught me many of the com­ plexities of Hokkaido archaeology. To him I extend my deep gratitude. I am extremely grateful to William Hurley for his support and assistance throughout the period of this project. I thank both him and Peter Bleed for their permission to include the Yagi site data in this monograph. Eisuke Yokoyama and Hajime Chiyo made available the Hakodate Air­ port soil samples for study. In addition, I acknowledge the cooperation of the Hakodate City Museum and the Hakodate City Board of Education. The people of Minamikayabe made living in their town a wonderful experience. In particular, I am indebted to the Takehara family for their fine hospitality. I thank Yasuyo Tsubakisaka for her kind assistance during my visits to Sapporo. At the Botany Department of the University of North Carolina, Jimmy Massey made available the herbarium collec­ tion for my study. I am also grateful for the use of the UNC Botany ix

Department photography laboratory and the assistance of its pho­ tographer and artist-Betsy Burgner. Some of the plates were provided by Steven Jaunzems and John Brodie of the Erindale Photography Department at the University of Toronto. The lettering in many of the figures was provided by Mabel Riendeau. Others who have assisted in aspects of this project include: Mr. Akutsu, Clare Fawcett, Theresa Ferguson, David Johnson, Tatsuro Matsuoka, Mr. Nakai, Koichi Naoi, Susan Rowley, Koichiro Tsurumi, and the Yagi site excavation teams of 1979 and 1980, which included, in addition to some of the individuals already mentioned: Connie Bennett, Carole Lord, Hiroaki Miya, Douglas Rice, and KaoruTanaka. Many others provided advice, hospitality, and materials. They include Ann and Peter Bleed, Kisao Ishizuki, Kankichi Sohma, J. Edward Kidder, Y oshinobu Kotani, Michiko Morii-Chiura, Chosuke Serizawa, and the T ohoku Botanical Gardens. This book is a revised version of my doctoral dissertation written at the University of North Carolina. Clara Stewart typed the manuscript. Richard Yarnell, Donald Brockington, Joffre Coe, Carole Crumley, John D. Eyre, and William Hurley served on my committee for the original dissertation. Richard Yarnell's guidance throughout this study is especially appreciated. Richard Ford and Barry Stewart gave helpful advice in completing this monograph. In addition, this manuscript has benefited from the comments of several anonymous reviewers. I thank them for their valuable input. The final responsibility for the contents of this monograph is, of course, my own.

x

JAPANESE ABSTRACT

xi

*~v;t,

~ti'itL@:@Wnm 0) ttrn ~~ vcJilH± LJ..::*Jll!3t~f

"

E

S

64.9

2.4 48.8

2.9 41.2 33.3 83.3

-

3.3 56.7 -100.0 -100.0

8.8 52.9

20.0 33.3 72.7 62.5

;:>

C

...,C

~ 0

C

Other Seeds

Table 3. Flotation Method Comparison for Usujiri Band Hamanasuno: Light Fraction and Seed Densities; and Seed Types as Percentage of Total N umber of Carbonized Seeds

~

'-l

V,

~

c;')

0 t'-< 0

§

~

58

KAMEDA lOMON PALEOETHNOBOTANY

there is an actual difference in the archaeological soil contents. This interpretation is supported by the observation that the other three sam­ ples have 65% to 75% grain seeds with knotweed seeds comprising roughly 40% of the total number of seeds, followed by grasses at between 24% and 30%. The main contrast between the froth samples and the other samples seems to be a factor of sample size. Seeds which are rare in the large froth floated samples such as elderberry, grape, and blackberry, do not show up in the small non-froth samples. Another contrast, of yet unknown signifi­ cance, is the variability of the unidentifiable seed percentages. These seeds have their identifying characteristics obscured. Few of the unidentifiable seeds seem to be freshly broken or otherwise distorted, indicating that the different flotation methods do not differentially damage seeds. Seed densities are comparatively low in the other samples, with most seed types represented by only one seed each per sample portion. In summary, only one of the samples taken for flotation method comparison contained enough information to make a reliable evaluation. The three methods do not appear to select differentially for seeds or types of seeds, contrary to the results anticipated by Jarman et al. (1972). The immersion methods seem to be effective in collecting subsistence informa­ tion relevant to this study. My purpose was not to determine the absolute effectiveness of each method but to evaluate their comparability, and for the purposes of this study, they will be considered to be directly com­ parable. In all the samples, wood charcoal collection may have been b"iased, and hence it may be under-represented in the froth samples. This would be expected since the soils were screened, prior to froth flotation, to remove small compacted soil clods which probably contained some wood charcoal. The principal advantage of the froth apparatus was its increased effi­ ciency over the immersion methods. Jarman et al. (1972:45) report that 1000 to 2000 liters of soil from Hahal Oren in Palestine could be pro­ cessed in a day "under optimum working conditions." The flotation rate in Japan was considerably lower. The maximum quantity of soil that we processed in a day was 374 liters from a single pit. On most good days, about 130 to 150 liters, and occasionally 200 liters were processed. This rate was still two to three times that of the immersion techniques. A number of factors influenced this rate: the number of people processing the samples (usually only one or two), the number of samples floated (one large sample can be floated faster than several smaller ones which require a greater cumulative clean-up time), and the soil which often tended to clump in the flotation cell and made discharge from the cell time-con­ suming. Furthermore, flotation was carried out concurrently with the

METHODOLOGY

59

excavation on a daily basis. Not only did I have to operate the machine myself, but I had to supervise excavation, soil sampling, and often collect samples.

LABORATORY ANALYSIS Analysis of the flotation samples followed a standardized procedure which closely follows the one outlined by Richard Yarnell (1974: 113-14). A sample is weighed and then passed through a series of 10 geological sieves of the following mesh sizes: 6.35, 4.00, 2.83, 2.38,2.00, 1.41, 1.00, 0.71, 0.42, and 0.21 millimeters. Each of the 11 fractions into which a sample is divided is weighed. The first sample weight and the fraction weights are not used in the final tabulations of the sample composition but are used as a check against subsequent weighings. The fractions above the 2.38 millimeter screen are separated entirely into their constituent components, which may number as many as nine: wood charcoal, nut­ shell, amur cork tree berry remains, sasa stem fragments, unidentified plant remains, bone, stone flakes, mineral (mainly pumice), and uncar­ bonized organic material. Each of these components is weighed. Mineral and uncarbonized organic material are not included in the final results. Each of the other components is reported as a percentage of the total component weight. Below the 2.38 millimeter screen only carbonized seeds are removed. The seeds are weighed together as a component. After the seeds have been removed, the fractions from the 2.00,1.41, and 1.00 millimeter screens are combined and weighed together. The fractions with particles smaller than 1.00 millimeter are examined, but their weights are not included in the final calculations and tabulations. The expected weight of each com­ ponent that would be found in the entire sample down to the 1.00 millimeter screen is then calculated. It is based on the assumption that the ratios of each component weight above the 2.38 millimeter screen to the total weight of the sample above 2.38 millimeters is the same as the ratio of the total component weight to the total sample weight. Seeds are usually found in the fractions between the 2.38 and 1.00 milli­ meters screens, and their weight is usually quite small. The seed weights are not included in the calculations but are tabulated in the final report of the flotation sample contents. For the sites where large quantities of seeds were recovered, the seeds are reported as a percentage of the total number of seeds per sample, otherwise only the number of seeds is recorded. Nevertheless, the raw data are reported in the appendices to this study.

60

KAMEDA JOMON PALEOETHNOBOTANY

Identifications of the remains were obtained by examination under a low-power (7X to 30X), binocular microscope. Fresh reference material collected by myself or received from the Tohoku Botanical Gardens seed collection was used for comparisons. Seed photographs in this text were taken using a Canon Fl, bellows attachment, a 50 millimeter macro lens reverse mounted on the bellows, and two electronic flash units. The film used is Hand W Control VTE Panchromatic film developed in the Hand W Company's 4.5 developer.

Chapter 4 HAKODATE AIRPORT SITE AND YAGI SITE PLANT REMAINS

Several archaeological subphases are found at the Hakodate Airport localities and the Yagi site. Two Initial lomon subphases are the earliest in the area tested for plant remains-Sumiyoshi-cho I at the Nakano B locality and Higashikushiro I at Yagi. Later occupations also occur at Hakodate Airport and Yagi. One, Todokawa "a" of the Early lomon, is represented both at Locality 4 of Hakodate Airport and Yagi (Table I). Some 4100 liters of soil from Yagi were processed while only 203 liters come from the two Hakodate Airport localities. The range of plant remains variation is rather small from'the latter localities, but the general composition of the remains is quite similar to those of similar age at Yagi. The paleoethnobotany of the early lomon occupations of the Kameda Peninsula seems to be adequately represented.

NAKANO B AND LOCALITY 4 Figure 4 shows the location of Nakano B and Locality 4 of Hakodate Airport. The land on which both sites are located is not well drained. The Uba River drains a narrow marsh which separates Locality 4 from the Nakano site. Within 3 kilometers of the sites are two rivers, the Matsukura and the Shiodomari rivers, which connect the coast with the interior of Kameda Peninsula. Both rivers have relatively wide flood­ plains. Most of the land around the sites as far as the lOa-meter contour line is under cultivation or is grassland. The only wooded areas are adjacent to the rivers and in the hills and mountains. Pollen from the

61

62

KAMEDA JOMON PALEOETHNOBOTANY

occupation levels suggest a prehistoric vegetation of buckeye, oak, and pine (Chiyo 1977). Pollen from herbaceous plants in the families Umbelli­ ferae and Compositae indicate the existence of unforested land. Nakano B Twenty-one houses were found at Nakano B but only one house was sampled for plant remains. The soil samples are from Grid 3 (Fig. 7), House 11 (Fig. 8) and were collected in 1975 by Eisuke Yokoyama, the excavation director that year. Two radiocarbon dates from House 11 are 7910 ± 80 B.P. and 7690 ± 125 B.P. A total of 125 liters of soil were processed from four levels: 1, 2, 3, and IIIb (Fig. 8). Level 2 was divided into two portions: one called "level 2," the other called "lower level 2." Although no specific information on the meaning of this distinction is available, "lower level 2" is probably the portion of level 2 immediately overlying level 3. "Level 2" is either general level 2 fill or the portion of level 2 adjacent to level 1. Level 3 is floor fill. All of the House 11 samples were processed by the garbage-can technique. The heavy fractions were found to have a significant propor­ tion of pottery sherds, stone flakes, small bone, and carbonized plant remains unlike the samples from Hamanasuno, Usujiri B, and Yagi. All of the sample listed in Table 4 was recovered from the heavy fractions except for some of the carbonized plant remains. About 62% of the plant remains is from the light fraction. Since none of the components except for the carbonized material were found in the light fractions from any of the other sites, the carbonized plant remains are summarized separately in Table 5 for comparative purposes. The carbonized items are comprised of wood charcoal, Japanese walnut shell, carbonized amur corktree berry flesh, unidentified plant remains, and 38 carbonized seeds. The seeds are comprised of only seven taxa, three of which are unknown. Fleshy fruit is the most abundant class of plant food represented; almost all of the seeds in this class are amur corktree. The single knotweed seed is the same as Type B seeds recovered from the other sites. Two of the unknown seeds are the same type and are not present in the samples from the other sites. A few patterns in the vertical distributions of plant remains are evident. Levels 2 and 3 have relatively high concentrations of plant remains, while levels 1 and lower level 2 have considerably lower concentrations. Seeds show the same pattern: higher densities are in levels 2 and 3 than in levels 1 and lower level 2. The distinction between the contents of level 3 and lower level 2 apparently confirms the distinction made in the field, but the

@

©;C8J

~

House 13

o Other Pits

NAKANO B, GRID 3

@

House 7

~

I

10m

o~

~O"":

Figure 7. Nakano B site, grid 3 (adapted from Chiyo 1977).

\

~

,~

U~

Q

0(~

~~

l(

) "',

~

V:J

~

~~

~

~

~~

~

~

~

~

§

~

~

~

~

~

64

KAMEDA JOMON PALEOETHNOBOTANY

U(

"

{

K

"

2m NAKANO B House 11

Figure 8. Nakano B site, House 11 (adapted from Chiyo 1977).

I

HAKODATE AND YAGI SITE PLANT REMAINS

65

Table 4. Nakano B, House 11, Flotation Samples: Contents as Percentage of Total Sample Weight and Light Fraction and Seed Densities Densities

Total Sample Weight (g)

Pottery

Stone Flakes

Small Bone

1 2 2 (lower) 3 IIIb2

40.48 362.09 32.16

21.5 61.9 33.5

55.4 28.3 46.5

a a

244.86 49.15

45.5 87.7

43.4 6.9

TOTAL

728.74

54.7

34.2

Level

Carbonized Plant Remains

Light Fraction (gjliter)

Seeds (no.jliter)

23.1 9.8 20.0

0.3 1.2 0.2

b 0.4 0.1

a

11.1 4.4

0.9 0.4

0.8 b

a

11.1

0.6

0.3

aLess than 1%. bless than 0.1 seeds/liter.

basis for the difference is unknown. No explanation for any of these patterns can be offered at the present time. Locality 4 Twenty-six flotation samples, totaling 78 liters, were processed from Locality 4. Seventeen of these samples were from House 1. The remainder was from Houses 10, 15,46, 47, and 52. The nine radiocarbon dates from Locality 4 range from 5430 ± 140 B.P. to 4720 ± 90 B.P.-three ofthe dates are from House 1 (4960 ± 110 B.P.), House 47 (4720 ± 90 B.P.), and House 52 (5100 ± 105 B.P.). The carbonized remains totaled 26.32 grams; 98.6% of this is wood charcoal. Walnut shell comprises the remaining 1.4%. The nutshell is all from Pit 2 in House 1. One knotweed seed, from House 1 Pit 3, was the only identifiable seed among the four seeds that were found. The flotation samples from Locality 4 were probably too small to have produced many plant remains, but the drop in percentage of walnut shell compared to Nakano B is probably significant. Interestingly, out of the ten tools found on the floor of House 1, seven were ishizara (metates). This is a considerably greater proportion than was found in House 11 at Nakano B. This is consistent with the hypothesis that the grinding stones were used for a purpose other than nut processing (Crawford et al. 1976).

YAGI SITE The Yagi site is situated on the coastal terrace between the Yagi and Osatsube rivers (Fig. 3). It is in the Osatsube community in Minamika-

80.61

aLess than 1%.

TOTAL

9.36 35.54 6.45 27.12 2.14

74.0

85.3 75.6 35.0 75.3 100.0

25.7

a

14.7 24.1 65.0 24.1

a

a

a(?) a

0.3

a a

0.3

a

a a a a 30

I 8 I 20

4

2

2

38

I II 2 24

Amur Total Weight (g) Corktree Unidentified Total of Carbonized Fruit Uniden- No. Wood Jug/ans Plant Amur Type B Plant Remains Charcoal Nutshell Remains Remains Seeds Corktree Dogwood Elderberry Knotweed Unknown tifiable Seeds

Table 5. Nakano B, House 11, Flotation Samples: Carbonized Plant Remains Components as Percentage of Total Weight of Carbonized Plant Remains and Number of Carbonized Seeds (Samples are Listed in Same Order as in Table 4)

0..:::

~

o ~

~

~

""3

~

~

~

~

~

~

~

&

HAKODATE AND YAGI SITE PLANT REMAINS

67

o

o

D VAGI SITE

o

0

o excavated orea H house

F feature contour interval 1 m

o

Figure 9. Vagi site.

10m

68

KAMEDA lOMON PALEOETHNOBOTANY

yabe at longitude 140° 14'E and latitude 41°53'N. The Vagi River has the broadest valley bottom of any ofthe rivers between Usujiri and Osatsube. This broad lowland extends only about 1.5 to 2 kilometers inland however. The course of the Vagi River is about 8 kilometers long; its source is at about 400 meters above sea level. The Osatsube River, less than a kilometer from the Vagi River at the Vagi Site, is slightly smaller and faster flowing, originating about 5 kilometers inland. The whole area of the terrace where the site is situated (Fig. 9) is under cultivation. Only the terrace slope and the hills backing the site vicinity are not cultivated; these areas are covered by secondary forest. The samples reported herein were collected as an integral part of the interdisciplinary Vagi project directed by William Hurley, Peter Bleed, and Masakazu Yoshizaki (Bleed et al. 1979; Hurley and Yoshizaki 1979). One of several purposes of this project is to describe the subsistence base of the lomon Vagi popUlation. The Vagi paleoethnobotanical data, in addition, directly relate to the goals of this broader Kameda Peninsula study. Analysis of other archaeological data from Vagi is still ongoing at the time of this writing, so issues specific to problems depending on integrating much of the Vagi information can be examined only tenta­ tively. The plant remains data from Vagi, as well as from Hamanasuno and Usujiri, clarify several subsistence-ecological related issues in spite of minimal data on stone and ceramic technology. Samples were taken from a variety of contexts. The contexts include deposits associated with six houses (1, la, 2, 3, 5, 6, and 7) and five external house features (101, 1008, 1010, 1017, and 1043), as well as deposits that are general sheet midden. Figure 9 illustrates the general location of relevant areas at Vagi. The chronology of the Vagi occupations represents a longer time period than that represented at any of the other sites reported in this investi­ gation. The Nakano B and Locality 4 localities of Hakodate Airport, the / Hamanasuno and the Usujiri B occupations represent sites classified as Initial, early Early, late Early, and Late Middle lomon, respectively. Vagi has Initial, early to late Early and late lomon occupations in evidence. Interpretation of the plant remains from Vagi must take this long time span into account. The midden and associated features in the northernmost portion of the site contained some Higashikushiro III (described by Hurley and Yoshizaki 1979) and mainly Ento-Kaso A (Todokawa) pottery-late Initial to early Early lomon. Radiocarbon dates from Feature WI are: level 2, 5180 ± 110 B.P. (Gak 8988); level 3a, 5770 ± 150 B.P. (Gak 8982) and 6650 ± 170 B.P. (Gak 8983). Feature 101 contained Ento-Kaso A pottery (Hurley, personal communication). This area of Vagi represents

HAKODATE AND YAGI SITE PLANT REMAINS

69

the same lomon subphase found at Hakodate Airport Locality 4, al­ though two of the radiocarbon da tes from Feature 101 are earlier than the dates from Locality 4. Houses 3 and 5, and perhaps 6 and 7, have some temporal overlap with Hamanasuno which was occupied from 5500 to 4500 B.P. (Hurley et al. 1976). The pottery from these contexts is primarily Saibesawa II, with some Saibesawa III. The upper fill of House 5, in addition to House 4, contained Late lomon pottery, so evidence for some fourth millennium B.P. activity exists at Yagi. Two radiocarbon dates from post holes in House 5 are 6290 ± 150 B.P. (Gak 8981) and 6520 ± 180 B.P. (Gak 8984). One sample from fill above the House 5 floor is 5170 ± 160 B.P. (Gak 8986). Radiocarbon dates from House 3 are not so consistent. A radio­ carbon date from level X2 is 4870 ± 140 B.P. (Gak 8987) or final Early lomon. A radiocarbon date on another fill sample is 5530 ± 150 B.P. (Gak 8990), which is comparable with material dated to 5550 ± 180 B.P. (Gak 8985) from the House 3 floor. However, three other floor radiocarbon dates are rather late: 4410 ± 140 B.P. (Gak 7588), 3790 ± 140 B.P. (Gak 7593), and 2590 ± 120 B.P. (Gak 8989). Some contamination by later charcoal may be evident, or some of the dates may be simply in error. The House 2 samples have an ambiguous cultural and temporal con­ text. The radiocarbon dates cluster in the fifth millennium B.P. (Bleed et al. 1979), but pottery from the pit fill is Initial lomon Higashikushiro I and Kojohama (Hurley and Yoshizaki 1979:56-57). One Early lomon sherd also came from this fill (ibid.:59). During the three field seasons at Yagi, about 7700 liters of soil were processed with the froth flotation machine. Analysis has been completed for residues from about 53% (4102 liters) of the processed soil (52 sam­ ples). Heavy fractions were not collected as a rule, but when screening the soil to prepare it for flotation indicated concentrations of potentially non­ flotable material, heavy fractions were obtained. Only three samples were chosen for this procedure. The plant remains data are summarized in Table 6 and Figure 10. The 1134 grams of light fraction are almost entirely wood charcoal (Fig. 10). The heavy fractions collected from three samples and nutshell collected by hand change this perspective somewhat. The three types of remains, including stone flakes and bone, total 2148.75 grams (32.0% flakes, 10.3% bone, 54.4% wood charcoal, 0.3% Sasa, and 2.9% walnut). Walnut is 5.0% of the total carbonized plant remains by weight. These statistics characterize the total carbonized remains from Yagi, but Yagi is not a single component site. Careful consideration of the archaeological context must be given in order to begin to understand the meaning of these data.

16

N81E98 level 4

N79E99 fill

N79E98 fill

2

2

2

nil

2

N76EIOO level 5

2

15

4

N76EIOO level 2

N76EIOO level 4

II

N84EIOO fill

49

24

(g)

level XI

N90E98 Fea. 3

Context

Total No. Seeds

2

la

House No.

Total Seed Weight

~

v

0.

N

~

>,

" I-

0.

...v

..c

,

;0

"~

-0 0 0. 0 0:

E:> 0.

Greens Knotweeds

~

~

~

0

~

~l(?)

~

~ 0"

:z ... "0.~

I-~ ~

• •

• • • • •

r?J IJ' '->"'1 :$~ ~

~~ .~ f2'

.:s

.~l!i

>-..~

rt>!

Sample Components (%)

Figure 23. Usujiri B site, Daigi VIIlb component, flotation samples: light fraction and seed densities, and sample components as percentage of total sample weight.

20

18

~

;::,

'*

~.

/

Sample Provenience

~

~

t.;

~

c"

~

~~

::0

~

..,

~

~ ~

tx:l c" .....,

;;.....,

S5

'less than 0.05 g. +less than 0.5%.

TOTAL

X3

Level X3 Hearth Total

18

18 20

Context

House

0.26

0.26

Total Seed Weight (g)

625

148 477

10

138

Total No. Seeds

19.5

0.7 25.4

0.7

~

Z :I:

C/l

f-

0:: ~ 0 0::

..!l

.......

00

~

t-