Indicators and Surrogates of Biodiversity and Environmental Change [1 ed.] 9781486304103, 9781486304097

Ecological indicators and surrogates are used widely by resource managers to monitor and understand complex biota and ec

159 92 14MB

English Pages 216 Year 2015

Report DMCA / Copyright

DOWNLOAD PDF FILE

Recommend Papers

Indicators and Surrogates of Biodiversity and Environmental Change [1 ed.]
 9781486304103, 9781486304097

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Indicators and Surrogates of Biodiversity and Environmental Change provides insights into the use of indicators and surrogates in natural resource management and conservation – where to use them, where not to use them, and how to use them. Using an ecological approach, the chapters explore the development, application and efficacy of indicators and surrogates in terrestrial, aquatic, marine and atmospheric environments. The authors identify current gaps in knowledge and articulate the future directions for research needed to close those gaps. This book is written by the world’s leading thinkers in the area of indicators and surrogates. It is the first major synthesis of learnings about indicators and surrogates and will be a critical resource for the vast number of people developing and applying them in ecosystems around the world.

About the editors David Lindenmayer is a Research Professor at the Fenner School of Environment and Society at The Australian National University. He leads a range of large-scale, long-term field-based programs examining the impacts of land use practices such as forestry, plantation management and agricultural development on biodiversity. Philip Barton is a Research Fellow at the Fenner School of Environment and Society at The Australian National University. He is a community ecologist with a particular interest in insects. Philip is currently working on theoretical and empirical assessments of biodiversity surrogates.

Indicators and Surrogates of Biodiversity and Environmental Change

David Lindenmayer, Philip Barton and Jennifer Pierson

Jennifer Pierson is a Research Fellow at the Fenner School of Environment and Society at The Australian National University. She is a population ecologist interested in the interacting contributions of behaviour, demography and genetics to population dynamics.

Indicators and Surrogates of Biodiversity and Environmental Change

Ecological indicators and surrogates are used widely by resource managers to monitor and understand complex biota and ecosystem processes. Their potential to guide complex resource management has meant they have been proposed for use in all ecosystems worldwide. Despite extensive research into indicators and surrogates, there remains much controversy about their use, in addition to major issues and knowledge gaps associated with their identification, testing and application.

Editors: David Lindenmayer, Philip Barton and Jennifer Pierson

Indicators and Surrogates of Biodiversity and Environmental Change Editors: David Lindenmayer, Philip Barton and Jennifer Pierson

This page intentionally left blank

‡7HQ&RPPLWPHQWVSSLQGG

$0

Indicators and Surrogates of Biodiversity and Environmental Change

Editors: David Lindenmayer, Philip Barton and Jennifer Pierson

© Professor David Lindenmayer, Dr Philip Barton and Dr Jennifer Pierson 2015 All rights reserved. Except under the conditions described in the Australian Copyright Act 1968 and subsequent amendments, no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, duplicating or otherwise, without the prior permission of the copyright owner. Contact CSIRO Publishing for all permission requests. National Library of Australia Cataloguing-inPublication entry Indicators and surrogates of biodiversity and environmental change/editors: David Lindenmayer, Philip Barton and Jennifer Pierson. 9781486304097 (paperback) 9781486304103 (epdf) 9781486304110 (epub) Includes bibliographical references and index. Environmental indicators. Environmental monitoring. Environmental management. Biodiversity conservation. Lindenmayer, David, editor. Barton, Philip, editor. Pierson, Jennifer, editor. 363.7063 Published exclusively in Australia and New Zealand by CSIRO Publishing Locked Bag 10 Clayton South VIC 3169 Australia Telephone: +61 3 9545 8400 Email: [email protected] Website: www.publish.csiro.au

Published exclusively throughout the world (excluding Australia and New Zealand) by CRC Press, with ISBN 978-1-4987-4870-4 CRC Press 6000 Broken Sound Parkway, NW Suite 300, Boca Raton, FL 33487 Front cover (clockwise from top left): Mushrooms on bark (Martin Westgate); Blue groper (Achoerodus viridis) (Maria Beger); Cicada (Cicadidae) (Jennifer Pierson); Spotted salamander (Ambystoma maculatum) (Kristine Hoffmann) Set in 10.5/12 Adobe Minion Pro and ITC Stone Sans Edited by Peter Storer Editorial Services Cover design by Andrew Weatherill Typeset by Desktop Concepts Pty Ltd, Melbourne Index by Bruce Gillespie Printed in China by 1010 Printing International Ltd CSIRO Publishing publishes and distributes scientific, technical and health science books, magazines and journals from Australia to a worldwide audience and conducts these activities autonomously from the research activities of the Commonwealth Scientific and Industrial Research Organisation (CSIRO). The views expressed in this publication are those of the author(s) and do not necessarily represent those of, and should not be attributed to, the publisher or CSIRO. The copyright owner shall not be liable for technical or other errors or omissions contained herein. The reader/user accepts all risks and responsibility for losses, damages, costs and other consequences resulting directly or indirectly from using this information. Original print edition: The paper this book is printed on is in accordance with the rules of the Forest Stewardship Council ®. The FSC® promotes environmentally responsible, socially beneficial and economically viable management of the world’s forests.

Contents

Acknowledgements vii Author biographies

1 Introduction – disciplinary and multi-disciplinary perspectives on ecological indicators and surrogates

ix

1

David Lindenmayer, Jennifer Pierson and Philip Barton

2 Surrogates for the distribution and trajectory of biodiversity

5

Martin Westgate

3 Biodiversity surrogates

15

David Lindenmayer, Philip Barton, Martin Westgate, Peter Lane and Jennifer Pierson

4 Conservation by proxy: thoughts 5 years on

25

Tim Caro

5 Avian surrogates in terrestrial ecosystems: theory and practice

33

Kathy Martin, José Tomás Ibarra and Mark Drever

6 Using decision theory to select indicators for managing threats to biodiversity

45

Ayesha Tulloch

7 Invertebrate indicators and ecosystem restoration

59

Philip Barton and Melinda Moir

8 Mosses as passive and active indicator surrogates for investigations of atmospheric pollution and quality

69

Hanna Salo

9 Lichens as ecological indicators to track atmospheric changes: future challenges

77

Cristina Branquinho, Paula Matos and Pedro Pinho

v

vi

Indicators and Surrogates of Biodiversity and Environmental Change

10 Approaches, potential and pitfalls of applying bioindicators in freshwater ecosystems

91

Jani Heino

11 Searching for the holy grail of wetland integrity: are surrogates still relevant in conservation planning?

101

Hamish Greig and Aram Calhoun

12 Surrogates for coral reef ecosystem health and evaluating management success

113

Maria Beger

13 Abiotic surrogates in support of marine biodiversity conservation 125 Camille Mellin

14 Building indicators for coupled marine socio-ecological systems

137

Catherine Longo and Benjamin Halpern

15 The application of genetic indicators in wild populations: potential and pitfalls for genetic monitoring

149

Jennifer Pierson, Gordon Luikart and Michael Schwartz

16 Use of surrogates in medicine: ideas that may be useful for ecology 161 Peter Lane and Philip Barton

17 Application of surrogates and indicators to monitoring natural resources 169 John Gross and Barry Noon

18 Indicators and surrogates in environmental management

179

William H. McDowell

19 A diversity of approaches to ecological surrogates and key knowledge gaps

189

David Lindenmayer, Jennifer Pierson, Philip Barton, Peter Lane, Ayesha Tulloch and Martin Westgate Index 195

Acknowledgements

We thank Tabitha Boyer and Claire Shepherd for many aspects of book production as well as organising the workshop that instigated the writing of all of the chapters that feature in this book. We thank Mac Hunter for contributing to the workshop and assisting with the synthesis of ideas. We most gratefully acknowledge the support of John Manger from CSIRO Publishing in encouraging this book to be written. A number of the chapters in this volume feature scientists from the Fenner School of Environment and Society at The Australian National University. These scholars are supported through an Australian Research Council grant (to DBL). We are indebted to that organisation for its support. Finally, the editors thank all chapter authors for the rapid responses to constant badgering to deliver timely but high quality written material. David Lindenmayer Jennifer Pierson Philip Barton (The Editors) December 2014

vii

This page intentionally left blank

‡7HQ&RPPLWPHQWVSSLQGG

$0

Author biographies

Philip Barton is a Research Fellow in the Fenner School of Environment and Society, ANU. He is a community ecologist with a particular interest in insects. Philip is currently working on theoretical and empirical assessments of biodiversity surrogates. Maria Beger is a Research Fellow in the Centre for Biodiversity and Conservation Science at the University of Queensland, Australia. Her research employs interdisciplinary empirical and modelling approaches to  aid decisions about managing marine ecosystems.  She works at better incorporating bio-physical and ecological processes  into decision science  that challenge effective management in the face of climate change, an increasing human footprint and transforming ecosystems. Cristina Branquinho is an Associate Researcher at the Universidade de Lisboa. She is an ecologist with particular interest in lichens and plants. Cristina is currently interested in ecological indicators of nitrogen pollution and of climate change based on functional diversity. Aram Calhoun is a wetland ecologist in the Department of Wildlife, Fisheries, and Conservation Biology at the University of Maine, USA. Her research focuses on forested wetlands and vernal pool ecosystems. She is particularly interested in collaborative approaches to conserving natural resources on private lands. Dr Calhoun is active in working at all levels of government (local, state, and federal) on wetland policy and conservation issues. Tim Caro is a Professor of Wildlife Biology at the University of California at Davis whose research focuses on behavioural ecology and conservation biology, particularly strategies to conserve wildlife in tropical ecosystems. He runs a long term field site in Katavi National Park western Tanzania. He recently published a book entitled Conservation by Proxy: Indicator, Umbrella, Keystone, Flagship, and other Surrogate Species in which he tried to bring order to this area of conservation. His current conservation work in Africa includes identifying wildlife corridors and understanding threats to protected areas. Mark Drever is a population ecologist with the Canadian Wildlife Service of Environment Canada and Adjunct Professor with the University of British Columbia. He conducts research on the conservation of migratory birds, focusing on shorebirds and waterbirds. Hamish Greig is a freshwater ecologist in the School of Biology and Ecology at the University of Maine, USA. His research focusses on communities of invertebrates and fish in stream, river, lake and wetland ecosystems. He is particularly interested in the impacts of ix

x

Indicators and Surrogates of Biodiversity and Environmental Change

disturbance on freshwater communities, and is currently working on understanding and measuring the impacts of climate change and catchment modification on freshwaters. John Gross is an ecologist with the US National Park Service. He was with the NPS Inventory & Monitoring Program during the development and implementation of natural resource monitoring across nearly 300 park units. John’s background is in ecological modelling and his current focus is on climate change adaptation in protected areas. Ben Halpern is a Professor in the Bren School of Environmental Science & Management at UC Santa Barbara and Chair in Marine Conservation at Imperial College London. He focuses his research at the interface between marine ecology and conservation planning. He has led and participated in several key synthetic research projects that have advanced our understanding of the state of the world’s oceans and the potential for marine reserves to improve ocean condition. In particular he has led the development and mapping of cumulative impact assessments at global and regional scales in marine and freshwater systems and has been the lead scientist for the Ocean Health Index project. Jani Heino is a Research Professor at the Finnish Environment Institute. He is an ecologist interested in biodiversity, mechanisms underlying its origins, maintenance and variation, and its conservation and management in the face of environmental changes. His current studies focus mainly on streams, rivers, ponds and lakes. José Tomás Ibarra is a Postdoctoral Researcher in the Department of Forest and Conservation Sciences at the University of British Columbia, Canada. He is a community ecologist with broad interests in natural history and the influence of multiple factors (e.g. habitat, biotic interactions and land-use changes) in shaping the current state of taxonomic and functional animal biodiversity. During his PhD, José Tomás studied the ecology of poorly known owls in Andean temperate forests, and validated forest-specialist owls as surrogates for both taxonomic and functional biodiversity in this globally threatened eco-region. Peter Lane is a Statistical Consultant in the Fenner School of Environment and Society, ANU. He is a statistician, working in David Lindenmayer's team on surrogacy. Previously, he  worked for 13 years  as a research and consultant statistician in the pharmaceutical industry (GlaxoSmithKline), and for 25 years in agricultural research at Rothamsted in the UK. David Lindenmayer is a Research Professor at the Fenner School of Environment and Society, ANU. He leads a range of large-scale, long-term field-based programs examining the impacts of land-use practices such as forestry, plantation management and agricultural development on biodiversity. Catherine Longo is a project scientist at the National Center for Ecological Analysis and Synthesis (NCEAS, USA). Her expertise is in indicators for coupled socio-ecological marine systems (www.oceanhealthindex.org), and her fields of interest are fisheries, biodiversity, metabolic theory, data-poor assessments and evaluating uncertainty. Gordon Luikart is a professor of Conservation Genetics at the Flathead Lake Biological Station and University of Montana. He has conducted research on the genetics and conser-

Author biographies

vation of fish and wildlife for over 15 years in the USA, France, and Portugal. His primary focus has been on the development and application of molecular genetic and computational tools for assessing the genetic basis of fitness, and monitoring effective population size and landscape genetic connectivity. Kathy Martin is a Professor in the Department of Forest and Conservation Sciences at the University of British Columbia and a Senior Research Scientist with Environment Canada, Vancouver Canada. She is a population and community ecologist with a particular interest in responses of birds to environmental and habitat disturbance in forest and alpine ecosystems. Kathy is currently working on theoretical and empirical assessments of biodiversity surrogates and indicators for cavity-nesting vertebrates experiencing insect outbreaks and forest harvesting, as well as the responses of mountain birds to climate change impacts in the Americas. Paula Matos is a PhD Candidate at the Universidade de Lisboa and the Universidade de Aveiro, Portugal. She is an ecologist with particular interest in lichens. Paula is currently interested in ecological indicators of climate change and air pollution based on lichen functional diversity. William H. McDowell is a Professor of Environmental Science in the Department of Natural Resources and the Environment, University of New Hampshire, USA. He is an ecosystem ecologist and biogeochemist with a particular interest in understanding the role of dissolved organic matter in aquatic and terrestrial ecosystems. He is currently working on several long-term projects to understand land–water connections in forested watersheds. He has recently begun to deploy real-time water-quality sensors in order to understand drivers of water quality at time scales from minutes to decades. Camille Mellin is an ARC Discovery Early Career Research Award fellow at the Australian Institute of Marine Science (AIMS). Camille is a quantitative ecologist with a background in marine and coral reef ecology. She is currently working on the predictive modelling of Acanthaster planci (crown-of-thorns starfish) outbreaks, coral reef resilience to disturbance and biodiversity patterns. Melinda Moir is an Assistant Professor at the School of Plant Biology at the University of Western Australia. She is a research entomologist with projects on conservation biology, extinction risk, refugia in climate change, traits across trophic groups in ecological restoration, and invasion ecology. Barry Noon is a Professor in the Department of Fish, Wildlife, and Conservation Biology at the Colorado State University. He has conducted research on the effects of land management practices on wildlife populations in the USA and internationally for more than 40 years. His primary research focus has been on the population dynamics and conservation of imperilled species in terrestrial ecosystems. Jenny Pierson is a Research Fellow in the Fenner School of Environment and Society, ANU. She is a population ecologist interested in the interacting contributions of behaviour, demography and genetics to population dynamics. She is currently working on empirical assessments of biodiversity surrogates, with a focus on genetic indicators and habitat-based surrogates.

xi

xii

Indicators and Surrogates of Biodiversity and Environmental Change

Pedro Pinho is a Post-doctoral Fellow at the Universidade de Lisboa working on the use of vegetation functional groups and spatial patterns to assess the effect of environmental changes, with both natural and human origin: desertification and land degradation, climate change, eutrophication and land-use/cover. Hanna Salo is currently a PhD candidate in the Department of Geography and Geology, University of Turku, Finland. She is a geographer focusing on environmental research, with the main emphasis on air quality issues. At present, Hanna is working on magnetic biomonitoring of atmospheric pollution and especially heavy metals. Michael Schwartz is the Director of the National Genomics Center for Wildlife and Fish Conservation located in Missoula Montana. He is a Presidential Early Career Award in Science and Engineering recipient and an adjunct faculty member of the University of Montana’s College of Forestry and Conservation. Dr Schwartz’s research team focuses on the fields of population, conservation and landscape genetics. He seeks to provide practical answers to natural resource problems, combining field and laboratory work. Ayesha Tulloch is a Research Fellow at the Fenner School of Environment and Society, ANU. She is interested in solving problems of resource allocation to monitoring and management of biodiversity. She is currently working on applications of surrogates in adaptive management of threatened birds, and exploring how variability in species associations can affect decisions such as optimal indicator selection that rely on surrogacy data for informing monitoring choices. Martin Westgate is a Research Fellow in the Fenner School of Environment and Society, ANU. He is an ecologist interested in how ecosystems respond to  natural and humaninduced change, and is currently working on biodiversity surrogates and meta-analyses of cross-taxon congruence.

1

Introduction – disciplinary and multi-disciplinary perspectives on ecological indicators and surrogates David Lindenmayer, Jennifer Pierson and Philip Barton

Introduction The term surrogate can be broadly defined as a proxy for something else. The scientific study of ecological surrogacy, and its application to biodiversity and resource management, is widespread – more than 5000 scientific articles have been published on the topic (Westgate et al. 2014). The reason for this vast literature is the natural environment is so complex that it is neither logistically practical nor financially possible to directly measure and monitor all components of the environment in all locations and at all times. Surrogates or proxies are therefore needed to represent other attributes of the environment that are not able to be directly measured (Caro 2010; Collen and Nicholson 2014). Indeed, the application of ecological surrogates lies at the heart of many key areas of environmental science, ranging from the selection and monitoring of reserves, monitoring of pollutants in the atmosphere or drinking water, the effectiveness of vegetation restoration, and estimates of the number of the total number of plant and animal species in a given location. The primary focus of this edited volume is on ecological surrogates – although, as we discuss below, our book also includes a chapter on the use of surrogates in medicine and insights on surrogacy from statistical science. However, before we outline some of the content of this book, the reason why it was written and the structure used to guide each chapter, it is important to first define at the outset what we mean by the term ‘surrogacy’, and other terms that fall within its broader remit (in particular the term ‘indicator’). We also highlight the important distinction between surrogacy and direct measurement (sensu Lindenmayer and Likens 2011). For the purposes of this book, we deem the term surrogate to mean a proxy for some other entity. The primary focus of this book is on ecological surrogates; that is, the use of proxies in the environment, ecology and conservation. Direct measurement in the context of this book means measuring a given entity without implying that it is a proxy for some other entity (Lindenmayer and Likens 2011). For example, the concentration of E. coli in an aquatic ecosystem might be measured simply to determine how much of this microorganism occurs in a given water body. Similarly, the concentration of carbon dioxide in the 1

2

Indicators and Surrogates of Biodiversity and Environmental Change

atmosphere (e.g. the Keeling curve) may simply be a direct measurement of how much of it now occurs in the air above a given area. These direct measurements of E. coli and carbon dioxide do not have to imply anything about the condition of the airshed or of the water. However, if the amount of E. coli is considered to a proxy for the quality of water fit for human consumption or the concentration of carbon dioxide as a proxy for future increases in temperature, then they become ecological surrogates – that is, a proxy for some other entity (water quality or future temperature). Why is there a need to define these terms? The reason is that many terms in the relevant literature (including the term surrogate itself) have been used in quite different ways and to mean quite different things. This is evident even in this book. Hence, we have asked the author/s of each chapter to specifically define key terms within the text to ensure that readers understand the particular usage.

The aim of this book There have been many detailed reviews of ecological surrogacy, and the narrower field of species-based indicators, in recent years (e.g. McGeoch 1998; Niemi et al. 1997; Rodrigues and Brooks 2007; Caro 2010). The purpose of this book was not to cover the same ground as these previous reviews. Rather, it was to highlight the diverse ways that surrogacy is studied and applied in the broader environmental sciences, including terrestrial, marine, aquatic, atmospheric and policy realms. Our aim was then to draw together some of the key learnings from these different disciplines and compare and contrast the different ways the study and application of surrogates have evolved within them. In addition, we sought to explore the overlap and divergences between disciplines in the use of ecological surrogates with an explicit aim of trying to uncover ways to improve the application of surrogates in the future.

The structure of this book and the chapters As part of compiling this book, we asked chapter authors not to re-hash what they have written previously, but rather to expound their personal views on their important area of expertise. In particular, authors we asked to arrange their chapters around 10 points that cover things they have learnt or are known, and remaining challenges or unknowns in their field. The focus was largely on practical (realistic) applications of indicators and surrogates in environmental management and/or conservation management. Each chapter covers discipline-specific areas, and many use terminology developed and used by researchers in that field. Acronyms have been avoided where possible, and definitions provided where necessary. Each chapter in this book has been kept deliberately short and to the point. This was done for good reason – the information ‘super-glut’ of this era means that there is an overload of information that few people have time to touch on, evenly briefly. Our aim was therefore to request chapter authors to produce short pithy chapters that were readily accessible and with the key points summarised in a way that could be quickly comprehended. The challenge to all authors was to summarise their ideas in a few thousand words – and shorter if possible. To this end, each chapter is structured in a similar way. That is: • A box with the 10 dot points – one for each key issue that they considered. • A few short introductory paragraphs with limited background on their topic.

1: Introduction

• Approximately five things that have been learnt or are known, and five remaining challenges or unknowns in their field, with one or two paragraphs on each of the 10 key issues. • A paragraph on each issue laying out why it is essential that it must be tackled. • A small set of references directing readers to additional literature that further explores the topic. This book is broken into six main themes. Theme 1 contains four chapters and is a general overview of ecological indicators and surrogates, including global perspectives on their use. Three chapters comprise theme 2, which is focused on ecological indicators and surrogates in the terrestrial environment. Theme 3 also comprises two chapters and both are associated with the application of indicators and surrogates in the measurement and management of the atmosphere. Aquatic ecosystems are the core of the theme 4, with two chapters featured. Theme 5 comprises three chapters on marine ecosystems. Theme 6 is particularly fascinating as it contains two quite different kinds of chapters not normally associated with the field of ecological indicators and surrogates. The first chapter in this theme concerns the increasing use of genetic metrics and explores the exciting potential of this relatively new area in the ecological surrogate domain. The second chapter in Theme 6 pivots around discussions of the learnings for the use of ecological indicators and surrogates in ecology that can be drawn from medical science and statistical science. Theme 7 comprises two chapters about indicators and surrogates in policy. The final theme 8 contains just one chapter – an overview of learnings that come from the collected insights of the chapter authors. In particular, what works well and what doesn’t, and what common ideas can be used to guide future research on the selection and application of ecological surrogates. This chapter was developed from the diversity of author experience and chapter content, especially through a workshop held in south-east Queensland (Australia) in October 2014.

Caveats and notes As with any book comprising many topics written by different authors, there are some inconsistencies in the approach and flavour of each chapter and the volume editors decided that these could remain. The authors of each chapter were responsible for their own contribution and only limited content editing was conducted. Of course we were acutely aware that there are sometimes more than 10 key issues that could be considered in each sector, ecosystem and cross-cutting theme. However, we felt that asking each author team to highlight their ‘top 10’ was a good way to focus their writing around important insights and key knowledge gaps that need to be filled. A key concern we had at the commencement of this book writing project was the risk of extensive overlap between chapters. In reality, there proved to be an impressive diversity of topics, important synergies and limited overlap between chapters. We considered the little overlap that did occur was tolerable because each chapter needed to be a stand-alone contribution given the way many readers might ‘dip into and out of’ the book. Ecological surrogates are likely to become even more important in guiding environmental policy and management in future. Our sincere hope is that this book represents a leap forward in the understanding and application of ecological surrogacy in terms of breadth of expertise and examples, as well as the identification of core learnings that transcend individual disciplines.

3

4

Indicators and Surrogates of Biodiversity and Environmental Change

References Caro T (2010) Conservation by Proxy: Indicator, Umbrella, Keystone, Flagship, and Other Surrogate Species. Island Press, Washington DC. Collen B, Nicholson E (2014) Taking the measure of change. Science 346, 166–167. Lindenmayer DB, Likens GE (2011) Direct measurement versus surrogate indicator species for evaluating environmental change and biodiversity loss. Ecosystems 14, 47–59. doi:10.1007/ s10021-010-9394-6. McGeoch MA (1998) The selection, testing and application of terrestrial insects as bioindicators. Biological Reviews of the Cambridge Philosophical Society 73, 181–201. doi:10.1017/S000632319700515X. Niemi GJ, Hanowski JM, Lima AR, Nicholls T, Weiland N (1997) A critical analysis on the use of indicator species in management. The Journal of Wildlife Management 61, 1240–1252. doi:10.2307/3802123. Rodrigues AS, Brooks TM (2007) Shortcuts for biodiversity conservation planning: the effectiveness of surrogates. Annual Review of Ecology Evolution and Systematics 38, 713–737. doi:10.1146/annurev.ecolsys.38.091206.095737. Westgate MJ, Barton PS, Lane PW, Lindenmayer DB (2014) Global meta-analysis reveals low consistency of biodiversity congruence relationships. Nature Communications 5, 3899. doi:10.1038/ncomms4899.

2

Surrogates for the distribution and trajectory of biodiversity Martin Westgate

Things we know 1 2 3 4 5

Most taxa show congruent distribution patterns. Processes that influence congruence are scale-dependent. Surrogate research is shifting towards species-level inference. Functional ecology can help to identify biodiversity surrogates. Surrogates should measure progress towards management goals.

Knowledge gaps 6 A long-term view of biodiversity surrogates. 7 Greater integration is needed across taxonomic boundaries. 8 Research on the mechanisms underpinning surrogate relationships is not always worthwhile. 9 Complementary surrogates. 10 Adaptive surrogates.

Introduction Biodiversity is a broad concept that encompasses several real and perceived values of ecosystems. However, the idea that science can help to conserve biodiversity relies on the assumption that biodiversity can be measured. Unfortunately, ecosystems are too complex to allow comprehensive mapping of their specific or genetic diversity: only a small proportion of all species have been taxonomically described, while data on species’ distributions or risks of extinction are rarer still (Scheffers et al. 2012). Consequently, our ability to measure ‘total’ biodiversity is dependent on surrogates: practical metrics that allow us to identify locations or times where we find high numbers and/or distinct combinations of species. In practice, most assessments of the distribution or trajectory of biodiversity rely on information from a small number of frequently studied taxa, typically with an emphasis on vertebrate animals and vascular plants (Westgate et al. 2014). In this chapter, I will refer to such metrics (i.e. those whose goal is the measurement of biodiversity) under the umbrella term ‘biodiversity surrogates’. My aim is to explore the current state of the s­ cience 5

6

Indicators and Surrogates of Biodiversity and Environmental Change

of biodiversity measurement, discuss how a reliance on surrogates influences conservation decisions, and suggest future directions for this important area of ecological research.

Things we know 1.  Most taxa show congruent distribution patterns A major component of biodiversity surrogate research involves the assessment of ‘congruence’, which measures the extent to which information on one group of species provides information on a second group. Perhaps the most common application of this approach is to test the relationship between the spatial distributions of two distinct taxonomic groups (known as ‘cross-taxon’ congruence). The success of this approach, and its utility for informing conservation, varies strongly depending on a range of factors. In particular, differences in how target and surrogate groups are defined and measured can strongly influence patterns of congruence (Rodrigues and Brooks 2007), as can attributes of study design such as spatial scale or grain size (Westgate et al. 2014). Further, there is debate regarding the extent to which different metrics for quantifying and comparing the diversity of two groups provide useful information (see point 3 below). Therefore, caution must be taken when using congruence-based approaches to infer patterns of biodiversity. Despite these caveats, it is encouraging that the degree of congruence between taxa is typically greater than zero, irrespective of the metric under investigation (Rodrigues and Brooks 2007; Westgate et al. 2014). Pointing out that congruence is a common attribute in nature is important because the academic literature is often highly critical of the surrogate concept. Positive correlations occur between taxa because all life responds to similar processes, particularly at large spatial and temporal scales. For example, studies of the influence of biogeographic parameters such temperature, latitude or elevation on taxonomic diversity have a long history in ecology, and their respective influence on different taxa is relatively well understood (e.g. Qian and Kissling 2010). This suggests that progress can be made towards understanding patterns of congruence between different sets of species or locations. 2.  Processes that influence congruence are scale-dependent Understanding the circumstances where we would expect to observe biodiversity congruence would greatly assist conservation decision making. However, the number or diversity of species at a given location is influenced by many processes, each acting at particular spatial and temporal scales. For example, we still lack an overarching theory that describes how evolution, species traits and species interactions combine to determine biodiversity at macro scales (Nuismer and Harmon 2014). Moreover, much research on surrogates in ecology and conservation focuses on how biodiversity responds to landscape-scale processes such as disturbance or fragmentation, which are areas where cross-taxon synthesis has traditionally been difficult (Ewers et al. 2010). These difficulties may explain why many applications of the surrogacy concept lack a theoretical foundation, or are inconsistent with known theoretical concepts (Sætersdal and Gjerde 2011). Some authors have attempted to identify processes that influence cross-taxon congruence across a range of spatial scales. In particular, groups of species that display strong interactions such as predator–prey dynamics or mutualisms should be highly congruent, but this does not always occur (Dehling et al. 2014). Alternatively, species may share physiological attributes that ensure they display similar responses to climate or elevation, leading to high congruence (Hawkins and Porter 2003). Without a better knowledge of how

2: Surrogates for the distribution and trajectory of biodiversity

these processes act and interact, however, we remain unable to predict when we should expect sets of taxa to display high congruence.

3.  Surrogate research is shifting towards species-level inference Early attempts to identify biodiversity surrogates typically focused on correlations in species richness and/or diversity across taxa, but prioritising locations with high species richness for conservation ignored areas with highest threat, or those with a high proportion of endemic species (Orme et al. 2005). Further, many taxa display changes in species composition across ecological gradients, but not species richness (Supp and Ernest 2014). Consequently, methods that use diversity indices to assess congruence have been largely superseded by optimisation of species occurrences as the standard method for spatial prioritisation (Williams et al. 2006). Similarly, hierarchical models are now capable of assessing change in assemblage structure via logistic regression of individual species occurrences (Wang et al. 2012), rather than relying on aggregate metrics. These tools have provided a range of new insights into the distribution and trajectory of biodiversity. This increased focus on species-level information, while encouraging, brings new challenges. First, species occurrence data are highly scale-dependent, with issues resulting from spatial variation in species occupancy, abundance and detectability increasing at fine scales (Hurlbert and Jetz 2007). Second, different software algorithms, or changes to the underlying assumptions of the analysis, can lead to enormous differences in the ranking of locations or actions for conservation (Grantham et al. 2010). Finally, applications of spatial prioritisation can be more sensitive to the economic value of study units than to their biological composition (Bode et al. 2008): a property that is useful for efficient conservation, but risks promoting the lesson that biodiversity itself is of secondary concern. Consequently, care is needed when relying intensively on species-level datasets for conservation decision-making. 4.  Functional ecology can help to identify biodiversity surrogates The functional approach to ecological research is based on the premise that knowledge of species traits can be used to understand ecosystem structure. This concept is widely accepted and studied, both because of its intuitive appeal (species differ from each other in meaningful ways, and traits reflect that), and its usefulness for describing patterns in biological communities (models that ignore differences between species often perform worse than models that account for those differences). Functional ecology is relevant to the surrogate concept because it provides a robust mechanism that can explain patterns of cooccurrence, and can therefore be used to identify reliable surrogates (Dehling et al. 2014). While useful, however, several conceptual and practical issues remain in how species traits are applied to biodiversity monitoring and management. In particular, the idea that species traits mediate interactions between species, or between a single species and its environment, is well accepted. Precisely which traits are used, however, and how they are defined and measured, remains challenging. Unfortunately, learning about traits that influence the distribution of plant species can rarely be transferred to understand patterns in animals, and vice versa. Of those traits that can be compared across dissimilar taxa (such as body mass, or related indices such as metabolic rates), controversy remains regarding their mechanistic basis and degree of explanatory power (Isaac and Carbone 2010). Further, work on animal traits remains largely descriptive, despite advances in predictive science based on plant traits. For example, Lindenmayer and colleagues (2014) showed that species selected for monitoring according to their traits did not occupy sites with high

7

8

Indicators and Surrogates of Biodiversity and Environmental Change

numbers of other bird species, negating their usefulness for conservation decision making. While potentially useful, therefore, functional approaches need to be carefully evaluated before being adopted as standard methods in biodiversity monitoring.

5.  Surrogates should measure progress towards management goals Biodiversity surrogates are a tool to provide reliable, up-to-date and cost-effective information on ecosystem state. Much research in the surrogate ecology literature has focused on the efficacy (Rodrigues and Brooks 2007) or efficiency (Kessler et al. 2011) of proposed surrogates, but this ignores a pivotal question: why monitor at all? Decision theory tells us that we should only collect data when: (1) we need that data to answer a question, and (2) the answer to that question will help us manage the system better, and so meet our management goals (McDonald-Madden et al. 2010). Therefore, the attributes of the system that you should monitor (i.e. your choice of surrogate) depend on what your goal is and how the system is changing in relation to your goal. In this context, it becomes clear that there is no single ‘best’ indicator that should always be selected regardless of context. Instead, the surrogate that should be chosen in a given situation is highly dependent on the question that needs to be investigated. Indeed, for simple or well-designed questions, it may be that no surrogate is needed at all, and that direct measurement is more efficient (Lindenmayer and Likens 2011). Despite this, few ecologists state their reasons for monitoring, the expected value of the resulting information for improving management, or even their overall management goals (Westgate et al. 2013). Without a management target or goal, the proximate problem considered by many articles on surrogates and indicators (i.e. which taxa are effective surrogates for each other) risks perpetuating the misleading suggestion that there exists a single surrogate that is valuable in all instances.

Knowledge gaps 6.  A long-term view of ecological surrogates Surrogate relationships can result from several processes, but the most reliable associations are likely to be those that have a known ecological or evolutionary basis. It follows, therefore, that a long-term view of how surrogate interactions form and change would aid understanding in this field. For example, processes such as fragmentation or disturbance can take decades to influence species distributions (Wearn et al. 2012). Even more broadly, the imprint of co-evolution (Jetz et al. 2004) or geological processes such as glaciation (Essl et al. 2011) can further influence species distributions and patterns of co-occurrence. Despite the strong influence of long-term processes on biodiversity, most assessments of surrogate relationships consist of ‘snapshot’ studies, while long-term monitoring programs of value for understanding consistency in surrogate relationships are similarly rare (Westgate et al. 2013). This longer view of consistency and change in surrogate relationships is developing (Barton et al. 2014), but needs further work. One way that the short-term view of surrogate relationships could be improved is by greater use of genetic data. The full range of applications for genetics data in surrogate ecology will be discussed in a later chapter (see Chapter 15), but two valuable research directions are particularly relevant here. First, several new methods, including genetic barcoding and environmental DNA, have enormous potential to reduce the cost of biodiversity monitoring (e.g. Thomsen et al. 2012). This is important because cost is a major factor influencing the longevity and effectiveness of ecological monitoring programs (Linden-

2: Surrogates for the distribution and trajectory of biodiversity

mayer and Likens 2009). Second, genetic analysis allows assessment of evolutionary rates, which over long periods may fundamentally alter species responses to each other or to their environment (Nuismer and Harmon 2014). These advances have enormous potential to improve ecologists’ ability to detect and understand long-term trends in biodiversity.

7.  Greater integration is needed across taxonomic boundaries Most research on biodiversity surrogates compares multiple species in one taxon against multiple species in a second taxon. While conceptually straightforward, this approach places primacy on taxonomic categories of biodiversity, despite evidence that phylogenetic relatedness has little influence on patterns of cross-taxon congruence (Westgate et al. 2014). With increases in our ability to monitor several taxonomic groups simultaneously (Kessler et al. 2011) comes greater potential to test more complex cross-taxon interactions in detail. Several authors have investigated methods for incorporating information from several dissimilar taxa in biodiversity surrogate research. One useful approach is to monitor species that have disproportionately large impacts on the number or diversity of species in an ecosystem, such as keystone species or ecosystem engineers (Caro 2010). Alternatively, maintaining functional diversity (rather than taxonomic or phylogenetic diversity) can be a valuable approach to biodiversity conservation, particularly where ecosystems include large numbers of functionally similar species (Gerisch et al. 2012). Finally, recognition of ecosystem change resulting from changes to species interactions (such as trophic cascades) has prompted some authors to consider monitoring and management of species interaction networks, though this remains an emerging field of research (Tylianakis et al. 2010). Each of these approaches has enormous potential for informing ecosystem-wide conservation via surrogates from a range of taxa. 8.  Research on the mechanisms underpinning surrogate relationships is not always worthwhile I have argued above that the mechanisms that underpin biodiversity congruence relationships are poorly understood, and that this represents a hindrance to the use of biodiversity surrogates (e.g. Barton et al. 2014). However, it does not follow that the need for reliable surrogates justifies any level of investment in research on ecological processes. This is because information on biodiversity should be collected only if it will help to make management decisions that, in turn, lead to better outcomes for biodiversity than would be possible in the absence of that information (McDonald-Madden et al. 2010). Surrogates that have been more thoroughly evaluated are less likely to generate surprising and deleterious results (e.g. falsely believing there is no decline in biodiversity), but also cost more to validate. Further, management experiments can become prohibitively large (and therefore expensive) as the number of processes being investigated increases, and so successful examples are rare (Westgate et al. 2013). Consequently, it is possible for the process of testing biodiversity surrogates to become more expensive than direct measurement, removing the economic justification for using surrogates at all (Lindenmayer and Likens 2011). The argument that we don’t always need a detailed understanding of ecological mechanisms can also be justified from a purely scientific perspective (i.e. without discussion of cost). In particular, several types of model exist that make useful predictions about the distribution of biodiversity from a limited amount of data. For example, approaches based on the maximum entropy formalism can be used to estimate how the abundance distribution of a group of species will change given certain constraints, without specifying the

9

10

Indicators and Surrogates of Biodiversity and Environmental Change

mechanisms by which those changes occur (e.g. see Dewar and Porté 2008). Similarly, autocorrelation analysis can give valuable information on the rate at which system state changes over space or time (Soininen et al. 2007), allowing scientifically informed but mechanism-free assessment of when and where monitoring should occur. Finally, there are significant difficulties when inferring ecological processes from pattern, as many kinds of processes generate similar patterns (McGill and Nekola 2010). Consequently, there is a growing argument that process-based approaches should be considered to be one option for understanding ecological systems, rather than a necessity in all cases.

9.  Complementary surrogates The principle of complementarity is most perhaps best known in ecology for its application to spatial prioritisation, where it is used to identify sets of sites that maximise the number of previously unprotected species that are included in future zoning arrangements. However, this principle can be applied more broadly, such as where alternate data sources provide different but complementary information on ecosystem state. A particularly useful advance for biodiversity surrogates would be to identify sets of indicators that – when studied or monitored in combination – best describe the total range of ecological responses within that ecosystem (Larsen et al. 2012). Despite growing awareness that complementary forms of information can help managers to make better conservation decisions, many articles on biodiversity surrogates seek only to identify sets of taxa that display strong, positive congruence relationships. This is somewhat ironic, as optimal ‘surrogate sets’ are likely to be those groups of species that display incongruent distributions, either over space or ecological gradients. Predicting sets of taxa that will provide complementary information is difficult, but is likely to include groups that use different ecological domains (Abell et al. 2011) or display markedly different functional traits such as body size (Velghe and Gregory-Eaves 2013). Such an approach would reduce the risk that some sections of the ecosystem could undergo drastic loss of biodiversity or function without that loss being observed by more taxonomically restricted monitoring programs. 10.  Adaptive surrogates I have argued above that different problems require different surrogates (point 5), while still suggesting that some surrogate relationships may be applicable across a range of locations, times, or taxa (points 6–8). How might ecologists reconcile the need for detailed, location-specific information with the desire to draw on existing knowledge? This is a challenge to applied ecological research and monitoring, where all decisions occur with some degree of uncertainty. However, a useful advance would be to better understand the circumstances where monitoring programs should shift between alternative biodiversity surrogates. An adaptive monitoring approach (Lindenmayer and Likens 2009) would be a valuable model to apply to surrogate identification and testing, for several reasons. The key advantage would be to allow monitoring to begin using the best available ‘global’ information, and then to update monitoring protocols as ‘local’ data became available. For example, different animal groups differ in both their representativeness of alternative taxa, and in the cost of monitoring them to a suitable degree of precision (Kessler et al. 2011). Therefore, it is plausible that ecologists may wish to switch between monitored taxa either as different ecological processes become important to management, or as their budgetary situation changes. However, it is critical that the process of transition is

2: Surrogates for the distribution and trajectory of biodiversity

carefully managed to avoid data inconsistency over time (Lindenmayer and Likens 2009). It is also important that ‘adaptive’ monitoring not be used as an excuse for trial and error monitoring and management, which can lead to undesirable conservation outcomes (Westgate et al. 2013).

Conclusions In this chapter, I have provided an overview of what I see as the key trends and future directions in the use of surrogates for biodiversity monitoring. I have argued that such monitoring requires the use of biodiversity surrogates, even when that use is not explicitly stated, because the measurement of total biodiversity is impossible. Developments in monitoring technologies such as remote sensing and genetics are likely to change the details of the methods that I have discussed here. However, the major challenges for the application of surrogates in this field are unlikely to change, and remain largely conceptual: What is biodiversity, and how should we measure and value it? Consequently, the key message of this chapter is that ecologists should be aware of the assumptions they make when seeking to measure and monitor biodiversity, and of the extent to which those assumptions (which may be implicit or explicit) can have unexpected or undesirable implications for conservation.

References Abell R, Thieme M, Ricketts TH, Olwero TH, Ng R, Petry P, et al. (2011) Concordance of freshwater and terrestrial biodiversity. Conservation Letters 4(2), 127–136. doi:10.1111/j.1755-263X.2010.00153.x. Barton PS, Westgate MJ, Lane PW, MacGregor C, Lindenmayer DB (2014) Robustness of habitat-based surrogates of animal diversity: a multi-taxa comparison over time and after fire. Journal of Applied Ecology 51(5), 1434–1443. doi:10.1111/1365-2664.12290. Bode M, Wilson KA, Brook TM, Turner WR, Mittermeier RA, McBride MF, et al. (2008) Cost-effective global conservation spending is robust to taxonomic group. Proceedings of the National Academy of Sciences of the United States of America 105(17), 6498–6501. doi:10.1073/pnas.0710705105. Caro T (2010) Conservation by Proxy: Indicator, Umbrella, Keystone, Flagship, and Other Surrogate Species. Island Press, Washington DC. Dehling DM, Töpfer T, Schaefer HM, Jordano P, Böhning-Gaese K, Schleuning M (2014) Functional relationships beyond species richness patterns: trait matching in plant–bird mutualisms across scales. Global Ecology and Biogeography 23(10), 1085–1093. doi:10.1111/ geb.12193. Dewar RC, Porté A (2008) Statistical mechanics unifies different ecological patterns. Journal of Theoretical Biology 251(3), 389–403. doi:10.1016/j.jtbi.2007.12.007. Essl F, Dullinger S, Plutzar C, Willner W, Rabitsch W (2011) Imprints of glacial history on correlations between endemic plant and invertebrate species richness. Journal of Biogeography 38(3), 604–614. doi:10.1111/j.1365-2699.2010.02425.x. Ewers RM, Marsh CJ, Wearn OR (2010) Making statistics biologically relevant in fragmented landscapes. Trends in Ecology & Evolution 25(12), 699–704. doi:10.1016/j.tree.2010.09.008. Gerisch M, Agostinelli V, Henle K, Dziock F (2012) More species, but all do the same: contrasting effects of flood disturbance on ground beetle functional and species diversity. Oikos 121(4), 508–515. doi:10.1111/j.1600-0706.2011.19749.x.

11

12

Indicators and Surrogates of Biodiversity and Environmental Change

Grantham HS, Pressey RL, Wells JA, Beattie AJ (2010) Effectiveness of biodiversity surrogates for conservation planning: different measures of effectiveness generate a kaleidoscope of variation. PLoS ONE 5(7), e11430. doi:10.1371/journal.pone.0011430. Hawkins BA, Porter EE (2003) Does herbivore diversity depend on plant diversity? The case of California butterflies. American Naturalist 161(1), 40–49. doi:10.1086/345479. Hurlbert AH, Jetz W (2007) Species richness, hotspots, and the scale dependence of range maps in ecology and conservation. Proceedings of the National Academy of Sciences of the United States of America 104(33), 13384–13389. doi:10.1073/pnas.0704469104. Isaac NJB, Carbone C (2010) Why are metabolic scaling exponents so controversial? Quantifying variance and testing hypotheses. Ecology Letters 13(6), 728–735. doi:10.1111/j.1461-0248.2010.01461.x. Jetz W, Rahbek C, Colwell RK (2004) The coincidence of rarity and richness and the potential signature of history in centres of endemism. Ecology Letters 7(12), 1180–1191. doi:10.1111/j.1461-0248.2004.00678.x. Kessler M, Abrahamczyk S, Bos M, Buchori D, Dwi Putra D, Gradstein SR, et al. (2011) Cost-effectiveness of plant and animal biodiversity indicators in tropical forest and agroforest habitats. Journal of Applied Ecology 48(2), 330–339. doi:10.1111/j.1365-2664. 2010.01932.x. Larsen FW, Bladt J, Balmford A, Rahbek C (2012) Birds as biodiversity surrogates: will supplementing birds with other taxa improve effectiveness? Journal of Applied Ecology 49(2), 349–356. doi:10.1111/j.1365-2664.2011.02094.x. Lindenmayer DB, Likens GE (2009) Adaptive monitoring: a new paradigm for long-term research and monitoring. Trends in Ecology & Evolution 24(9), 482–486. doi:10.1016/j. tree.2009.03.005. Lindenmayer D, Likens G (2011) Direct measurement versus surrogate indicator species for evaluating environmental change and biodiversity loss. Ecosystems 14(1), 47–59. doi:10.1007/s10021-010-9394-6. Lindenmayer DB, Lane PW, Westgate MJ, Crane M, Michael D, Okada S, et al. (2014) An empirical assessment of the focal species hypothesis. Conservation Biology 28(6), 1594– 1603. doi:10.1111/cobi.12330. McDonald-Madden E, Baxter PWJ, Fuller RA, Martin TG, Game ET, Montambault J, et al. (2010) Monitoring does not always count. Trends in Ecology & Evolution 25(10), 547–550. doi:10.1016/j.tree.2010.07.002. McGill BJ, Nekola JC (2010) Mechanisms in macroecology: AWOL or purloined letter? Towards a pragmatic view of mechanism. Oikos 119(4), 591–603. doi:10.1111/j.1600-0706. 2009.17771.x. Nuismer SL, Harmon LJ (2015) Predicting rates of interspecific interaction from phylogenetic trees. Ecology Letters 18(1), 17–27. doi:10.1111/ele.12384. Orme CDL, Davies RG, Burgess M, Eigenbrod F, Pickup N, Olson VA, et al. (2005) Global hotspots of species richness are not congruent with endemism or threat. Nature 436, 1016–1019. doi:10.1038/nature03850. Qian H, Kissling WD (2010) Spatial scale and cross-taxon congruence of terrestrial vertebrate plant species richness in China. Ecology 91(4), 1172–1183. doi:10.1890/09-0620.1. Rodrigues ASL, Brooks TM (2007) Shortcuts for biodiversity conservation planning: the effectiveness of surrogates. Annual Review of Ecology Evolution and Systematics 38, 713–737. doi:10.1146/annurev.ecolsys.38.091206.095737. Sætersdal M, Gjerde I (2011) Prioritising conservation areas using species surrogate measures: consistent with ecological theory? Journal of Applied Ecology 48(5), 1236–1240. doi:10.1111/j.1365-2664.2011.02027.x.

2: Surrogates for the distribution and trajectory of biodiversity

Scheffers BR, Joppa LN, Pimm SL, Laurance WF (2012) What we know and don’t know about earth's missing biodiversity. Trends in Ecology & Evolution 27(9), 501–510. doi:10.1016/j. tree.2012.05.008. Soininen J, McDonald R, Hillebrand H (2007) The distance decay of similarity in ecological communities. Ecography 30(1), 3–12. doi:10.1111/j.0906-7590.2007.04817.x. Supp SR, Ernest SKM (2014) Species-level and community-level responses to disturbance: a cross-community analysis. Ecology 95(7), 1717–1723. doi:10.1890/13-2250.1. Thomsen PF, Kielgast JOS, Iversen LL, Wiuf C, Rasmussen M, Gilbert MTP, et al. (2012) Monitoring endangered freshwater biodiversity using environmental DNA. Molecular Ecology 21(11), 2565–2573. doi:10.1111/j.1365-294X.2011.05418.x. Tylianakis JM, Laliberté E, Nielsen A, Bascompte J (2010) Conservation of species interaction networks. Biological Conservation 143(10), 2270–2279. doi:10.1016/j.biocon.2009.12.004. Velghe K, Gregory-Eaves I (2013) Body size is a significant predictor of congruency in species richness patterns: a meta-analysis of aquatic studies. PLoS ONE 8(2), e57019. doi:10.1371/ journal.pone.0057019. Wang Y, Naumann U, Wright ST, Warton DI (2012) Mvabund – an r package for model-based analysis of multivariate abundance data. Methods in Ecology and Evolution 3(3), 471–474. doi:10.1111/j.2041-210X.2012.00190.x. Wearn OR, Reuman DC, Ewers RM (2012) Extinction debt and windows of conservation opportunity in the Brazilian amazon. Science 337(6091), 228–232. doi:10.1126/ science.1219013. Westgate MJ, Likens GE, Lindenmayer DB (2013) Adaptive management of biological systems: a review. Biological Conservation 158, 128–139. doi:10.1016/j.biocon.2012.08.016. Westgate MJ, Barton PS, Lane PW, Lindenmayer DB (2014) Global meta-analysis reveals low consistency of biodiversity congruence relationships. Nature Communications 5, 3899. doi:10.1038/ncomms4899. Williams P, Faith D, Manne L, Sechrest W, Preston C (2006) Complementarity analysis: mapping the performance of surrogates for biodiversity. Biological Conservation 128, 253–264. doi:10.1016/j.biocon.2005.09.047.

13

This page intentionally left blank

‡7HQ&RPPLWPHQWVSSLQGG

$0

3

Biodiversity surrogates David Lindenmayer, Philip Barton, Martin Westgate, Peter Lane and Jennifer Pierson

Things we know 1 2 3 4

There is a need to develop and then apply consistent terminology. Surrogates are essential for biodiversity conservation. There is a massive interest in surrogates and indicators. Almost all biotic components of ecosystems have been proposed as an indicator or surrogate. 5 Sometimes it is easier and more appropriate to directly measure a given entity than seek surrogates for that entity. 6 Some species have anti-surrogate or anti-indicator properties. Knowledge gaps 7 There is a paucity of testing and validation of indicators and surrogates. 8 There is disconnect between scientific and academic study of surrogates and their practical application on the ground. 9 There are lessons that can be learned from other disciplines – especially medicine. 10 There is a sequence of steps that can help improve the selection and application of indicators and surrogates.

Introduction One of the most extensive bodies of literature in the fields of ecology and conservation biology focuses on biodiversity surrogates, including indicator species and other kinds of related proxies (Westgate et al. 2014). Indeed, many kinds of surrogates and indicators have been proposed for use in conservation and environmental management (Caro 2010; Westgate et al. 2014). The use of surrogates and indicators is important because it is impossible to measure, manage and conserve all elements of biota in all ecosystems and at all times. In this chapter, we outline 10 major developments, problems and opportunities associated with the identification and application of biodiversity surrogates.

15

16

Indicators and Surrogates of Biodiversity and Environmental Change

Things we know 1.  There is a need to develop and then apply consistent terminology Within the extensive literature on biodiversity surrogates, the terms ‘indicators’, ‘indicator species’, ‘focal species’ and other kinds of proxies have come to have various meanings (Caro 2010). However, an overarching definition of a biodiversity surrogate that is appropriate for this chapter is (adapted from McGeoch 1998, p. 185): Any measure ‘… that readily reflects: the biotic or abiotic state of an environment; represents the impact of an environmental change on a habitat, community or ecosystem; or is indicative of the diversity of a subset of taxa, or of wholesale diversity, within an area.’ This overarching definition encompasses all three kinds of surrogates or indicators that some authors have classified as environmental indicators, ecological indicators and biodiversity indicators, respectively (McGeoch 1998; Duelli and Obrist 2003). The definition and concept of indicator species (and related kinds of biodiversity surrogates) has become confused in the past decade. Indeed, some modern applications of the term ‘indicator’ no longer relate to a relationship between a proxy and a target entity but simply to whatever entity is being measured (Woodward et al. 1999). That is, whatever entity that is being measured is simply termed ‘an indicator’, even if surrogacy relationships are not claimed (see Lindenmayer and Likens 2011). The focal species approach is another example in the surrogate and indicators literature where terminology is used inconsistently and in a confused way. The focal species approach has specific ecological meaning (Lambeck 1997). Under the approach, known threatening processes in a given landscape are described. The species most sensitive to each threat are then identified. One or more species may be identified for each threat. These are the focal species (Lindenmayer et al. 2014). However, some workers use the term focal species to refer to any taxa that is the focus for study, irrespective of whether they have surrogate or indicator value (or otherwise).

2.  Surrogates are essential for biodiversity conservation A wide range of kinds of surrogates have been proposed for use in conservation and environmental management (Caro 2010). Many arguments have been made in support of the widespread use of biodiversity surrogates and various kinds of indicator species in ecology and conservation. These include: (1) there is a biodiversity crisis (and conservation biology is a crisis discipline) and therefore conservation action must be taken now, even with imperfect information using indicators and surrogates to guide key interventions such as the establishment of protected areas; (2) it is financially and logistically impossible to work on the management and conservation of every environment and every species in all environments (Duelli and Obrist 2003); and (3) targeting particular biodiversity surrogates for conservation action is a powerful and readily understood communication tool as well as a good ‘social hook’ for human community engagement in conservation activities (Lindenmayer and Fischer 2003). 3.  There is a massive interest in surrogates and indicators Over the past few decades a very large, and rapidly expanding, volume of literature has accumulated on biodiversity surrogates (e.g. Landres et al. 1988; Niemi et al. 1997; McGeoch 1998; Rodrigues and Brooks 2007; Lewandowski et al. 2010; Lindenmayer and Likens

3: Biodiversity surrogates

Table 3.1.  Examples of the use of biodiversity surrogates and various kinds of indicator species in different arenas of ecology, conservation and environmental management. Field

Examples

Restoration ecology

(Zeppelini et al. 2009; Fagan et al. 2010)

Agri-ecological conservation

(Billeter et al. 2008; Kaiser et al. 2010)

Forest ecology and management

(Drever et al. 2008; Saha and Haldar 2009)

Plantation tree management

(Roundtable for Sustainable Palm Oil 2007; Smith et al. 2008)

Pollution ecology

(Dixit et al. 1992; Edmondson et al. 2010; Basile et al. 2011)

Pest management

(Bernard et al. 2010)

Conservation planning

(reviewed by Rodrigues and Brooks 2007; Mitrovich et al. 2010)

Rare and threatened species management

(Wenger 2008)

Climate change

(Ellis et al. 2009; Pearman et al. 2011)

Ecological monitoring

(Cantarello and Newton 2008)

2011; Westgate et al. 2014). Biodiversity surrogate approaches are now used widely in all forms of applied ecology (Table 3.1).

4.  Almost all biotic components of ecosystems have been proposed as an indicator or surrogate Almost all biotic components of ecosystems – ranging from viruses, bacteria, fungi, plants and insects, to all groups of vertebrates – have been proposed as surrogates (Lindenmayer and Likens 2011). However, it is impractical for all of them to be surrogates. It is critical to develop new approaches and methods to better determine how or when to prioritise among such a huge array of suggested proxies. 5.  Sometimes it is easier and more appropriate to directly measure a given entity than seek surrogates for that entity Surrogate approaches contrast with a direct measurement approach; in the latter approach, the focus is on a single entity or a highly targeted subset of entities in a given ecosystem, but no surrogacy relationships with unmeasured entities are assumed (Lindenmayer and Likens 2011). A goal of the direct measurement approach is to demonstrate mechanistic, causal or other relationships between key attributes of the target ecosystem system (e.g. particular environmental conditions) and the entities selected for measurement. The direct measurement approach can be valuable because some applications of surrogate approaches appear to be circular and inefficient. That is, it requires more work to identify and apply a proxy than it does to directly measure the entity for which it is considered to be indicative (Seddon and Leech 2008; Lindenmayer and Likens 2011). As an example, the Capercaillie Grouse Tetrao urogallus has been claimed as an indicator of open habitats in Europe (Braunisch and Suchant 2008). In this case, it is likely to be easier, less expensive and more tractable to measure directly open habitats than take the step of finding a surrogate for this ecosystem. Nevertheless, the direct measurement approach targeted at a given variable of specific interest is not without limitations. These include that the entities targeted for measurement are well understood and can be accurately measured, and that there is sufficient

17

18

Indicators and Surrogates of Biodiversity and Environmental Change

understanding about key ecological processes influencing those entities. The direct measurement approach is reductionist, and many elements of the biota, biotic processes and environmental factors must be ignored because of practical considerations.

6.  Some species have anti-surrogate or anti-indicator properties in landscapes Much work has focused on identifying those species that will act as proxies for the occurrence of high levels of richness of other species, or the richness of species of conservation concern (Westgate et al. 2014). However, recent work has shown there can also be value in identifying those species that are typically associated with low levels of species richness. For example, in a study of Australian woodland birds, Lindenmayer and colleagues (2014) found strong evidence of two inter-related kinds of negative surrogacy patterns: those in which a particular species of conservation concern was negatively associated with the occurrence of a given focal species (sensu Lambeck 1997), and those in which a given nominated focal species typically occurred on sites with low species richness. Negative surrogacy implies that although the occurrence of some focal species may be associated with many other species (including species of conservation concern), there will nevertheless be other focal species that might be associated with low levels of co-occurrence of other species. This implies that management actions targeted for a given species may not benefit many other species (including species of conservation concern).

Knowledge gaps 7.  There is a paucity of testing and validation of indicators and surrogates Few proposed indicator species or biodiversity surrogates have been validated through rigorous testing in which the spatial, temporal and taxonomic boundaries for their useful application have been set. However, many authors (see McGeoch 1998; Duelli and Obrist 2003; Heink and Kowarik 2010; Lewandowski et al. 2010; Lindenmayer and Likens 2011) have highlighted how the correlative relationships that underpin the application of the indicator species and biodiversity surrogate approaches have rarely been subjected to detailed testing and validation (e.g. see Fraser et al. 2009; Mistry et al. 2008; Saha and Haldar 2009), especially with additional data that were not used to construct the initial relationship (Fleishman and Murphy 2009). Testing and validation is needed to: (1) assess the strength of surrogacy relationships and, in turn, determine levels of predictive ability; (2) elucidate the ecological processes underlying patterns of surrogacy and create a pathway to connect work on indicator species and surrogates with other areas of ecological research, including theory; and, (3) using insights from 1 and 2 above, define the spatial, temporal and taxonomic boundaries where the use of a biodiversity surrogate is valid or not (e.g. see Roth and Weber 2008). Testing and validation are important because many claims about the validity of indicators have been made on the basis of assertion, advocacy and the public appeal of charismatic groups of organisms (reviewed by Fleishman and Murphy 2009). However, there is often a lack of empirical support about such assertions (McGeoch 1998; Duelli and Obrist 2003; Heink and Kowarik 2010; Lindenmayer and Likens 2011). A second reason to conduct rigorous testing and validation is to identify the ecological (causal) processes that result in particular species or groups being robust indicator species or biodiversity surrogates. In some cases, establishing causality may require a detailed understanding of natu-

3: Biodiversity surrogates

ral history (Fleishman and Murphy 2009) and/or careful experimental work, as highlighted in the case of indicator plants and nitrogen levels in moorland environments in the United Kingdom (e.g. Edmondson et al. 2010).

8.  There is disconnect between scientific and academic study of surrogates and their practical application on the ground The scientific literature is replete with an ever-expanding number of studies of indicators and surrogates (Caro 2010; Westgate et al. 2014). There is no doubting the increasing quality of this body of work. Yet, this research momentum has not resonated particularly strongly with applied ecologists or practical conservation biologists working on the ground or in seascapes. There is no doubt that practical application will always lag behind research, and perhaps that is appropriate to ensure that ideas and approaches are reasonably settled before they underpin on-the-ground work. However, issues associated with the strengths and weaknesses of surrogate approaches rarely appear to be well understood by practitioners, leading to risks of management mistakes (e.g. Walker 1981) (see Chapter 11). This suggests that researchers need to temper their desire to oversell their research findings in science marketing (to get their papers published) so that practitioners can more readily separate substance from spin. 9.  There are lessons that can be learned from other disciplines – especially medicine Some of the best applications of surrogate-like approaches come, arguably, from the medical sciences (Niemi et al. 1997; Barton et al. 2015). Most importantly, there is a well-developed conceptual, mathematical and statistical framework for the explicit use of objectives, surrogates and outcomes (endpoints) in the medical sciences (see Chapter 16). This framework is particularly powerful when applied to clinical trials, where results obtained from the direct measurement of a given entity (e.g. the rate of patient mortality resulting from lung cancer) must be very strongly concordant with results obtained from a proposed surrogate (such as the amount and length of time of smoking; Begg and Leung 2000; Barton et al. 2015). Under such a statistical framework, the greater the statistical association between the direct measure and the indicator or surrogate, the lower the probability of a misleading inference arising from the use of that indicator or surrogate. Of course, we fully appreciate that natural ecosystems and their associated biodiversity are far more complex than a single species system (i.e. humans). But we suggest the general principles and underlying thinking developed from approaches generated from the medical sciences will often also apply in the ecological and conservation sciences. 10.  There is a sequence of steps that can help improve the selection and application of indicators and surrogates Despite the burgeoning use of indicator species and biodiversity surrogates around the world (reviewed by Westgate et al. 2014), much controversy remains about their use. Indeed, there have been many critiques of the indicator species and biodiversity surrogate approaches (Landres et al. 1988; Andelman and Fagan 2000; Lindenmayer et al. 2000; Roth and Weber 2008; Seddon and Leech 2008; Cushman et al. 2010). Some of the key criticisms to date have been: 1 Almost all groups of organisms, ranging from viruses to all groups of vertebrates, have been proposed as indicator species or biodiversity surrogates. However, few

19

20

Indicators and Surrogates of Biodiversity and Environmental Change

2

3 4 5

studies have discussed how or when to prioritise among such an array of suggested proxies. Numerous applications of biodiversity surrogate approaches have been narrow and idiosyncratic and correspondingly appear to have limited predictive ability to other sets of circumstances, including between groups of taxa and across spatial and temporal scales. Some applications of biodiversity surrogates require more work to identify and apply a proxy than the direct measurement of the entity for which it is considered to be indicative. Few biodiversity surrogates have been validated through rigorous testing in which the spatial, temporal and taxonomic boundaries for their useful (and not useful) application have been set. The ecological processes underlying pattern-based surrogacy relationships have rarely been elucidated. Hence, work on biodiversity surrogates has often remained distant from other fields of ecological work like niche theory, and community ecology.

To tackle the problems (at least partially), we present a checklist of considerations to guide the selection and application of robust indicator species and biodiversity surrogates (see Table 3.2). Our hope is that the application of the checklist may catalyse a greater rigour in the use of biodiversity surrogates that we suggest has often been lacking to date. Table 3.2.  Checklist of factors to consider in guiding the selection and application of robust biodiversity surrogates Question regarding biodiversity surrogate

Explanation

Can the objectives be explicitly and precisely articulated?

Statements of objectives should be framed explicitly to encompass a well-identified target (e.g. the population of a target species), explicit spatial context (e.g. management region), explicit temporal context (e.g. 100 years) and an estimate of uncertainty (e.g. 95% confidence).

Can information about the biodiversity surrogate be readily communicated to others?

Ecological information about the biodiversity surrogate needs to be readily communicated to stakeholders. This ability should include: what the measure is (e.g. the species richness of a particular group like ants); the entities for which it is a proxy (e.g. total species richness of all invertebrates, richness dissimilarity of all vascular plants); why it has been selected (e.g. the proxy is too difficult or too expensive to measure); what the objectives and outcomes (endpoints) are (e.g. species protection), and how the surrogate links both with objectives and endpoints.

Is it efficient and readily measured?

The biodiversity surrogate must be more easily and accurately measured than the entity or entities for which it is purported to be indicative.

Is it cost-effective?

The biodiversity surrogate must cost less to identify and then develop (including testing; see below) than the direct measurement of the entity for which it is a proxy. This cost may include assessments of the relative cost-effectiveness of different biodiversity surrogates (including composite sets of different groups of taxa).

3: Biodiversity surrogates

Question regarding biodiversity surrogate

Explanation

Is it risk averse? Can the uncertainty of risk be determined and the uncertainty of risk and application be specified?

The risks of poor management decisions associated with the selection of the wrong surrogate or a weak surrogate need to be assessed and articulated well to stakeholders.

Is it testable and can it be validated?

Any valid biodiversity surrogate must be subjected to rigorous testing including the strength of surrogacy relationships. For example, this testing may include quantification of correlations between the population size of a proxy species and that of the species for which it is considered to be indicative, or the correlation between different groups of taxa.

What is the nature and strength of spatial, temporal or taxonomic transferability?

As a consequence of testing, it should be possible to determine the predictive ability of a biodiversity surrogate. This predictability would aim to articulate the spatial, temporal and taxonomic boundaries within which an indicator species or biodiversity surrogate is valid, but beyond which there is weak surrogacy and the risk of failure is high.

Is it possible to identify the ecological processes underlying the patterns of surrogacy relationships?

A valuable step in work on biodiversity surrogates is to identify the mechanism underpinning relationships between a proxy and the entity for which it is indicative. These relationships would have the dual aim of better determining spatial, temporal and taxonomic predictive ability, as well as strengthening links between surrogate research and other areas of work in ecology, particularly ecological theory.

Does the biodiversity surrogate correlate strongly with the direct measurement of the desired outcome (i.e. the endpoint)?

The critical test of the validity of a proxy is a strong statistical association between it and a direct measure of the outcome or endpoint. For example, reserves selected to include wellknown large predators (the surrogate) also capture high species richness of other groups (the endpoint). The stronger this association, the lower the probability of a misleading inference arising from the use of that indicator or surrogate.

Conclusion The field of surrogates is a huge and rapidly expanding area of research in ecology and conservation. We suggest there are key areas of work needed to improve the application of surrogates. In particular, we argue there is an urgent need to: 1 Ensure that there is a broadly and generally agreed set of terms that can be applied by researchers and managers. Models for developing common terminology are well developed in other fields. As an example, there is a dictionary of forestry terms that has been developed and routinely applied by the Society of America Foresters (Helms 1998). It is time for those working on surrogates to follow suit. 2 Recognise that the very hard work associated with identifying and applying robust surrogates may not be warranted in some cases as the direct measurement of particular entities may be a more cost-effective and tractable alternative (Lindenmayer and Likens 2011). 3 Ensure that there is more and better testing and validation of surrogates so that their strengths and limitations can be made clearer to those developing and/or

21

22

Indicators and Surrogates of Biodiversity and Environmental Change

applying them. This is essential to determine the taxonomic, temporal and spatial boundaries within which surrogates are valid and those beyond which they are not (for example, see Westgate et al. 2014 for work on cross-taxon congruency). 4 Look beyond ecology to other disciplines, such as medicine, for ideas, concepts and innovations that can be adopted in the application of surrogates (Barton et al. 2015).

Acknowledgements The work in this chapter is based on an Australian Research Council Laureate Fellowship.

References Andelman SJ, Fagan WF (2000) Umbrellas and flagships: efficient conservation surrogates or expensive mistakes? Proceedings of the National Academy of Sciences 97, 5954–5959. Barton PS, Pierson JC, Westgate MJ, Lane PW, Lindenmayer DB (2015) Learning from clinical medicine to improve the use of surrogates in ecology. Oikos 124, 391­–398. Basile ER, Avery HW, Bien WF, Keller JM (2011) Diamondback terrapins as indicator species of persistent organic pollutants: using Barnegat Bay, New Jersey as a case study. Chemosphere 82, 137–144. doi:10.1016/j.chemosphere.2010.09.009. Begg CB, Leung DHY (2000) On the use of surrogate end points in randomized trials. Journal of the Royal Statistical Society. Series A, (Statistics in Society) 163, 15–28. doi:10.1111/1467985X.00153. Bernard MB, Cole P, Kobelt A, Horne PA, Altmann J, Wratten SD, et al. (2010) Reducing the impact of pesticides on biological control in Australian vineyards: pesticide mortality and fecundity effects on an indicator species, the predatory mite Euseius victoriensis (Acari: Phytoseiidae). Ecotoxicology (London, England) 103, 2061–2071. Billeter R, Liira J, Bailey D, Bugter R, Arens P, Augenstein I, et al. (2008) Indicators for biodiversity in agricultural landscapes: a pan-European study. Journal of Applied Ecology 45, 141–150. doi:10.1111/j.1365-2664.2007.01393.x. Braunisch V, Suchant R (2008) Using ecological forest site mapping for long-term habitat suitability assessments in wildlife conservation - demonstrated for Capercaillie (Tetrao urogallus). Forest Ecology and Management 256, 1209–1221. doi:10.1016/j. foreco.2008.06.027. Cantarello E, Newton AC (2008) Identifying cost-effective indicators to assess the conservation status of forested habitats in Natura 2000 sites. Forest Ecology and Management 256, 815–826. doi:10.1016/j.foreco.2008.05.031. Caro T (2010) Conservation by Proxy: Indicator, Umbrella, Keystone, Flagship, and Other Surrogate Species. Island Press, Washington DC. Cushman SA, McKelvey KS, Noon BR, McGarigal K (2010) Use of abundance of one species as a surrogate for abundance of others. Conservation Biology 24, 830–840. Dixit SS, Smol JP, Kingston JC, Charles DF (1992) Diatoms – powerful indicators of environmental change. Environmental Science & Technology 26, 22–33. doi:10.1021/ es00025a002. Drever MC, Aitken KE, Norris AR, Martin K (2008) Woodpeckers as reliable indicators of bird richness, forest health and harvest. Biological Conservation 141, 624–634. doi:10.1016/j.biocon.2007.12.004. Duelli P, Obrist MK (2003) Biodiversity indicators: the choice of values and measures. Agriculture, Ecosystems & Environment 98, 87–98. doi:10.1016/S0167-8809(03)00072-0.

3: Biodiversity surrogates

Edmondson JL, Carroll JA, Price EA, Caporn SJ (2010) Bio-indicators of nitrogen pollution in heather moorland. The Science of the Total Environment 408, 6202–6209. doi:10.1016/j. scitotenv.2010.08.060. Ellis CJ, Yahr R, Coppins BJ (2009) Local extent of old-growth woodland modifies epiphyte response to climate change. Journal of Biogeography 36, 302–313. doi:10.1111/j.1365-2699. 2008.01989.x. Fagan KC, Pywell RF, Bullock JM, Marrs RH (2010) Are ants useful indicators of restoration success in temperate grasslands? Restoration Ecology 18, 373–379. doi:10.1111/j.1526-100X. 2008.00452.x. Fleishman E, Murphy DD (2009) A realistic assessment of the indicator potential of butterflies and other charismatic taxonomic groups. Conservation Biology 23, 1109–1116. doi:10.1111/j.1523-1739.2009.01246.x. Fraser SE, Beresford AE, Peters J, Redhead JW, Welch AJ, Mayhew PJ, et al. (2009) Effectiveness of vegetation surrogates for parasitoid wasps in reserve selection. Conservation Biology 23, 142–150. doi:10.1111/j.1523-1739.2008.01069.x. Heink U, Kowarik I (2010) What criteria should be used to select biodiversity indicators? Biodiversity and Conservation 19, 3769–3797. doi:10.1007/s10531-010-9926-6. Helms JA (Ed.) (1998) The Dictionary of Forestry. Society of American Foresters, Bethesda, MD. Kaiser T, Rohner M-S, Matzdorf B, Kiesel J (2010) Validation of grassland indicator species selected for result-orientated agri-environmental schemes. Biodiversity and Conservation 19, 1297–1314. doi:10.1007/s10531-009-9762-8. Lambeck RJ (1997) Focal species: a multi-species umbrella for nature conservation. Conservation Biology 11, 849–856. doi:10.1046/j.1523-1739.1997.96319.x. Landres PB, Verner J, Thomas JW (1988) Ecological uses of vertebrate indicator species: a critique. Conservation Biology 2, 316–328. doi:10.1111/j.1523-1739.1988.tb00195.x. Lewandowski AS, Noss RF, Parsons DR (2010) The effectiveness of surrogate taxa for the representation of biodiversity. Conservation Biology 24, 1367–1377. doi:10.1111/j.1523-1739. 2010.01513.x. Lindenmayer DB, Fischer J (2003) Sound science or social hook - a response to Brooker’s application of the focal species approach. Landscape and Urban Planning 62, 149–158. doi:10.1016/S0169-2046(02)00147-0. Lindenmayer DB, Likens GE (2011) Direct measurement versus surrogate indicator species for evaluating environmental change and biodiversity loss. Ecosystems 14, 47–59. doi:10.1007/ s10021-010-9394-6. Lindenmayer DB, Margules CR, Botkin DB (2000) Indicators of biodiversity for ecologically sustainable forest management. Conservation Biology 14, 941–950. Lindenmayer DB, Lane PW, Westgate MJ, Crane M, Michael D, Okada S, Barton PS (2014) An empirical assessment of the focal species hypothesis. Conservation Biology 28, 1594–1603. doi:10.1111/cobi.12330. McGeoch MA (1998) The selection, testing and application of terrestrial insects as bioindicators. Biological Reviews of the Cambridge Philosophical Society 73, 181–201. doi:10.1017/S000632319700515X. Mistry J, Berardi A, Simpson M (2008) Birds as indicators of wetland status and change in the North Rupunini, Guyana. Biodiversity and Conservation 17, 2383–2409. doi:10.1007/ s10531-008-9388-2. Mitrovich MJ, Matsuda T, Pease KH, Fisher RN (2010) Ants as a measure of effectiveness of habitat conservation planning in southern California. Conservation Biology 24, 1239– 1248. doi:10.1111/j.1523-1739.2010.01486.x.

23

24

Indicators and Surrogates of Biodiversity and Environmental Change

Niemi GJ, Hanowski JM, Lima AR, Nicholls T, Weiland N (1997) A critical analysis on the use of indicator species in management. The Journal of Wildlife Management 61, 1240–1252. doi:10.2307/3802123. Pearman PB, Fuisan A, Zimmermann NE (2011) Impacts of climate change on Swiss biodiversity: an indicator taxa approach. Biological Conservation 144, 866–875. doi:10.1016/j.biocon.2010.11.020. Rodrigues AS, Brooks TM (2007) Shortcuts for biodiversity conservation planning: the effectiveness of surrogates. Annual Review of Ecology Evolution and Systematics 38, 713–737. doi:10.1146/annurev.ecolsys.38.091206.095737. Roth T, Weber D (2008) Top predators as indicators for species richness? Prey species are just as useful. Journal of Applied Ecology 45, 987–991. doi:10.1111/j.1365-2664.2007.01435.x. Roundtable for Sustainable Palm Oil (2007) Principles and Criteria for Sustainable Palm Oil Production. RSPO, Kuala Lumpur, Malaysia. Saha JJ, Haldar P (2009) Acridids as indicators of disturbance in dry deciduous forest of West Bengal in India. Biodiversity and Conservation 18, 2343–2350. doi:10.1007/s10531-0099591-9. Seddon PJ, Leech T (2008) Conservation short cut, or long and winding road? A critique of umbrella species criteria. Oryx 42, 240–245. Smith GF, Gittings T, Wilson M, French L, Oxbrough A, O'Donoghue S, et al. (2008) Identifying practical indicators of biodiversity for stand-level management of plantation forests. Biodiversity and Conservation 17, 991–1015. doi:10.1007/s10531-007-9274-3. Walker KF (1981) Ecology of freshwater mussels in the River Murray. Australian Water Resources Council Technical Paper No. 63. Australian Government Publishing Service, Canberra. Wenger SJ (2008) Use of surrogates to predict stressor response of imperiled species. Conservation Biology 22, 1564–1571. doi:10.1111/j.1523-1739.2008.01013.x. Westgate M, Barton PW, Lane P, Lindenmayer DB (2014) Global meta-analysis reveals biodiversity congruence is highest at spatial extremes. Nature Communications 5, 3899. doi:10.1038/ncomms4899. Woodward A, Jenkins K, Schreiner EG (1999) The role of ecological theory in long-term monitoring: report on a workshop. Natural Areas Journal 19, 223–233. Zeppelini D, Bellini BC, Creao-Durate AJ, Hernandez M (2009) Collembola as bioindicators of restoration in mined sand dunes of northeastern Brazil. Biodiversity and Conservation 18, 1161–1170. doi:10.1007/s10531-008-9505-2.

4

Conservation by proxy: thoughts 5 years on Tim Caro

Things we know 1 We need to continue the search for effective indicators of biodiversity, particularly those used to establish reserves. 2 The concept of umbrella species may need to be abandoned because of persistent failure to find evidence that species or populations predict population health or species richness at a local scale. 3 Apex predators are often threatened but they are keystone components in terrestrial and aquatic ecosystems. 4 Bird species diversity and abundance are often used as ecological disturbance indicator species to assess consequences of land use change. 5 Flagship species should be chosen using social marketing tools rather than ad hoc methods. Knowledge gaps 6 It may be important to directly measure the effects of land use change on species rather than the questionable use of other species as cross-taxonresponse indicator species. 7 There is value in simplifying and extending the use of landscape species to identify wildlife corridors. 8 Use sentinel species to alter public opinion about global warming and establish new reserves to mollify effects of climate change. 9 Reduce the use of management indicator species to assess management activities and the success of restoration projects. 10 Use marketing tools to advertise the importance of individual protected areas.

Introduction In this chapter I outline some developments in our understanding of surrogate species since my book Conservation by Proxy was published 5 years ago (Caro 2010) and then suggest where surrogate concepts might be useful in the future. This chapter therefore takes a species-based approach to conservation. 25

26

Indicators and Surrogates of Biodiversity and Environmental Change

Things we know 1.  We need to continue the search for effective indicators of biodiversity, particularly those used to establish reserves There is general consensus on where the location of, and threats to, biodiversity are found at a global scale. For instance, most of the nine major terrestrial biodiversity prioritisation protocols that examine vulnerability and irreplaceability of regions agree that tropical forests and Mediterranean biomes require conservation attention (Brooks et al. 2006). Therefore, further large-scale exercises that pinpoint areas of biodiversity using indicator species are becoming redundant. Instead, the more pressing task is to use biodiversity indicators at a medium spatial scale to identify where reserves or reserve networks should be placed. Another key task is to develop methods that rely on easy-to-measure species’ presence– absence data as surrogates for other species (e.g. Halme et al. 2009; Larsen et al. 2012; Nicholson et al. 2013). Biological indicators of species richness work best in complementarity algorithms that identify where to establish networks of reserves (Lewandowski et al. 2010). In reality, however, reserve establishment often depends more on trying to protect relatively untouched areas, local stakeholder compliance and governmental sanction, rather than on detailed biological measures, at least in terrestrial biomes. In marine habitats, locations of high diversity and whereabouts of important ecosystems are less well documented, so biodiversity indicators at a large scale may still prove useful (Stuart-Smith et al. 2013; Olds et al. 2014). 2.  The concept of umbrella species may need to be abandoned because of persistent failure to find evidence that species or populations predict population health or species richness at a local scale Whether umbrella species are used to predict species richness at a local level (local umbrella species), or guarantee the survival of populations of other sympatric species if they are protected (classic umbrella species), or are used by managers to monitor the state of other populations (management indicator species) (sensu Caro 2010), the evidence for their efficacy is still mixed at best. Generally, studies have failed to find evidence for large species, or particular functional groups, or certain trophic levels, or similar taxonomic groups acting as effective umbrellas for other species at medium and small scales (e.g. Branton and Richardson 2011; Cushman et al. 2010). Almost out of desperation, studies have explored using multiple umbrella species simultaneously (Sattler et al. 2014), although this necessarily undermines their utility as a shortcut. Currently then, there are few generalisations that can be made about choosing appropriate umbrella species and we may need to accept that umbrella species are not a particularly robust surrogate concept in conservation planning or in management. However, they are still (unfortunately) seen as a viable shortcut to quantifying aspects of biodiversity or management success at a small scale. 3.  Apex predators are often threatened but they are keystone components in terrestrial and aquatic ecosystems Apex predators in both terrestrial and marine systems are now viewed as pivotal to structuring ecosystems through alternating positive and negative effects on lower trophic levels (Heithaus et al. 2008; Jorge et al. 2013). Additionally, they restructure lower tropic levels by controlling the abundance of mid-sized carnivores (Terborgh and Estes 2010; Hollings et al. 2014). Apex predators are viewed as keystone species because of their disproportionate effects on lower trophic levels and ecologists are pressing for their conservation for this reason, as well as their role as flagship species (Estes et al. 2011; Ripple et al. 2014). For

4: Conservation by proxy: thoughts 5 years on

example, Letnic and colleagues (2009) even advocate maintaining an alien top predator, the Dingo (Canis lupus dingo), in Australia because by limiting Red Fox (Vulpes vulpes) populations, they increase the diversity of small native mammals (but not kangaroos). Many frugivorous birds and mammals in neotropical forests are also keystone species because they are such important seed dispersers (Vidal et al. 2013) and we face the challenge of maintaining their populations in the face of bushmeat consumption.

4.  Bird species diversity and abundance are often used as ecological disturbance indicator species to assess consequences of land use change Recently, much research on the effects of land use change in the tropics and subtropics, farming in Europe, and urbanisation in developed countries has used ecological-disturbance indicator species to measure biological effects of habitat alteration. Birds are often viewed as the most cost-effective indicator species of choice (e.g. Gardner et al. 2008; Peck et al. 2014). For instance, seed-dispersing birds are more sensitive to the extent of rainforest cover around forest fragments rather than fragment size itself in subtropical Queensland, Australia (Moran and Catterall 2014). Growth in agricultural production can occur in two main ways: either through highyield farming on existing agricultural land and (hopefully) concomitant protection of natural habitats from land conversion, or lower yield outcomes that seek to conserve some biodiversity on farmed land. These ‘land sparing’ or ‘land sharing’ outcomes are often assessed using either guilds of birds or trees. For example, Phalan et al. (2011) showed that birds and trees in India and Ghana, especially species with small global ranges, benefited more from ‘land sparing’ than ‘land sharing’. Of course this raises the issue of whether these groups can be surrogates for other taxa such as amphibians and arthropods. Some studies examining taxon-specific responses across land uses suggest that it is risky to assume taxa can act as proxies for each other (Barlow et al. 2007; Gardner et al. 2007). In areas of urbanisation, the presence of nesting raptors continues to attract attention as a marker of species richness (Burgas et al. 2014) but there are no clear-cut patterns emerging as yet. For example, birds and beetles do not respond similarly to increasing housing density (although this depends on whether these are forest guild or introduced species) but they do respond similarly to differences in housing densities on adjacent land (Gagné and Fahrig 2011). Increasingly, studies seem to be taking a cautious approach of examining one or two taxonomic groups with respect to land use but do not attempt to extrapolate to other taxa (Evans et al. 2009). 5.  Flagship species should be chosen using social marketing tools rather than ad hoc methods Flagship species are ‘chosen for their charisma, to increase public awareness of conservation issues and rally support for the protection of that species habitat. Protection of other species is accomplished through the umbrella effect of the flagship species’ (Favreau et al. 2006). Flagship species are used principally by non-governmental organisations, local conservation groups, zoos and governments to promote conservation awareness, raise funds, to establish a reserve or in self-promotion. Flagship species have hitherto been chosen in an ad hoc fashion with the great majority being large mammals, principally carnivores or primates, and large birds, primarily with a western audience in mind. Local people’s choice of flagship species and those of western donors often fail to coincide (Borgerhoff Mulder et al. 2009), however, and systematic stakeholder-driven approaches in choosing flagship species have recently been developed (Verissimo et al. 2014). These methods are based on

27

28

Indicators and Surrogates of Biodiversity and Environmental Change

social marketing protocols that involve identifying a conservation issue, a target audience, and characterising aspects of the campaign, marketing, evaluation and assessment (Verissimo et al. 2011). Pilot studies asking villagers to choose between species that vary in say appearance, population size, geographic distribution, visibility and whether they are commonly kept in captivity, generate flagship preferences that are sensibly tied to local context. This method makes it more likely that local stakeholders can relate to local conservation campaigns. Additionally, there is now some recognition that invertebrates and plants can act as useful flagship species. For example, colourful large butterflies, as well as dragonflies, honeybees and earthworms (species providing material benefits or ecological services), are admired in Assam, India (Barua et al. 2012), while giant squid can raise awareness of marine conservation issues (Guerra et al. 2011).

Knowledge gaps 6.  It may be important to directly measure the effects of land use change on species rather than the questionable use of other species as cross-taxon-response indicator species Certainly, the importance of monitoring the effects of land use change on biodiversity and ecosystem processes continues to be critical and is central to current debate over whether conservation effort should be targeted principally at areas altered by human activity or at pristine areas (Kareiva and Marvier 2012; Wuerthner et al. 2014). Compared with 5 years ago, great progress has been made in understanding the multiplicity of effects of defaunation on ecosystems (e.g. Dirzo et al. 2014). However, the use of cross-taxon response indicator species whose presence or population size of species may predict the response of other taxa to environmental change has not grown substantially in its wake. Rather, studies simply examine changes in populations directly using, for instance, multi-species occupancy models (e.g. Carrillo-Rubio et al. 2014). It may be better to take a few species chosen for their ecological importance or political capital to measure the biological consequences of land use change. 7.  There is value in simplifying and extending the use of landscape species to identify wildlife corridors Connectivity between protected areas is believed to promote genetic variability, to reduce demographic stochasticity, and to ‘rescue’ populations that have dwindled in one of the reserves. Yet, opportunities for providing corridors between reserves are rapidly diminishing in many parts of the world as agriculture expands and new roads block wildlife movements. Efforts to identify wildlife corridors are still in their infancy (Caro et al. 2009) but they often use single species’ movements, such as those of African Elephants (Loxodonta africana), as an untested proxy for movements of other sympatric species (but see Epps et al. 2011). De facto, these species resemble landscape species: ‘species that use large ecologically diverse areas and often have significant impacts on the structure and function of natural ecosystems. Their requirements in time and space make landscape species particularly susceptible to human alteration and use of natural landscapes’ (Sanderson et al. 2002). Landscape species are chosen formally in a computationally iterative and somewhat complex process (Coppolillo et al. 2004). This site-based conservation method should be simplified and political and publicity values should be incorporated more explicitly into the process when identifying functional corridors. This is important because public relations are vital in establishing corridors that pass through agricultural land and close to villages.

4: Conservation by proxy: thoughts 5 years on

8.  Use sentinel species to alter public opinion about global warming and establish new reserves to mollify effects of climate change Originally used as biological monitors of pollution (Beeby 2001), sentinel species are being employed to highlight global warming. To give two illustrations, Polar Bears (Ursus maritimus) are reported as being unable to find ice on which to travel in the arctic summer, and the Golden Toad (Incilius periglenes) has become extinct as a result of global warming and chytrid disease (Pounds et al. 2006). Sentinel species are also beginning to be used to highlight the inadequacies of current reserve networks to protect species under various global warming scenarios. This involves predicting species distributions after shifts in geographic ranges, combining these with habitat suitability, and pinpointing which reserves might still hold such species in the future (e.g. Cianfrani et al. 2011). Where these are inadequate, pleas are made for setting up new reserves. 9.  Reduce the use of management indicator species to assess management activities and the success of restoration projects For protected areas to become more effective, managers need to monitor the outcome of their activities and change their practices where necessary. Additionally, restoration projects (Seddon et al. 2014) that seek to renew farmland to its former wild state need to know which species to plant first, which animals to reintroduce, and which environmental regimes can return areas most faithfully to their former states. In both instances, practitioners would like to use selected species as markers of effective management and restoration success: that is, management indicator species. Unfortunately, these are chosen for many different reasons including sensitivity, or being species of special interest, species that managers hope will recover, or indicators of ecological change, each of which may have different attributes. Worryingly, they often fail to predict structural habitat characteristics and other species’ population trends. Despite their popularity, it may be better for practitioners to measure a combination of structural components and species population attributes than relying on guilds or single species as ecosystem proxies (Lindenmayer et al. 2000). 10.  Use marketing tools to advertise the importance of individual protected areas Looking to the future, I see a role for flagships being extended to national parks. Increasingly, protected areas will have to pay their way if their borders are to be respected by surrounding communities, states and governments. To generate income, they must attract visitors who will only come to a park to see a particular species found there, a particular habitat, or witness a phenomenon such as a migration or a breeding arena. Many national parks are established to represent a particular environment, such as a wetland or montane forest, but there is a danger such areas will be degazetted unless they have something ‘special’ to offer fee paying visitors. One possibility is that each reserve in a country markets itself as providing a special wildlife experience. This will involve careful thought as to how parks advertise themselves without undue overlap.

Conclusion I have discussed some recent developments in uses of surrogate species over the last 5 years, concentrating on medium-scale biodiversity indicators, umbrella species, apex predators, indicators of effects of land use change, and firming up of methods for choosing flagship species. Turning to areas where surrogate species may be able to help future conservation

29

30

Indicators and Surrogates of Biodiversity and Environmental Change

efforts, I draw attention to their utility in identifying wildlife corridors, in focusing public attention on global warming, and of making flagship protected areas to attract visitors. Yet surrogate species are not always useful. For example, I advocate measuring effects of land use change directly rather than using other species as proxies, and in measuring many ecosystem attributes when assessing management and restoration successes, not using management indicator species. More generally, while some sorts of surrogates have had their day in the sense that the goals for which they were first used have been achieved (e.g. indicators of large-scale patterns of biodiversity), and others have proven difficult to use effectively (e.g. umbrella species), many surrogate concepts are nevertheless still helpful in achieving conservation objectives. Nature is too complicated not to use species’ proxies. Moreover, the public needs natural complexities to be simplified, so surrogate species can be useful and may further the conservation agenda if employed properly and carefully.

References Barlow J, Gardner TA, Araujo AI, Avila-Pires TC, Bonaldo AB, Costa JE, et al. (2007) Quantifying the biodiversity value of tropical primary, secondary, and plantation forests. Proceedings of the National Academy of Sciences of the United States of America 104, 18555–18560. doi:10.1073/pnas.0703333104. Barua M, Gurdak DJ, Ahmed RA, Tamuly J (2012) Selecting flagships for invertebrate conservation. Biodiversity and Conservation 21, 1457–1476. doi:10.1007/s10531-012-0257-7. Beeby A (2001) What do sentinels stand for? Environmental Pollution 112, 285–298. doi:10.1016/S0269-7491(00)00038-5. Borgerhoff Mulder M, Schacht R, Caro T, Schacht J, Caro B (2009) Knowledge and attitudes of children of the Rupunini: implications for conservation in Guyana. Biological Conservation 142, 879–887. doi:10.1016/j.biocon.2008.12.021. Branton M, Richardson JS (2011) Assessing the value of the umbrella-species concept for conservation planning with meta-analysis. Conservation Biology 25, 9–20. doi:10.1111/j.1523-1739.2010.01606.x. Brooks TM, Mittermeier RA, da Fonesca GAB, Gerlach J, Hoffmann M, Lamoreux JF, et al. (2006) Global biodiversity conservation priorities. Science 313, 58–61. doi:10.1126/ science.1127609. Burgas D, Byholm P, Parkkima T (2014) Raptors as surrogates of biodiversity along a landscape gradient. Journal of Applied Ecology 51, 786–794. doi:10.1111/1365-2664.12229. Caro T (2010) Conservation by Proxy: Indicator, Umbrella, Keystone, Flagship, and Other Surrogate Species. Island Press, Washington DC. Caro T, Jones T, Davenport TRB (2009) Realities of documenting wildlife corridors in tropical countries. Biological Conservation 142, 2807–2811. doi:10.1016/j.biocon.2009.06.011. Carrillo-Rubio E, Kery M, Morreale SJ, Sullivan PJ, Gardner B, Cooch EG, et al. (2014) Use of multispecies occupancy models to evaluate the response of bird communities to forest degradation associated with logging. Conservation Biology 28, 1034–1044. doi:10.1111/ cobi.12261. Cianfrani C, Le Lay G, Maiorano L, Satizabal HF, Loy A, Guisan A (2011) Adapting global conservation strategies to climate change at the European scale: the otter as a flagship species. Biological Conservation 144, 2068–2080. doi:10.1016/j.biocon.2011.03.027. Coppolillo P, Gomez H, Maisels F, Wallace R (2004) Selection criteria for suites of landscape species as a basis for site-based conservation. Biological Conservation 115, 419–430. doi:10.1016/S0006-3207(03)00159-9.

4: Conservation by proxy: thoughts 5 years on

Cushman SA, McKelvey KS, Noon BR, McGarigal K (2010) Use of abundance of one species as a surrogate for abundance of others. Conservation Biology 24, 830–840. doi:10.1111/j.1523-1739.2009.01396.x. Dirzo R, Young HS, Galettit M, Ceballos G, Isaac NJB, Collen B (2014) Defaunation in the Anthropocene. Science 345, 401–406. doi:10.1126/science.1251817. Epps CW, Mutayoba BM, Gwin L, Brashares JS (2011) An empirical evaluation of the African elephant as a focal species for connectivity planning in East Africa. Diversity & Distributions 17, 603–612. doi:10.1111/j.1472-4642.2011.00773.x. Estes JA, Terborgh J, Brashares JS, Power ME, Berger J, Bond WJ, et al. (2011) Trophic downgrading of planet earth. Science 333, 301–306. doi:10.1126/science.1205106. Evans KL, Newson SE, Gaston KJ (2009) Habitat influences on urban avian assemblages. Ibis 151, 19–39. doi:10.1111/j.1474-919X.2008.00898.x. Favreau JM, Drew CA, Hess GR, Rubino MJ, Koch FH, Eschelbach KA (2006) Recommendations for assessing the effectiveness of surrogate species approaches. Biodiversity and Conservation 15, 3949–3969. doi:10.1007/s10531-005-2631-1. Gagné SA, Fahrig L (2011) Do birds and beetles show similar responses to urbanization? Ecological Applications 21, 2297–2312. doi:10.1890/09-1905.1. Gardner TA, Caro T, Fitzherbert E, Banda T, Lalbhai P (2007) Conservation value of multiple use areas in East Africa. Conservation Biology 21, 1516–1525. Gardner TA, Barlow J, Araujo IS, Avila-Pires TC, Bonaldo AB, Costa JE, et al. (2008) The cost-effectiveness of biodiversity surveys in tropical forests. Ecology Letters 11, 139–150. doi:10.1111/j.1461-0248.2007.01133.x. Guerra A, Gonzalez AF, Pascual S, Dawe EG (2011) The giant squid Architeuthis: an emblematic invertebrate that can represent concern for the conservation of marine biodiversity. Biological Conservation 144, 1989–1997. doi:10.1016/j.biocon.2011.04.021. Halme P, Monkkonen M, Kotiaho JS, Ylisirnio A-L, Markkanen A (2009) Quantifying the indicator power of an indicator species. Conservation Biology 23, 1008–1016. doi:10.1111/j.1523-1739.2009.01206.x. Heithaus MR, Frid A, Wirsing AJ, Worm B (2008) Predicting ecological consequences of marine top predator declines. Trends in Ecology & Evolution 23, 202–210. doi:10.1016/j. tree.2008.01.003. Hollings T, Jones M, Mooney N, McCallum H (2014) Trophic cascades following the diseaseinduced decline of an apex predator, the Tasmanian devil. Conservation Biology 28, 63–75. doi:10.1111/cobi.12152. Jorge MLSP, Galetti M, Ribeiro MC, Ferraz KMPMB (2013) Mammal defaunation as surrogate of tropic cascades in a biodiversity hotspot. Biological Conservation 163, 49–57. doi:10.1016/j.biocon.2013.04.018. Kareiva P, Marvier M (2012) What is conservation science? Bioscience 62, 962–969. doi:10.1525/bio.2012.62.11.5. Larsen FW, Bladt J, Balmford A, Rahbek C (2012) Birds as biodiversity surrogates: will supplementing birds with other taxa improve effectiveness? Journal of Applied Ecology 49, 349–356. doi:10.1111/j.1365-2664.2011.02094.x. Letnic M, Koch F, Gordon C, Crowther MS, Dickman CR (2009) Keystone effects of an alien top-predator stem extinctions of native mammals. Proceedings. Biological Sciences 276, 3249–3256. doi:10.1098/rspb.2009.0574. Lewandowski AS, Noss RF, Parsons DR (2010) The effectiveness of surrogate taxa for the representation of biodiversity. Conservation Biology 24, 1367–1377. doi:10.1111/j.1523-1739. 2010.01513.x.

31

32

Indicators and Surrogates of Biodiversity and Environmental Change

Lindenmayer DB, Margules CR, Botkin DB (2000) Indicators of biodiversity for ecologically stable forest management. Conservation Biology 14, 941–950. doi:10.1046/j.1523-1739.2000. 98533.x. Moran C, Catterall CP (2014) Responses of seed-dispersing birds to amount of rainforest in the landscape around fragments. Conservation Biology 28, 551–560. doi:10.1111/ cobi.12236. Nicholson E, Lindenmayer DB, Frank K, Possingham HP (2013) Testing the focal species approach to making conservation decisions for species persistence. Diversity & Distributions 19, 530–540. doi:10.1111/ddi.12066. Olds AD, Connolly RM, Pitt KA, Maxwell PS, Aswani S, Albert S (2014) Incorporating surrogate species and seascape connectivity to improve marine conservation outcomes. Conservation Biology 28, 982–991. doi:10.1111/cobi.12242. Peck MR, Maddock ST, Morales JN, Onate H, Mafla-Endara P, Penafiel VA, et al. (2014) Cost-effectiveness of using small vertebrates as indicators of disturbance. Conservation Biology 28, 1331–1341. doi:10.1111/cobi.12373. Phalan B, Onial M, Balmford A, Green RE (2011) Reconciling food production and biodiversity conservation: land sharing and land sparing compared. Science 333, 1289– 1291. doi:10.1126/science.1208742. Pounds JA, Bustamante MR, Coloma LA, Consuegra JA, Fogden MP, Foster PN, et al. (2006) Widespread amphibian extinctions from epidemic disease driven by global warming. Nature 439, 161–167. doi:10.1038/nature04246. Ripple WJ, Esters JA, Beschta RL, Wilmers CC, Ritchie EG, Hebblewhite M, et al. (2014) Status and ecological effects of the world’s largest carnivores. Science 34310.1126/ science.1241484 Sanderson EW, Redford KH, Vedder A, Coppolillo PB, Ward ES (2002) A conceptual model for conservation planning based on landscape species requirements. Landscape and Urban Planning 58, 41–56. doi:10.1016/S0169-2046(01)00231-6. Sattler T, Pezzatti GB, Nobis MP, Obrist MK, Roth T, Moretti M (2014) Selection of multiple umbrella species for functional and taxonomic diversity to represent urban biodiversity. Conservation Biology 28, 414–426. doi:10.1111/cobi.12213. Seddon P, Griffiths CJ, Soorae PS, Armstrong P (2014) Reversing defaunation: restoring species in a changing world. Science 345, 406–412. doi:10.1126/science.1251818. Stuart-Smith RD, Bates AE, Lefcheck JS, Duffy JE, Baker SC, Thomson RJ, et al. (2013) Integrating abundance and functional traits reveals new global hotspots of fish diversity. Nature 501, 539–542. doi:10.1038/nature12529. Terborgh J, Estes JA (2010) Trophic Cascades: Predators, Prey and the Changing Dynamics of Nature. Island Press, Washington DC. Verissimo D, MacMillan DC, Smith RJ (2011) Toward a systematic approach for identifying conservation flagships. Conservation Letters 4, 1–8. doi:10.1111/j.1755-263X.2010.00151.x. Verissimo D, Pongiluppi T, Santos MCM, Develey PF, Fraser I, Smith RJ, et al. (2014) Using a systematic approach to select flagship species for bird conservation. Conservation Biology 28, 269–277. doi:10.1111/cobi.12142. Vidal MM, Pires MM, Guimaraes PR, Jr (2013) Large predators as the missing components of seed-dispersal networks. Biological Conservation 163, 42–48. doi:10.1016/j. biocon.2013.03.025. Wuerthner G, Crist E, Butler T (2014) Keeping the Wild: Against the Domestication of Earth. Island Press, Washington DC.

5

Avian surrogates in terrestrial ecosystems: theory and practice Kathy Martin, José Tomás Ibarra and Mark Drever

Things we know 1 Surrogates should be special, but not too specialised: niche theory may help. 2 Reliable conservation surrogates can be transferrable among forest ecosystems. 3 Focusing on functional diversity can reveal underlying mechanisms for the surrogacy relationship. 4 Surrogacy relationships are dynamic. 5 Accounting for detectability increases the reliability of the surrogacy relationship. Knowledge gaps 6 How do we account for the influence of ecological structure and life history variation on surrogate–target relationships? 7 How can citizen science initiatives provide opportunities for testing and validating surrogacy relationships? 8 When should we abandon conservation surrogates in favour of direct measurements? 9 Can we use artificial habitat attributes as surrogates? 10 Can we use anti-surrogates to measure loss or recovery of biodiversity composition or function?

Introduction Terrestrial realms comprise 29% of the Earth’s land surface and support most of the studied biodiversity. Forests comprise 30% of the global landmass, ~22% of which remain as natural primary forests, and only 5% of primary forest is located in protected areas (Mackey et al. 2015). For primary forests, the conservation questions often focus on which species are present and what species/services might be lost with forest degradation and loss. About 65% of global forest lands have been harvested and, since most are regenerating naturally (7% in plantations; FAO 2010), the biodiversity conservation and management questions usually relate to species persistence, and whether, and at which stages, species 33

34

Indicators and Surrogates of Biodiversity and Environmental Change

groups and ecological processes are lost or may recover. Continued investigations of biodiversity trends have resulted in urgent calls to action, given concerns that we are experiencing a global wave of anthropogenically driven biodiversity loss with species and population extirpations and steep declines in local species abundance (Dirzo et al. 2014). While the problems of conserving biodiversity in terrestrial realms can be described and prioritised at a global level (Brooks et al. 2006), generally conservation action and management is undertaken within geopolitical boundaries, and thus reliable surrogates and indicators are needed at operational scales. The complexity of environmental problems, such as habitat degradation and loss, have led to the development of surrogates that can be used to identify and track the status and trends of species richness in ecological communities (Caro and O’Doherty 1999; Caro 2010). However, conservation focused on taxonomic diversity alone will not reliably maintain critical ecosystem processes (Díaz and Cabido 2001), especially in areas with high rates of endemism in biodiversity hotspots (Kerr 1997). Thus, surrogates are required that represent aspects of functional diversity given that an understanding of the mechanisms underpinning surrogacy relationships is a key component for improving the efficacy of the surrogate species approach (Lindenmayer and Likens 2011; Sattler et al. 2014). Efficient conservation surrogates should be: (1) relevant to either ecologically important or conservation important phenomena; (2) sufficiently sensitive, but not too sensitive, to offer a warning of anthropogenic habitat disturbance; (3) able to provide an estimate of the status of target biodiversity across wide environmental gradients, (4) easy and cost-effective to survey; and (5) distributed over a broad geographical area (Noss 1990). Before the use of these surrogates is deployed in the management of ecosystems, careful tests of their reliability are needed to determine the strength of the surrogates and the extent of their applicability across time and space.

1.  Surrogates should be special, but not too specialised: niche theory may help Ecological specialisation can be viewed as the restricted niche width of a species (Futuyma and Moreno 1988). Niche models have been used for assessing the condition of communities (Hirzel and Le Lay 2008; Wiens et al. 2010; Clavel et al. 2011; Ibarra et al. 2014b), and thus can be deployed to identify biodiversity surrogates. Habitat-specialists, relative to habitat-generalists, are more likely to be sensitive to habitat degradation and fragmentation, and thus have been recommended as ideal surrogates (Pearson 1994; Cabeza et al. 2008). Meeting the needs of habitat-specialists is expected to also provide for the requirements of generalist species, unless the surrogate species is so highly specialised it occupies a very narrow niche. Few studies have tested the importance of being a specialist as a key criterion for putative surrogates (but see Ibarra 2014). Birds may be suitable surrogates because their ecology is well understood, they respond to habitat changes at multiple spatial scales, they include a wide range of trophic guilds, and standardised monitoring protocols have been developed (Carignan and Villard 2002; Padoa-Schioppa et al. 2006). Recent advances suggest avian surrogates including single species, groups of related species, or functional guilds (e.g. woodpeckers, top predators and common birds) can be used to monitor avian or vertebrate responses to habitat disturbance and environmental change. Drever et al. (2008) found that woodpeckers were a reliable surrogate guild for all birds in a range of mixed deciduous and coniferous forest site types that experienced significant habitat change (harvest, fire and insect outbreaks), with the exception of a brief period during the peak of the bark beetle outbreak where wood-

5: Avian surrogates in terrestrial ecosystems: theory and practice

pecker abundance continued to increase while density of all other birds decreased. Studies in Europe have found that one or several raptor species were reliable surrogates of overall species richness and richness of vulnerable species in forests (Sergio et al. 2006; Burgas et al. 2014). Common species make significant contributions to biomass and often to ecosystem function (Gaston and Fuller 2008); thus, common species may be good surrogates for the richness of, or changes to, systems (Gaston 2008). Broad-scale monitoring programs often have greater statistical power to monitor common species than rare species. Empirical studies in British Columbia tested the efficacy of common birds as indicators of all forest land bird responses to habitat disturbance, and found that densities of common species corresponded more strongly with changes in total bird density and overall species richness than rare species (Koch et al. 2011). These patterns were non-linear and species with intermediate–high commonness showed similar or better correspondence than the most abundant species. In summary, studies of terrestrial systems have identified reliable avian surrogates that include single and multiple species, and species from top and bottom trophic levels, as well as several functional guilds. Furthermore, several of the most reliable guilds (woodpeckers, owls and common species) perform well as surrogates when the target areas are scaled up from the stand level (15–25 ha) to landscapes (100 ha to 10 000 ha), and the relationships hold over several continents (Gaston and Williams 1993; Mikusiński et al. 2001; Sergio et al. 2004; Sergio et al. 2006; Roberge et al. 2008; Koch et al. 2011; Burgas et al. 2014).

2.  Reliable conservation surrogates can be transferrable among forest ecosystems The reliability of potential surrogates is typically assessed at local scales, and it is often unknown whether such surrogate species have transferrable analogues in other ecosystems. The need for transferrable and reliable surrogates is especially important for ecosystems that have been classified as having high vulnerability (usually related to time-sensitive conservation options) or high irreplaceability (e.g. faunal or floral endemism in Atlantic subtropical or tropical forest ecosystems; Buchanan et al. 2011). In these systems, there is a strong imperative to proceed with the use of surrogates for biodiversity given the need for urgent conservation action. Thus, identifying general traits of surrogate species across a variety of ecosystems remains an important research question. Ideally these traits should be based on general ecological theory (e.g. trophic level) and could identify a key mechanism for the surrogacy relationship. For example, similar raptor species have been identified as surrogates in a variety of forest ecosystems (e.g. Strix owl species in forests of western North America, northern Europe and southern Chile), indicating the potential for such generalised surrogates to exist (Noon and McKelvey 1996; Burgas et al. 2014; Ibarra 2014). 3.  Focusing on functional diversity can reveal underlying mechanisms for the surrogacy relationship Studies assessing the impacts of habitat degradation and loss often assume that these disturbances reduce taxonomic diversity or species richness, resulting in similar losses of ecosystem functioning and stability (Milder et al. 2008). This assumption, however, exceeds both empirical and theoretical validation (Schwartz et al. 2000; Mayfield et al. 2010). Research on surrogacy relationships has chiefly explored whether the occurrence of surrogates is indicative of high taxonomic diversity. However, functional diversity (defined as the value and range of functional traits – phenological, behavioural, physiological or mor-

35

36

Indicators and Surrogates of Biodiversity and Environmental Change

Fig. 5.1.  Example of a surrogate–target biodiversity validation study for the Andean temperate forests of southern Chile. A field study of forest birds was designed to test the reliability of specialist and generalist forest owls to monitor taxonomic diversity (species richness) and several measures of functional diversity (e.g. large tree users, vertical profile users, bamboo understory users, shrub users) across forest conditions ranging from old growth stands to highly degraded forests. Before testing for surrogacy reliability, occurrence estimates (ψ) from the stand to the landscape level for the potential surrogates (owls), and the density estimates for the target biodiversity (21 other bird species) were adjusted for detectability (p). The habitat specialist owl was determined to be a reliable surrogate of both taxonomic and functional diversity in Andean temperate forests, whereas the generalist owl was an unreliable surrogate (Ibarra 2014).

phological – present in a community) is now considered to be an important metric for linking diversity with ecosystem processes and stability, and thus should be considered when identifying surrogate candidates (Fig. 5.1; Ibarra 2014; Trivellone et al. 2014; Sattler et al. 2014). Assessment using surrogates for functional diversity may provide further insights into the effects of such anthropogenic impacts on the services that humans derive from ecosystems (Chapin et al. 2000).

4.  Surrogacy relationships are dynamic The surrogate concept was developed for assessing (and tackling) environmental change and biodiversity loss (Lindenmayer and Likens 2011). Limited research has been done on the long-term resiliency of surrogacy relationships that can inform strategies for enhancing the adaptive capacity of ecosystems (Filotas et al. 2014). In the short term, a single species or habitat attribute can show high congruence with target biodiversity. However, because terrestrial ecosystems are characterised by their heterogeneity, non-linearity and uncertainty (Filotas et al. 2014), a congruency relationship may change over longer time

5: Avian surrogates in terrestrial ecosystems: theory and practice

periods (Thomson et al. 2005; Hess et al. 2006; Westgate et al. 2014). For example, Lindenmayer et al. (2014) reported the relationship between the abundance of hollow-bearing trees (surrogates) and arboreal cavity-dwelling marsupials (target biodiversity) remained positive over a period of 30 years. However, the decline in abundance of hollow-bearing trees over time, which promoted either a behavioural shift among animals to such changes in nesting resources or an alteration of patch- and landscape-level population dynamics, weakened the surrogacy relationship. In some cases, the most appropriate surrogate may change over time. In a commensal community network of cavity-using vertebrates that experienced an intense insect outbreak, the keystone species in the excavator guild in the nest web changed from Northern Flickers (Colaptes auratus) before the outbreak, to a broader suite of generally less common woodpeckers after the peak of the bark beetle outbreak (Cockle and Martin 2015). For longer term programs with appropriate data, we recommend using a Species Accumulation Index as a general approach to test the strength and endurance of the surrogate–target relationship in terrestrial realms (cross-taxon congruence; Westgate et al. 2014). Thus, we recommend some effort be devoted to testing the strength of the surrogate–target relationship over time and space.

5.  Accounting for detectability increases the reliability of the surrogacy relationship Frequently researchers assume that the species of interest – either the surrogate candidate or target biodiversity – are absent from a site when they were not detected. The probability of detecting both the surrogate candidate (e.g. top predators) and target biodiversity depends on many factors, including wind speed, temperature, and date and time of survey (Ibarra et al. 2014a). Thus, the failure to detect individuals at a site does not mean they were absent unless detection probability is perfect (MacKenzie et al. 2006). The assumption of perfect detectability can result in misleading inferences about the consistency of cross-taxon congruence and, therefore, of the strength of the biodiversity–surrogacy relationships. Advances have been made in the analyses of site-occurrence and species richness data that enable accounting for imperfect detection, along with site- and survey-specific covariates, to assess a range of ecological and conservation-related questions at both the site and landscape scales (Noon et al. 2012). Even in poorly known systems, surrogacy studies need clear objectives that are explicitly linked to science or management questions, and should be designed in advance so that the probability of detecting the species in a survey can be estimated (MacKenzie et al. 2006; Moilanen 2002; Tyre et al. 2003; but see Welsh et al. 2013). Examples of models that can adjust for detectability include, among others, occupancy (MacKenzie et al. 2006), N-mixture (Royle 2004), multinomial–Poisson mixture (Royle et al. 2004), and conventional and multiple covariate distance sampling (Buckland et al. 2001; Marques et al. 2007). Of course, detectability is only one of numerous issues to consider when designing and refining surrogate-target studies, albeit it is an important one.

Knowledge gaps 6.  How do we account for the influence of ecological structure and life history variation on surrogate-target relationships? Community assemblages are structured by trophic levels and nest site types, in addition to other ecological processes such as direct and indirect competition or predation, which influence population persistence and abundance. Therefore, an ecological understanding of the system and knowledge of resilience of taxa to environmental or habitat change is

37

38

Indicators and Surrogates of Biodiversity and Environmental Change

necessary for validating the use of surrogates, given that resilience at different trophic or nidic levels may reduce the signal of responses to habitat change differently between the surrogate and target biodiversity. For example, plants may respond to habitat change or climate change more dramatically than the next highest trophic level, because herbivores or insectivores may be able to use altered habitat components differently to mitigate their responses (Gauthier et al. 2013). Thus, knowledge of community structure (nest webs and food webs) and linkages may facilitate identification of the most appropriate groups to choose as the surrogate and target biodiversity. Life history variation also can influence species responses to habitat change and environmental variation. Two congeneric sympatric ptarmigan species responded differently to severe environmental conditions in northern Canada, with the ‘slow life style’ Rock Ptarmigan (Lagopus muta) reducing its reproductive effort strongly during cold late breeding seasons while the sympatric ‘fast life style’ White-tailed Ptarmigan (L. leucura saxatalis), showed strong reproductive resilience to the same conditions (Wilson and Martin 2012). Finally, the losses and gains in ecosystems will not be symmetrical, due to differences in the resilience of functional guilds as well as the replacement of native and rare species by exotic species and common species. Overall, a lack of ecological knowledge for both the surrogate and target biodiversity greatly increases the uncertainty in adopting surrogate approaches to conservation problems.

7.  How can citizen science initiatives provide opportunities for testing and validating surrogacy relationships? Increasingly, citizen science data such as ebird (http://ebird.org/content/ebird/) or other publically accessible web-based data submission programs are being used to generate species distributions at a range of scales in response to various landscape changes (Wood et al. 2011). Many of these programs sample a wide diversity of species and thus provide opportunities to identify species or suites of species that can serve as reliable surrogates. Citizen science data provided estimates of suitable habitat comparable to intensively collected field observation data in a study using several modelling approaches to predict supply of coastal alpine habitat for the Vancouver Island White-tailed Ptarmigan, an alpine specialist bird (Jackson et al. 2015). Earlier in the study, it was confirmed that the distribution of this ptarmigan had not decreased by using historical records submitted by hikers and contemporary field surveys of these sites (Martin et al. 2004). The collection of publicly submitted records of species detections has tremendous potential to describe and predict changes in species or communities that have undergone spatial or temporal changes. The development of machine learning models and methods to determine pseudo-absences enables the reliable use of such surrogate data by ecologists and conservation biologists. In addition, these data can be collected economically and efficiently in areas where limited resources are available to dedicate expert field professionals to data collection. 8.  When should we abandon conservation surrogates in favour of direct measurements? Lindenmayer and Likens (2011) proposed a critical evaluation of the use of surrogates for conservation, and suggested that direct measurement of the particular entities of environmental or conservation interest will be the best option in some circumstances (see Chapter 4). Given the typical complexity and high diversity of terrestrial ecosystems, it is unclear when direct measurement may be the best option. Direct measures may serve in simplified

5: Avian surrogates in terrestrial ecosystems: theory and practice

forests such as plantations, where particular stand attributes can be carefully controlled and may themselves be the management goals (e.g. vertical stand structure variables; Humphrey et al. 1999). Recent evidence suggests habitat-based surrogates (the structure of vegetation) can serve equally or better than species-based surrogates (Lindenmayer et al. 2014). These habitat-based surrogates, such as trees in large-diameter or older age-classes, may be more amenable to apply than species-based surrogates to the prescriptive nature of forest management operations, and may thus provide a better alternative than direct measures. However, the use of direct measures may also require validation to ensure methodological issues do not undermine their usefulness. For example, surveys of the number of cavities for opossums, or stick nest surveys for raptors and waterbirds, may not provide accurate estimates of attribute availability or use. Cockle et al. (2010) found that only 19% of the tree cavities identified from ground-based surveys in primary Atlantic forest stands were in fact usable for nesting and roosting. Thus, conducting visual checks of tree cavities or direct counts of cavity-using vertebrates such as opossums, gliders or raptors may in general be more useful than conducting ground-based counts of cavities in the forest without some preliminary ecological knowledge of the cavity-using vertebrate community in that system.

9.  Can we use artificial habitat attributes as surrogates? Artificial habitat attributes are sometimes chosen as proxy attributes to estimate resource availability or demographic rates of target species (e.g. surveys of nest boxes, or platform nests for raptors and large waterbirds). Nest boxes are often used as a proxy habitat attribute for tree cavities as they are convenient for monitoring demography for cavity-nesting birds. However, if vital rates differ for birds using nest boxes and natural tree nests, it may be problematic to rely on demographic data from nest boxes to accurately diagnose the factors most responsible for the population declines. Most of the demographic data collected for the Tree Swallow (Tachycineta bicolor) – a species showing population declines across North America – comes from nest box studies, but birds nesting in boxes have larger clutches and higher nesting success than those in tree cavities (Shutler et al. 2012). Therefore, careful validation is necessary when using artificial structures to monitor biodiversity trends in ecosystems with multi-annual resources that vary in time and space and across successional stages. 10.  Can we use anti-surrogates to measure loss or recovery of biodiversity composition or function? The presence of certain species or groups of species (e.g. generalists, pests and invasive species) may be used to indicate reliably the loss of habitat function or biodiversity. For example, the proportion of conifer species in mixed deciduous stands was negatively associated with all bird richness (target) and woodpecker abundance (the surrogate group, Drever et al. 2008). In some cases, it may be easier to establish the statistical reliability of a target–surrogate relationship over a gradient of environmental change with a dominant habitat type or condition that supports low biodiversity levels (e.g. conifers) rather than the more rare, but valuable, habitat attribute (in this case, deciduous trees, especially Trembling Aspen Populus tremuloides). It is likely easier to scale up this relationship of forest types or conditions with the abundant anti-surrogate in using remote sensing, Landsat imagery, or other landscape level habitat data. In other systems, certain invasive species may serve as reliable anti-surrogates if their presence or impact increases when habitat conditions reach a certain level of degradation, or their exit from the system could

39

40

Indicators and Surrogates of Biodiversity and Environmental Change

indicate a certain stage of recovery. For example, cavity-excavating woodpeckers and secondary cavity-nesting passerines readily co-exist with the dominant European Starling (Sturnus vulgaris) when there is an abundant supply of tree cavities (Koch et al. 2012). However, when too many suitable trees for excavation have been removed or the most suitable forest types or nesting trees have reached a critically low abundance, starlings and sometimes other larger woodpeckers may usurp cavities of endangered woodpeckers (Blanc and Walters 2008). To ensure the utility of the anti-surrogate concept, it would be necessary to be able to distinguish between unreliable surrogates and those having antisurrogate properties.

Conclusion Given the complexity of most terrestrial ecosystems, the use of surrogates will have enduring appeal as a way to simplify and rationalise management objectives that attempt to maintain biodiversity under economic incentives to extract resources from ecosystems. Biodiversity surrogates have been criticised because generally they do not provide much insight into changes in ecosystem function. However, several studies using avian surrogates have made considerable progress in developing functional indicators of biodiversity responses across time, space and taxa. Therefore, the use of avian surrogates for biodiversity may be most successful when linked to an ecological mechanism, such as ‘bottom-up’ purveyance of resources (e.g. habitat features or tree species associated with structural diversity, or woodpeckers as primary ‘producers’ of cavities for other species) or ‘top-down’ control (e.g. presence and abundance of top predators such as owls, wolves or bears). Such theory also may be useful in identifying the target measures of biodiversity. More work is needed to rationalise what surrogates are meant to represent within the context of ecosystem management, and ensuring that their use contributes to diverse systems that are resilient to natural and anthropogenic disturbances. This research will most likely involve the use of ecological data collected over long time series and that span large spatial scales, such as data made available through citizen science initiatives.

Acknowledgements A variety of agencies funded this research, including support from Environment Canada, and the Natural Sciences and Engineering Research Council of Canada (NSERC Discovery Grants, Special Strategic and Northern Research Supplements, Postdoctoral Fellowship) to K. Martin and M. Drever. D. B. Lindenmayer, J. Pierson and P. Barton, and anonymous reviewers provided valuable comments on an earlier draft of this manuscript. J. T. Ibarra received funding support from Rufford Small Grants Foundation and The Peregrine Fund. He also acknowledges the post-graduate scholarship from Comisión Nacional de Investigación Científica y Tecnológica (CONICYT).

References Blanc LA, Walters JR (2008) Cavity excavation and enlargement as mechanisms for indirect interactions in an avian community. Ecology 89, 506–514. doi:10.1890/07-0219.1. Brooks TM, Mittermeier RA, da Fonseca GAB, Gerlach J, Hoffmann M, Lamoreux JF, et al. (2006) Global biodiversity conservation priorities. Science 313, 58–61. doi:10.1126/ science.1127609.

5: Avian surrogates in terrestrial ecosystems: theory and practice

Buchanan GM, Donald PF, Butchart SHM (2011) Identifying priority areas for conservation: a global assessment for forest-dependent birds. PLoS ONE 6, e29080. doi:10.1371/journal. pone.0029080. Buckland ST, Anderson DR, Burnham KP, Laake JL, Borchers DL, Thomas L (2001) Introduction to Distance Sampling: Estimating Abundance of Biological Populations. Oxford University Press, Oxford, UK. Burgas D, Byholm P, Parkkima T (2014) Raptors as surrogates of biodiversity along a landscape gradient. Journal of Applied Ecology 51, 786–794. Cabeza M, Arponen A, Van Teeffelen A (2008) Top predators: hot or not? A call for systematic assessment of biodiversity surrogates. Journal of Applied Ecology 45, 976–980. doi:10.1111/ j.1365-2664.2007.01364.x. Carignan V, Villard M-A (2002) Selecting indicator species to monitor ecological integrity: a review. Environmental Monitoring and Assessment 78, 45–61. doi:10.1023/A: 1016136723584. Caro T (2010) Conservation by Proxy: Indicator, Umbrella, Keystone, Flagship, and Other Surrogate Species. Island Press, Washington DC. Caro T, O’Doherty G (1999) On the use of surrogate species in conservation biology. Conservation Biology 13, 805–814. doi:10.1046/j.1523-1739.1999.98338.x. Chapin FSI, Zavaleta ES, Eviner VT, Naylor RL, Vitousek PM, Reynolds HL, et al. (2000) Consequences of changing biodiversity. Nature 405, 234–242. doi:10.1038/35012241. Clavel J, Julliard R, Devictor V (2011) Worldwide decline of specialist species: toward a global functional homogenization? Frontiers in Ecology and the Environment 9, 222–228. doi:10.1890/080216. Cockle KL, Martin K (2015) Temporal dynamics of a commensal network of cavity-nesting vertebrates: increased diversity during an insect outbreak. Ecology 96, 1093–1104. Cockle KL, Martin K, Drever MC (2010) Supply of tree-holes limits nest density of cavitynesting birds in primary and logged subtropical Atlantic forest. Biological Conservation 143, 2851–2857. doi:10.1016/j.biocon.2010.08.002. Díaz S, Cabido M (2001) Vive la différence: plant functional diversity matters to ecosystem processes. Trends in Ecology & Evolution 16, 646–655. doi:10.1016/S0169-5347(01)02283-2. Dirzo R, Young HS, Galetti M, Ceballos G, Isaac NJB, Collen B (2014) Defaunation in the Anthropocene. Science 345, 401–406. doi:10.1126/science.1251817. Drever M, Aitken K, Norris A, Martin K (2008) Woodpeckers as reliable indicators of bird richness, forest health and harvest. Biological Conservation 141, 624–634. doi:10.1016/j. biocon.2007.12.004. FAO (Food and Agriculture Organization of the United Nations) (2010) Key Findings: Global Forest Resources Assessment. FAO, Rome, Filotas E, Parrott L, Burton PJ, Chazdon RL, Coates KD, Coll L, et al. (2014) Viewing forests through the lens of complex systems science. Ecosphere 5, 1–23. doi:10.1890/ES13-00182.1. Futuyma DJ, Moreno G (1988) The evolution of ecological specialization. Annual Review of Ecology and Systematics 19, 207–233. doi:10.1146/annurev.es.19.110188.001231. Gaston KJ (2008) Biodiversity and extinction: the importance of being common. Progress in Physical Geography 32, 73–79. doi:10.1177/0309133308089499. Gaston KJ, Fuller RA (2008) Commonness, population depletion and conservation biology. Trends in Ecology & Evolution 23, 14–19. doi:10.1016/j.tree.2007.11.001. Gaston KJ, Williams PH (1993) Mapping the world’s species – the higher taxon approach. Biodiversity Letters 1, 2–8. doi:10.2307/2999642.

41

42

Indicators and Surrogates of Biodiversity and Environmental Change

Gauthier G, Bêty J, Cadieux M-C, Legagneux P, Doiron M, Chevallier C, et al. (2013) Longterm monitoring at multiple trophic levels suggests heterogeneity in responses to climate change in the Canadian Arctic tundra. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 368, 20120482. doi:10.1098/rstb.2012.0482. Hess GR, Bartel Ra, Leidner AK, Rosenfeld KM, Rubino MJ, Snider SB, et al. (2006) Effectiveness of biodiversity indicators varies with extent, grain, and region. Biological Conservation 132, 448–457. doi:10.1016/j.biocon.2006.04.037. Hirzel AH, Le Lay G (2008) Habitat suitability modelling and niche theory. Journal of Applied Ecology 45, 1372–1381. doi:10.1111/j.1365-2664.2008.01524.x. Humphrey J, Hawes C, Peace AJ, Ferris-Kaan R, Jukes M (1999) Relationships between insect diversity and habitat characteristics in plantation forests. Forest Ecology and Management 113, 11–21. doi:10.1016/S0378-1127(98)00413-7. Ibarra JT (2014) Andean temperate forest owls: detectability, habitat relationships and reliability as biodiversity surrogates. PhD thesis. Faculty of Forestry, University of British Columbia, Canada. Ibarra JT, Martin K, Altamirano TA, Vargas FH, Bonacic C (2014a) Factors associated with the detectability of owls in South American temperate forests: implications for nocturnal raptor monitoring. The Journal of Wildlife Management 78, 1078–1086. doi:10.1002/ jwmg.740. Ibarra JT, Martin K, Drever MC, Vergara G (2014b) Occurrence patterns and niche relationships of sympatric owls in South American temperate forests: a multi-scale approach. Forest Ecology and Management 331, 281–291. doi:10.1016/j.foreco.2014.08.027. Jackson MM, Gergel SE, Martin K (2015) Citizen science and field survey observations provide comparable results for mapping alpine species distributions. Biological Conservation 181, 162–172. Kerr J (1997) Species richness, endemism, and the choice of areas for conservation. Conservation Biology 11, 1094–1100. doi:10.1046/j.1523-1739.1997.96089.x. Koch AJ, Drever MC, Martin K (2011) The efficacy of common species as indicators: avian responses to disturbance in British Columbia, Canada. Biodiversity and Conservation 20, 3555–3575. doi:10.1007/s10531-011-0148-3. Koch AJ, Martin K, Aitken KEH (2012) The relationship between introduced European starlings and the reproductive activities of mountain bluebirds and tree swallows in British Columbia, Canada. The Ibis 154, 590–600. doi:10.1111/j.1474-919X.2012.01242.x. Lindenmayer DB, Likens GE (2011) Direct measurement versus surrogate indicator species for evaluating environmental change and biodiversity loss. Ecosystems 14, 47–59. doi:10.1007/ s10021-010-9394-6. Lindenmayer DB, Barton PS, Lane PW, Westgate MJ, McBurney L, Blair D, et al. (2014) An empirical assessment and comparison of species-based and habitat-based surrogates: a case study of forest vertebrates and large old trees. PLoS ONE 9, e89807. doi:10.1371/ journal.pone.0089807. MacKenzie DI, Nichols JD, Royle JA, Pollock KH, Bailey LL, Hines JE (2006) Occupancy Estimation and Modeling: Inferring Patterns and Dynamics of Species Occurrence. Academic Press, London, UK. Mackey B, DellaSala DA, Kormos C, Lindenmayer D, Kumpel N, Zimmerman B, et al. (2015) Policy options for the world’s primary forests in multilateral environmental agreements. Conservation Letters 8, 139–147. doi:10.1111/conl.12120.

5: Avian surrogates in terrestrial ecosystems: theory and practice

Marques TA, Thomas L, Fancy SG, Buckland ST (2007) Improving estimates of bird density using multiple-covariate distance sampling. The Auk 124, 1229–1243. doi:10.1642/00048038(2007)124[1229:IEOBDU]2.0.CO;2. Martin K, Brown GA, Young JR (2004) The historic and current distribution of the Vancouver Island white-tailed ptarmigan (Lagopus leucurus saxatilis). Journal of Field Ornithology 75, 239–256. Mayfield MM, Bonser SP, Morgan JW, Aubin I, McNamara S, Vesk Pa (2010) What does species richness tell us about functional trait diversity? Predictions and evidence for responses of species and functional trait diversity to land-use change. Global Ecology and Biogeography 19, 423–431. Mikusiński G, Gromadzki M, Chylarecki P (2001) Woodpeckers as indicators of forest bird diversity. Conservation Biology 15, 208–217. doi:10.1046/j.1523-1739.2001.99236.x. Milder JC, Lassoie JP, Bedford BL (2008) Conserving biodiversity and ecosystem function through limited development: an empirical evaluation. Conservation Biology 22, 70–79. doi:10.1111/j.1523-1739.2007.00812.x. Moilanen A (2002) Implications of empirical data quality for metapopulation model parameter estimation and application. Oikos 96, 516–530. doi:10.1034/j.1600-0706.2002. 960313.x. Noon BR, McKelvey KS (1996) Management of the spotted owl: a case history in conservation biology. Annual Review of Ecology and Systematics 27, 135–162. doi:10.1146/annurev. ecolsys.27.1.135. Noon BR, Bailey LL, Sisk TD, McKelvey KS (2012) Efficient species-level monitoring at the landscape scale. Conservation Biology 26, 432–441. doi:10.1111/j.1523-1739.2012.01855.x. Noss RF (1990) Indicators for monitoring biodiversity: a hierarchical approach. Conservation Biology 4, 355–364. doi:10.1111/j.1523-1739.1990.tb00309.x. Padoa-Schioppa E, Baietto M, Massa R, Bottoni L (2006) Bird communities as bioindicators: the focal species concept in agricultural landscapes. Ecological Indicators 6, 83–93. doi:10.1016/j.ecolind.2005.08.006. Pearson DL (1994) Selecting indicator taxa for the quantitative assessment of biodiversity. Philosophical Transactions of the Royal Society of London 345, 75–79. doi:10.1098/ rstb.1994.0088. Roberge J-M, Mikusiński G, Svensson S (2008) The white-backed woodpecker: umbrella species for forest conservation planning? Biodiversity and Conservation 17, 2479–2494. doi:10.1007/s10531-008-9394-4. Royle JA (2004) N-mixture models for estimating population size from spatially replicated counts. Biometrics 60, 108–115. doi:10.1111/j.0006-341X.2004.00142.x. Royle JA, Dawson DK, Bates S (2004) Modeling abundance effects in distance sampling. Ecology 85, 1591–1597. doi:10.1890/03-3127. Sattler T, Pezzatti GB, Nobis MP, Obrist MK, Roth T, Moretti M (2014) Selection of multiple umbrella species for functional and taxonomic diversity to represent urban biodiversity. Conservation Biology 28, 414–426. doi:10.1111/cobi.12213. Schwartz MW, Brigham CA, Hoeksema JD, Lyons KG, Mills MH, van Mantgem PJ (2000) Linking biodiversity to ecosystem function: implications for conservation ecology. Oecologia 122, 297–305. doi:10.1007/s004420050035. Sergio F, Marchesi L, Pedrini P (2004) Integrating individual habitat choices and regional distribution of a biodiversity indicator and top predator. Journal of Biogeography 31, 619–628. doi:10.1046/j.1365-2699.2003.01002.x.

43

44

Indicators and Surrogates of Biodiversity and Environmental Change

Sergio F, Newton I, Marchesi L, Pedrini P (2006) Ecologically justified charisma: preservation of top predators delivers biodiversity conservation. Journal of Applied Ecology 43, 1049– 1055. doi:10.1111/j.1365-2664.2006.01218.x. Shutler D, Hussell DJT, Norris DR, Winkler DW, Robertson RJ, Bonier F, et al. (2012) Spatiotemporal patterns in nest box occupancy by tree swallows across North America. Avian Conservation and Ecology 7, 3. doi:10.5751/ACE-00517-070103. Thomson J, Fleishman E, Macnally R, Dobkin D (2005) Influence of the temporal resolution of data on the success of indicator species models of species richness across multiple taxonomic groups. Biological Conservation 124, 503–518. doi:10.1016/j.biocon.2005.02.013. Trivellone V, Schoenenberger N, Bellosi B, Jermini M, de Bello F, Mitchell EAD, et al. (2014) Indicators for taxonomic and functional aspects of biodiversity in the vineyard agroecosystem of Southern Switzerland. Biological Conservation 170, 103–109. doi:10.1016/j.biocon.2013.12.008. Tyre AJ, Tenhumberg B, Field SA, Niejalke D, Parris K, Possingham HP (2003) Improving precision and reducing bias in biological surveys by estimating false negative error rates in presence–absence data. Ecological Applications 13, 1790–1801. doi:10.1890/02-5078. Welsh AH, Lindenmayer DB, Donnelly CF (2013) Fitting and interpreting occupancy models. PLoS ONE 8, e52015. doi:10.1371/journal.pone.0052015. Westgate MJ, Barton PS, Lane PW, Lindenmayer DB (2014) Global meta-analysis reveals low consistency of biodiversity congruence relationships. Nature Communications 5, 1–8. doi:10.1038/ncomms4899. Wiens JJ, Ackerly DD, Allen AP, Anacker BL, Buckley LB, Cornell HV, et al. (2010) Niche conservatism as an emerging principle in ecology and conservation biology. Ecology Letters 13, 1310–1324. doi:10.1111/j.1461-0248.2010.01515.x. Wilson S, Martin K (2012) Influence of life history strategies on sensitivity, population growth and response to climate for sympatric alpine birds. BMC Ecology 12, 9. doi:10.1186/1472-6785-12-9. Wood C, Sullivan B, Iliff M, Fink D, Kelling S (2011) eBird: engaging birders in science and conservation. PLoS Biology 9, e1001220.

6

Using decision theory to select indicators for managing threats to biodiversity Ayesha Tulloch

Things we know 1 We need clear objectives related to quantifiable monitoring and management outcomes. 2 Thinking about complementarity avoids redundancies and reduces uncertainty. 3 Considering monitoring costs and indicator cost-effectiveness is crucial. 4 Choosing indicators necessitates trade-offs in representativeness, certainty in outcomes and costs. 5 Strategic decision-theoretic approaches help make difficult choices by meeting explicit objectives under uncertainty. Knowledge gaps 6 Current indicator selection methods do not evaluate indicators in terms of their ability to reduce uncertainty 7 Most of our systems for selecting indicators assume we have prior information on how they respond to threats and actions 8 Indicators can represent actions, threats or ultimate biodiversity outcomes – what are the consequences of not knowing all of these parts of the management cycle? 9 Indicators are rarely chosen with more than one threat and mitigating action in mind – how is it best to account for the complexities of interacting threats? 10 Indicators are about accepting risk – can we use risk analysis to assess what the consequences might be if the indicator is wrong?

Introduction In an ideal world, with unlimited resources, we could monitor or track the progress of anything we might be interested in, from changes in the weather, to responses of declining populations of species of conservation concern, to threat mitigation actions. However, it is rarely possible to measure everything. Not only are budgets limited, but the ecology and 45

46

Indicators and Surrogates of Biodiversity and Environmental Change

behaviour of many species make them hard to detect or difficult to measure. Furthermore, natural systems have so many physical processes, biological interactions and species that it is impossible to keep track of them all. In cases when the states or trends of environmental phenomena are too difficult, inconvenient, or expensive to measure directly, indicators provide simple cost-effective measurable proxies. Indicators help us communicate the state of the environment or the effectiveness of management actions to policy makers and other stakeholders, by providing us with a summary of what is going on in our world. In biodiversity conservation, indicators are useful for three things: (1) tracking performance (evidence-based management); (2) discriminating among competing hypotheses (scientific exploration); and (3) discriminating among alternative policies (evaluation) (Failing and Gregory 2003). The use of indicators is not new. They have been proposed and used at local, national and international scales (see examples in Table 6.1). A classic regional example is the use of the Northern Spotted Owl (Strix occidentalis) in north-west USA as a management indicator species to measure the effect of old-growth logging on lower trophic levels (Dawson et al. 1987). At a national scale, indicator frameworks include New Zealand’s core national environmental indicators to estimate the state of the environment (Ministry for the Environment 2007) and Canada’s use of the Woodland Caribou (Rangifer tarandus) to indicate the health of forests (Wedeles et al. 2014). Multi-national indicators include: the Global Wild Bird Index developed by the non-government conservation organisations BirdLife and the Royal Society for the Preservation of Birds (RSPB); the 10 climate change flagship species developed by the International Union for Conservation of Nature (IUCN) (Foden and Stuart 2009); and the aggregate set of 55 indicators recently evaluated to represent progress towards the global Aichi Targets that aim to reduce, and eventually halt, the loss of biodiversity (Tittensor et al. 2014). Indicators can be common species, species under threat (Foden and Stuart 2009), threats or pressures themselves (e.g. invasive alien species in Europe: European Environment Agency 2010), environmental attributes (Niemeijer and de Groot 2008), or policy responses and economic impacts of activities. In conservation applications, the concept of indicators has been used to solve problems of where to locate protected areas in systematic conservation planning (Rodrigues and Brooks 2007), or how the environment is faring (e.g. state of the environment reports: ANZECC State of the Environment Reporting Task Force 2000). Other conservation applications include tracking the impacts of land use on biodiversity (e.g. the Wild Bird Index), and measuring the outcomes of conservation management (e.g. the Northern Spotted Owl).

Things we know 1.  We need clear objectives related to quantifiable monitoring and management outcomes Regardless of the scale or purpose of use, indicators need to be selected based on explicit objectives related to how and why they will be used, linking the indicator clearly with expected monitoring and management outcomes. The inability to define explicit objectives has been identified as a significant problem in environmental management and monitoring (Lindenmayer et al. 2012). Understanding the objectives – the ‘why’, ‘what’ and ‘how’ of indicators – is an essential pre-requisite for ensuring that monitoring is efficient and relevant (i.e. fit for purpose). Currently, only 5% of management projects from major conservation organisations have a question-driven monitoring approach (Muir 2010), and it is likely that for indicator approaches, this figure is even lower. Composite indicators such as IUCN’s Red List assess the number of species listed as threatened to

6: Using decision theory to select indicators for managing threats to biodiversity

indicate change in an ecosystem, and can only indicate a long-term response to management due to up-listing or down-listing of a species (IUCN 2008). Ideally, we would like to be able to measure changes on a much shorter time frame: that is, before a species declines so much that it must be listed as threatened, or even extinct. A clear objective sets out both the goals of the use of the indicator: for example, ‘find the indicator species to monitor that maximise the likelihood of detecting an increase of 10% in the growth rate of the target species due to management’, but also the constraints, that might relate to funding ‘within a budget of $1 million’, or risk ‘and will tell us about whether species have responded positively to management with a confidence level of 90%’, or time ‘within the next 50 years’. The more specific and targeted the objective with respect to expected outcomes, the fewer indicators are likely required. For instance, in a case study of finding species to indicate responses to invasive Red Fox (Vulpes vulpes) management in Western Australia, focusing objectives to find indicators that represented only the responses of threatened species changed the indicators selected for monitoring (Tulloch et al. 2013). Fewer species were needed to detect changes in the system, reducing the costs of monitoring and leaving more money for other actions. The consequences of choosing the wrong indicator can be severe. If we had 100% certainty in the outcomes of management, we wouldn’t need indicators at all. But systems are variable. The more uncertainty we have, the more we need to know, and the less a single indicator will be useful for telling us about what is going on. For example, concluding that a management action for the recovery of multiple threatened species has been successful based on a positive response for a single species might result in the cessation of monitoring or management, and the redistribution of funding to other actions. If that conclusion is wrong and other species are declining, this could lead to extinctions (Field et al. 2004). At the least, there will be monetary losses involved with ineffective management and having wasted funding on monitoring the wrong species. In the example of selecting indicators for measuring the success of invasive Red Fox control in Australia (Fig. 6.1), species such as Tammar Wallabies (Macropus eugenii) and Brushtail Possums (Trichosurus vulpecula) are likely to respond positively to management of this predator. A simple objective targeting only these indicators to monitor the success of fox control projects would conclude that predator management has been successful. However, at least three threatened mammals that coincide with these positively responding species are more likely to decline than increase when foxes are managed (Fig. 6.1: Woylie (Bettongia penicillata, Ringtail Possum Pseudocheirus occidentalis and Bandicoot Isoodon obesulus). This is because of complex bottom-up and top-down effects such as food and shelter availability, mesopredator release and interacting threats. An objective of finding indicators to monitor the likelihood of positive or negative responses to fox control would ensure all potential outcomes of management (including perverse) are covered.

2.  Thinking about complementarity avoids redundancies and reduces uncertainty When uncertainty is high, or there is a range of potential responses by species to management, we should select sets of indicator species rather than a single top-ranked indicator (Wade et al. 2014). Most indicator selection frameworks result in a ranked list of species, but a simple ranking approach cannot account for the realities of inter-dependencies between species and actions. Basing indicator selection on the principle of complementarity avoids redundancies. Redundant indicators exhibit significant positive correlations: that is, they provide the same information. However, incorporating complementarity to avoid redundancies is a complex problem because it requires an understanding of interac-

47

48

Indicators and Surrogates of Biodiversity and Environmental Change

Fig. 6.1.  A food-web approach to selecting indicators, adapted from Tulloch et al. (2013). The thickness of each line represents how well one node might indicate changes in another. Environmental attributes can be used as an indicator of the state of the environment. In the diagram, poor soil nutrients can be an indicator of Woylie declines (like beavers, Woylies and bandicoots are ecosystem engineers; Fleming et al. 2014). Threatened species can be used to indicate the state of the environment (such as in the IUCN Red List reporting) or to indicate the success of threat mitigation actions. In the diagram, Woylie declines can be an indicator of increased feral predators, and Woylie increases might indicate response to invasive predator management. However, Woylie responses interact strongly with fire and its management. Tammar wallabies might serve as a better indicator species to monitor as their responses are more specific to invasive predator management. Bandicoots might serve as a surrogate for Woylies (indicated by the dashed lines) as they are easier to detect, occupy similar habitats and are threatened by similar things.

tions between species, threats and responses to actions. Redundancy is rarely tested in the terrestrial world. However, methods have been proposed to assess indicator complementarity post-indicator selection, using statistical techniques such as correlation analysis, multivariate analysis and mutual information analysis (Blanchard et al. 2010; Greenstreet et al. 2012). The objective is to reduce the suite of indicators to a parsimonious number, thereby ensuring all ecosystem attributes/properties are captured and avoiding bias when synthesising across indicators. This kind of post hoc analysis is useful to save money, but explicitly accounting for complementarity within the decision-making framework is a better way to do it. By including the level of certainty that one indicator could represent another species, studies have found that some species currently used to indicate the effectiveness of conservation management decisions are redundant, as they respond in similar way – adding a second indicator species to the monitoring regime provides no additional benefits (Tulloch et al. 2013).

6: Using decision theory to select indicators for managing threats to biodiversity

Fig. 6.2.  Trade-offs between selecting many indicators and few. If we have a budget that can be spent on either management or monitoring of species, selecting more indicator species to monitor and increasing our confidence in the outcomes we report (towards the right of the x-axis) means that we have less money for managing. More indicators selected also means more chance of redundancies due to double counting, but lower risk of missing a perverse outcome due to management (e.g. one species declining while the indicator increases). The red line indicates a potential optimal outcome for indicator selection if all trade-offs and likely outcomes are evaluated using a decision-theoretic approach. Note that curves are hypothetical and may take many different forms. (Adapted from Possingham et al. 2012.)

3.  Considering monitoring costs and indicator cost-effectiveness is crucial One of the main reasons to use indicators is to save time or money or both – if the costeffectiveness of an indicator has not been considered, it is hard to justify an indicator’s use against simply measuring the item of interest. Finding cost-effective indicators to measure the success of a management action can save managers a huge amount of money that can be spent on mitigation or monitoring of other threats (Gardner et al. 2008, Peck et al. 2014). For instance, by using a cost-effectiveness analysis to inform indicator selection decisions, Tulloch et al. (2013) found that the most informative indicator set had a combined cost of half that of the currently monitored set of species in a case study of monitoring invasive fox management in south-western Australia. Ignoring costs in indicator selection could prioritise more expensive species for monitoring (Tulloch et al. 2011), leaving less money for management and ultimately resulting in poor management outcomes (Fig. 6.2). Despite this, many current frameworks for selecting indicators species still ignore costs.

49

50

Indicators and Surrogates of Biodiversity and Environmental Change

Table 6.1.  Examples of national and global scale indicator species selection approaches Example

Example species

Indicator selection methods

Potential management issues

Australia’s State of the Environment Indicators (2000) 1

1. Number of listed species and ecological communities. 2. Estimated populations of selected species

Pre-existing threatened species lists: species identified according to perceived conservation value or rarity

Lacks links to management objectives, actions and likely outcomes

New Zealand’s Core National Environmental Indicators (2013) 2

Lesser short tailed bat; Kiwi; Kākā; Kōkako; Mōhua; Wrybill; Dactylanthus

Expert scoring against criteria

Lacks consideration of cost-effectiveness and likely outcomes of actions

IUCN Climate Change Flagship Species (2011) 3, 4

Leatherback Turtle; Quiver Tree; Arctic Fox; Koala

Experts to choose charismatic flagship species

Lacks consideration of likely outcomes of actions

Global Wild Bird Index 5; Europe’s Common Farmland Bird Index 6

Farmland birds

Statistical trends: correlation with change (due to threat/ management)

Lacks consideration of cost-effectiveness

USA Management Indicator Species (MIS) Approach 7

Pileated Woodpecker; Ruffed Grouse; Meadow Vole; Virginia Northern flying Squirrel

Multi-criteria decision-making approach

Little guidance on dealing with uncertainty, complementarity and trade-offs

USA Surrogate Species Approach 8

Indiana Bat; Bald and Golden Eagles; Pileated Woodpecker; Black Bear; Cutthroat Trout

Dependent on context: might involve structured decision making, or experts to select based on perceived conservation value (landscape surrogates) or rarity (species requiring ‘special attention’)

Little guidance on dealing with uncertainty, complementarity and trade-offs

1. ANZECC State of the Environment Reporting Task Force (2000) 2. Monks et al. (2013) 3. Foden et al. (2011) 4. Barua et al. (2011) 5. Biodiversity Indicators Partnership (2014); Gregory and van Strien (2010) 6. PECBMS (2014) 7. e.g. Mosely et al. (2010) 8. US Department of the Interior (2012)

4.  Choosing indicators necessitates trade-offs in representativeness, certainty in outcomes and costs. Finding the best indicator or set of indicators is a complex question involving the exploration of trade-offs in costs, management outcomes, uncertainty and the ability of different species or systems to represent others. Many indicator selection frameworks are available (see examples in Table 6.1). However, most traditional approaches are qualitative, rather than quantitative, selecting indicators based on their appeal, perceived ecological, social or

6: Using decision theory to select indicators for managing threats to biodiversity

economic value or those that represent an expression of a desired biological outcome (e.g. flagship species, endangered species, species crucial for conservation tourism). The most commonly used frameworks are open to subjectivity and bias because they rely on expert opinion. For example, many frameworks ask stakeholders to score candidates against a range of criteria, rather than weighing the relative benefits of each indicator against the likely costs of measuring it.

5.  Strategic decision-theoretic approaches help make difficult choices by meeting explicit objectives under uncertainty. What has been lacking from indicator selection frameworks until recently is a simple approach that can pick indicators objectively and transparently while trading off costs, uncertainty and expected outcomes from management. Decision theory provides a transparent, systematic approach for choosing monitoring strategies that meet explicit objectives under uncertainty (Possingham 2001). Decision-theoretic approaches require the expected benefits and costs of alternative decisions to monitor candidate indicators to be calculated and related to management outcomes and objectives. By being quantitative, they avoid issues of bias and subjectivity. An informative indicator species should ideally provide greater statistical power and allow the detection of real, rather than spurious, changes due to natural variation, be representative of the responses of others in the landscape and be cost-effective to monitor. New decision-theoretical advances meet these fundamental requirements for indicators by incorporating the likelihood of detecting change due to management (outcomes), what that change tells us about other species (representativeness), and our confidence in the information we have (uncertainty), all while accounting for the costs of monitoring.

Knowledge gaps 6.  Current indicator selection methods do not evaluate indicators in terms of their ability to reduce uncertainty If we had perfect information about our system, or high confidence in our management actions, we might not need to monitor at all. This is rarely the case. We use indicators to learn about environmental change and the consequences of our threat mitigation actions. However, many indicators used today are redundant, providing little additional benefits to the monitoring program and often giving the same information as another indicator. The reason redundant or uninformative indicators are still being used is that most indicator selection methods do not evaluate indicators in terms of their ability to reduce uncertainty in decision making. We need approaches that help decision makers explore the relationships between our level of uncertainty in ecosystem processes and their threats, and the number or type of indicators we need to select. The more uncertainty we have in our system, the more indicators we might pick, to ensure ourselves against unforeseen events (Fig. 6.2). Uncertainty will be related not only to the amount of biodiversity in our system, but also how many threats, and potential actions are available. Value of information (VOI) analysis evaluates the benefits of collecting additional information to reduce uncertainty in a specific decision-making context (Raiffa and Schlaifer 1961). The approach can be used to estimate quantitatively the expected benefit of monitoring a candidate indicator for management decision making, compare alternative indicators, and select those that are optimal given a management objective and budget constraints. Although VOI has been used to inform health, engineering, business, fisheries and biodiversity conservation deci-

51

52

Indicators and Surrogates of Biodiversity and Environmental Change

sions (Bratvold et al. 2009; Forsberg and Guttormsen 2006; Runge et al. 2011; Yokota and Thompson 2004), it has not yet been applied to indicator selection.

7.  Most of our systems for selecting indicators assume we have prior information on how they respond to threats and actions One of the disadvantages of VOI and many decision-theoretic approaches to selecting indicators is that they assume we have prior information on how species respond to threats and actions. For systems under new threats or for new management actions, how do we pick thresholds of action or inaction without prior information? How much change is acceptable before we change our management action? Experts have been used in the past to determine responses when there are few empirical data available, but more work is needed in synthesising existing information and identifying knowledge gaps. To identify useful indicators, existing long-term monitoring programs and initiatives need to be identified and the information they have derived used to parameterise decision-support tools for indicator selection such as VOI. Time-series data are useful and important for managers to be able to calculate population variability over time rather than simply estimating an overall predicted trend (Wilson et al. 2011). A meta-database identifying existing monitoring efforts would allow us to take into consideration the ability of past monitoring activities to detect trends in target species in relation to where, how and why the monitoring is taking place, as a basis for work to develop a robust and repeatable indicator prioritisation approach. A useful database would include information such as the monitoring method, spatial location and extent, duration, data quality and approximate costs for each candidate indicator species, as well as the ability to link each species to threats or ongoing management actions when available. At a national scale, these databases are largely unavailable worldwide, although some countries such as Australia and the UK have begun to synthesise information for some taxonomic groups (e.g. http://www.nerpdecisions.edu.au/special-project-indicator-species-monitoring.html). Progress has also been made on a regional basis towards consolidating dynamic online repositories of remotely sensed information supporting data upload, analysis, dissemination and reporting (e.g. NatureServe’s Biodiversity Indicators Dashboard; http://www.natureserve.org/conservation-tools/projects/ biodiversity-indicators-dashboard). These new ‘dashboards’ help managers and policy makers to visualise and document progress of the key biodiversity indicators towards achieving the Aichi Targets. 8.  Indicators can represent actions, threats or ultimate biodiversity outcomes – what are the consequences of not knowing all of these parts of the management cycle? Indicators can represent actions (how much we are doing to mitigate a threatening process?), threats (how is the threatening process changing in response to action?) or ultimate biodiversity outcomes (how is the ecosystem and the target biodiversity responding to management?). Uncertainty exists in what situations we need to know all of these parts of the management cycle, and what the consequences are of not knowing all of these components. This cycle is much like a food web, in which some species might represent other lower trophic levels with more certainty than others, which often relates to how strong the link is between the species (Fig. 6.1). The same is likely true of the management cycle. The threats with fewer possible actions, fewer candidate indicators, and more certainty about links with actions and species responses, are likely easier to be monitored by only parts of this cycle than others with complex relationships. Approaches such as food-web theory

6: Using decision theory to select indicators for managing threats to biodiversity

allow nodes with probabilistic values of survival for candidate species, to be linked with possible management actions, their costs and their likely outcomes. Despite many highly cited papers on food-web and network theory over the past few decades, none have used this approach to link indicators to management outcomes. The use of networks would allow us to visualise the relationships and interactions between actions, threats, indicators and ultimate biodiversity outcomes (the target species), and enable the selection of complementary indicator species. The next step for this is to account for further complexities of the real world. In most cases, more than one level of surrogacy often exists. For example, species 1 informs on species 2, which informs on species 3, so therefore species 1 gives some information on species 3. Furthermore, the optimal indicator is likely to change over time if adaptive management is being conducted. These time- and state-dependent questions are computationally difficult, and require probabilistic approaches such as Bayesian Belief Networks or Markov Decision Processes (Chadès et al. 2012; Nyberg et al. 2006).

9.  Indicators are rarely chosen with more than one threat and mitigating action in mind – how is it best to account for the complexities of interacting threats? One of the issues with selecting indicators that are linked to management outcomes is the fact that rarely is there only a single threat that acts on a species – nearly all ecological systems face multiple threats. Multiple interacting threats and species make it difficult to determine to what threat and to what action a species is responding. Current indicator selection frameworks deal with only one threat at a time (e.g. Tulloch et al. 2013). Designing indicator selection strategies to learn about the effects of multiple actions is not trivial, and requires knowledge of the web of interactions between species (Lane et al. 2014), threats and actions (Fig. 6.1). Frameworks that deal with interdependencies, such as through a food-web approach, could allow managers to evaluate trade-offs between different objectives. These trade-offs might necessitate choosing a suite of highly specific indicators that represent with high certainty responses to each threat in the ecosystem but can’t be easily extrapolated to other threats. In Fig. 6.1, this might correspond with selecting the wallaby to represent invasive fox management, which is an excellent indicator of fox management success but a poor indicator of fire impacts, and is unable to tell us anything about feral cats (Felis catus) and the species that respond to them. Alternatively, managers might trade off finding indicators that are more representative of the whole-of-ecosystem state, but might provide less confidence in responses of individual species to specific threats or actions. In Fig. 6.1, this might include monitoring soil-digging mammals such as bandicoots to inform on functional biodiversity – the condition of a habitat, or provisioning services, which are tied to soil nutrients. 10.  Indicators are about accepting risk – can we use risk analysis to assess what the consequences might be if the indicator is wrong? Finally, those choosing indicators need to face the reality that any decision to accept only a partial understanding of the ecosystem and its threats, such as through an indicator approach, is an acceptance of risk. The level of risk that a decision maker is willing to accept can be incorporated into the indicator selection process. In this way, managers can weigh up the pros and cons of each indicator in relation to the confidence in management outcomes that it can provide. Some indicator selection frameworks incorporate uncertainty in a probabilistic way into selection of indicators. Most focus only on the benefits of the indicators (how responsive are they and how specific?). Instead of assessing

53

54

Indicators and Surrogates of Biodiversity and Environmental Change

against how good an indicator might be (i.e. how much information it provides on your chosen threat-action), risk assessment allows managers to assess what the consequences might be if an indicator is wrong. The more risk-averse managers are, the more information they need to be happy with the outcomes, and likely more indicators need to be monitored (the right side of Fig. 6.1). Most managers will have a threshold of certainty below which consequences are not accepted. Risk assessment is widely conducted in health, business and environmental management contexts such as mining accident risk analysis and fisheries production (Markowitz 1959; Burgman 2005), and is increasingly proposed in some conservation contexts (e.g. Mouysset et al. 2013; Tulloch et al. 2015). Risk analysis is just as applicable to an indicator selection approach, and it would be interesting to see this avenue explored further.

Conclusions Indicator choices should be driven by the decision context and associated uncertainties. Indicator species can, if chosen carefully, provide managers with a systematic, cost-effective and repeatable way to measure and monitor the outcomes of conservation actions. However, for maximum utility of decisions for adaptive management, consideration of the costs and benefits of monitoring each indicator, including the ways in which species and threats interact and the range of potential responses, is crucial. Surprisingly, systematic evaluation of the most cost-effective indicator to evaluate a specific management goal is a step that is often ignored. Without this, decisions will likely select suboptimal suites of indicators, leaving less money for management or, worse, leading to further declines in a system where the responses of species have not been adequately detected. Picking the wrong indicator to represent responses of ecosystems to threatening processes and their mitigating actions can be disastrous. Perverse outcomes include: (1) missing a short-term response because the chosen indicator responded too slowly; (2) wasted money if the chosen indicator is redundant and results in double counting of a management impact; (3) ignoring potential negative outcomes on other species or communities’ and (4) loss of biodiversity if objectives are too tightly constrained to evaluating only the success of management. It is important for those selecting indicators to acknowledge that species interact, and actions targeted at one species will impact other species either directly or indirectly. Furthermore, decisions that ignore uncertainty run the risk of a lack of confidence in the outcome. Most current frameworks deal with some, but not all, of the components for incorporating decision theory into indicator choices for management decisions. A decision-theoretic resource allocation framework for indicators provides several benefits. It allows for the consideration of multiple objectives, multiple actions or threats, the level of confidence we have in the ability to measure responses accurately, and the risk of getting things wrong. More importantly, an objective and transparent way of making choices under uncertainty might ultimately save money by integrating actions that benefit multiple species, and avoiding cascading problems due to system changes that negatively affect other species in the system.

References ANZECC State of the Environment Reporting Task Force (2000) Core Environmental Indicators for Reporting on the State of the Environment. Australian and New Zealand Environment and Conservation Council (ANZECC), Canberra ACT, .

6: Using decision theory to select indicators for managing threats to biodiversity

Barua M, Root-Bernstein M, Ladle R, Jepson P (2011) Defining flagship uses is critical for flagship selection: a critique of the IUCN climate change flagship fleet. Ambio 40, 431–435. Biodiversity Indicators Partnership (2014) Global Wild Bird Index for Habitat Specialists. UNEP World Conservation Monitoring Centre, Cambridge, UK. . Blanchard JL, Coll M, Trenkel VM, Vergnon R, Yemane D, Jouffre D, et al. (2010) Trend analysis of indicators: a comparison of recent changes in the status of marine ecosystems around the world. ICES Journal of Marine Science 67, 732–744. doi:10.1093/icesjms/fsp282. Bratvold RB, Bickel JE, Lohne HP (2009) Value of information in the oil and gas industry: past, present, and future. SPE Reservoir Evaluation & Engineering 12, 630–638. doi:10.2118/110378-PA. Burgman MA (2005) Risks and Decisions for Conservation and Environmental Management. Cambridge University Press, New York. Chadès I, Carwardine J, Martin TG, Nicol S, Sabbadin R, Buffet O (2012) MOMDPs: a solution for modelling adaptive management problems. In Proceedings of the 26th AAAI Conference on Artificial Intelligence (AAAI12). 22–26 July, Toronto, Canada. . Accessed 22 March 2015. Dawson WR, Ligon JD, Murphy JR, Myers JP, Simberloff D, Verner J (1987) Report of the Scientific Advisory Panel on the Spotted Owl. The Condor 89, 205–229. doi:10.2307/1368783. European Environment Agency (2010) ‘Towards an early warning and information system for invasive alien species (IAS) threatening biodiversity in Europe’. Technical Report No. 5, European Environment Agency (EAA), Copenhagen, . Failing L, Gregory R (2003) Ten common mistakes in designing biodiversity indicators for forest policy. Journal of Environmental Management 68, 121–132. doi:10.1016/S03014797(03)00014-8. Field SA, Tyre AJ, Jonzen N, Rhodes JR, Possingham HP (2004) Minimizing the cost of environmental management decisions by optimizing statistical thresholds. Ecology Letters 7, 669–675. doi:10.1111/j.1461-0248.2004.00625.x. Fleming PA, Anderson H, Prendergast AS, Bretz MR, Valentine LE, Hardy GES (2014) Is the loss of Australian digging mammals contributing to a deterioration in ecosystem function? Mammal Review 44, 94–108. doi:10.1111/mam.12014. Foden W, Stuart SN (2009) Species and Climate Change: More than Just the Polar Bear. IUCN Species Survival Commission (SSC), Gland, Switzerland. Forsberg OI, Guttormsen AG (2006) The value of information in salmon farming. Harvesting the right fish at the right time. Aquaculture Economics & Management 10, 183–200. doi:10.1080/13657300600985215. Gardner TA, Barlow J, Araujo IS, Avila-Pires TC, Bonaldo AB, Costa JE, et al. (2008) The cost-effectiveness of biodiversity surveys in tropical forests. Ecology Letters 11, 139–150. doi:10.1111/j.1461-0248.2007.01133.x. Greenstreet SPR, Fraser HM, Rogers SI, Trenkel VM, Simpson SD, Pinnegar JK (2012) Redundancy in metrics describing the composition, structure, and functioning of the North Sea demersal fish community. ICES Journal of Marine Science 69, 8–22. doi:10.1093/ icesjms/fsr188. Gregory RD, van Strien A (2010) Wild bird indicators: using composite population trends of birds as measures of environmental health. Ornithological Science 9, 3–22. doi:10.2326/ osj.9.3.

55

56

Indicators and Surrogates of Biodiversity and Environmental Change

IUCN (2008) IUCN Red List of Threatened Species. IUCN, Gland, Switzerland and Cambridge, UK. Lane PW, Lindenmayer DB, Barton PS, Blanchard W, Westgate MJ (2014) Visualization of species pairwise associations: a case study of surrogacy in bird assemblages. Ecology and Evolution 4, 3279–3289. doi:10.1002/ece3.1182. Lindenmayer DB, Gibbons P, Bourke M, Burgman M, Dickman CR, Ferrier S, et al. (2012) Improving biodiversity monitoring. Austral Ecology 37, 285–294. doi:10.1111/j.1442-9993. 2011.02314.x. Markowitz HM (1959) Portfolio Selection: Efficient Diversification of Investments. John Wiley & Sons, New York. Ministry for the Environment (2007) Environment New Zealand 2007. Ministry for the Environment, Manatü Mö Te Taiao, Wellington, New Zealand. Monks JM, O’Donnell CFJ, Wright EF (2013) Selection of Potential Indicator Species for Measuring and Reporting on Trends in Widespread Native Taxa in New Zealand. Department of Conservation, New Zealand, . Mosely KR, Ford WM, Edwards JW, Strager MP (2010) A Multi-Criteria Decision-making Approach for Management Indicator Species Selection on the Monongahela National Forest, West Virginia. USA Forest Service, PA, . Mouysset L, Doyen L, Jiguet F (2013) How does economic risk aversion affect biodiversity? Ecological Applications 23, 96–109. doi:10.1890/11-1887.1. Muir M (2010) ‘Are we measuring conservation effectiveness? A survey of current resultsbased management practices in the conservation community’. Unpublished report, Conservation Measures Partnership, . Niemeijer D, de Groot RS (2008) A conceptual framework for selecting environmental indicator sets. Ecological Indicators 8, 14–25. doi:10.1016/j.ecolind.2006.11.012. Nyberg JB, Marcot BG, Sulyma R (2006) Using Bayesian belief networks in adaptive management. Canadian Journal of Forest Research 36, 3104–3116. doi:10.1139/x06-108. PECBMS (Pan-European Common Bird Monitoring Scheme) (2014) European wild bird indicators, 2014 update. European Bird Census Council, . Peck MR, Maddock ST, Morales JN, Oñate H, Mafla-Endara P, Peñatfiel VA, et al. (2014) Cost-effectiveness of using small vertebrates as indicators of disturbance. Conservation Biology 28, 1331–1341. doi:10.1111/cobi.12373. Possingham HP (2001) The Business of Biodiversity. Applying Decision Theory Principles to Nature Conservation. The Australian Conservation Foundation and Earthwatch Institute, South Melbourne. Possingham HP, Wintle BA, Fuller RA, Joseph LN (2012) The conservation return on investment from ecological monitoring. In Biodiversity Monitoring in Australia. (Eds D Lindenmayer and P Gibbons) pp. 49–61. CSIRO Publishing, Melbourne. Raiffa H, Schlaifer R (1961) Applied Statistical Decision Theory. Graduate School of Business Administration, Harvard University, MA. Rodrigues ASL, Brooks TM (2007) Shortcuts for biodiversity conservation planning: the effectiveness of surrogates. Annual Review of Ecology Evolution and Systematics 38, 713–737. doi:10.1146/annurev.ecolsys.38.091206.095737. Runge MC, Converse SJ, Lyons JE (2011) Which uncertainty? Using expert elicitation and expected value of information to design an adaptive program. Biological Conservation 144, 1214–1223. doi:10.1016/j.biocon.2010.12.020.

6: Using decision theory to select indicators for managing threats to biodiversity

Tittensor DP, Walpole M, Hill SLL, Boyce DG, Britten GL, Burgess ND, et al. (2014) A mid-term analysis of progress toward international biodiversity targets. Science 346, 241–244. doi:10.1126/science.1257484. Tulloch A, Possingham HP, Wilson K (2011) Wise selection of an indicator for monitoring the success of management actions. Biological Conservation 144, 141–154. doi:10.1016/j. biocon.2010.08.009. Tulloch AIT, Chadès I, Possingham HP (2013) Accounting for complementarity to maximize monitoring power for species management. Conservation Biology 27, 988–999. Tulloch AIT, Maloney RF, Joseph LN, Bennett JR, Di Fonzo MMI, Probert WJM, et al. (2015) Effect of risk aversion on prioritizing conservation projects. Conservation Biology 29, 513–524. US Department of the Interior (2012) ‘Draft technical guidance on selecting species for design of landscape scale conservation’. US Fish and Wildlife Service, USA, . Wade ASI, Barov B, Burfield IJ, Gregory RD, Norris K, Vorisek P, et al. (2014) A niche-based framework to assess current monitoring of European forest birds and guide indicator species’ selection. PLoS ONE 9, e97217. doi:10.1371/journal.pone.0097217. Wedeles C, Ray J, Dzus E, Korol C, Morel S (2014) ‘Proposed indicators to address species at risk, including Woodland Caribou, in Canada’s Forest Management Standard’. Report prepared for the Forest Stewardship Council Canada (FSC-CA), Toronto, Canada. Wilson HB, Kendall BE, Possingham H (2011) Variability in population abundance and the classification of extinction risk. Conservation Biology 25, 747–757. Yokota F, Thompson KM (2004) Value-of-information analysis in environmental health risk management decisions: past, present, and future. Risk Analysis 24, 635–650.

57

This page intentionally left blank

‡7HQ&RPPLWPHQWVSSLQGG

$0

7

Invertebrate indicators and ecosystem restoration Philip Barton and Melinda Moir

Things we know 1 2 3 4

Invertebrates have many attributes of a good indicator. Numerous taxa are available as indicators. Invertebrates can indicate a wide range of restoration objectives. Invertebrate assemblages rather than single species are typically used as indicators. 5 Mine rehabilitation leads the way in the application of invertebrate indicators. Knowledge gaps 6 The ‘build it and they will come’ paradigm is flawed. 7 Multiple indicators are needed for the restoration of multiple ecosystem characteristics. 8 Functional approaches show promise. 9 Assessment of invertebrate indicators should include control sites and longterm monitoring. 10 Choice of invertebrate indicator(s) needs clear conceptual justification.

Introduction Human activity has degraded or destroyed large areas of land and vegetation, and the ecological restoration of these areas is a major scientific and land management priority (Hobbs and Cramer 2008; Benayas et al. 2009). One of the most fundamental research questions in any restoration program is how to measure its progress or success. Biological indicators, such as the abundance of a species or diversity of a set of taxa, are often used in this endeavour. Invertebrates are a useful group of organisms for measuring ecosystem change and are commonly used to indicate either changes in biodiversity or particular ecological processes relative to a reference or control state. In this chapter, we discuss some of the key advances in knowledge of the use of invertebrate indicators, and some areas requiring improvement.

59

60

Indicators and Surrogates of Biodiversity and Environmental Change

Things we know 1.  Invertebrates have many attributes of a good indicator Invertebrates comprise the bulk of eukaryotic biodiversity (Mora et al. 2011), and correspondingly display a vast array of life histories, sizes and degrees of habitat specialisation. This enormous variety makes invertebrates attractive as potential indicators of many aspects of ecosystems because particular taxa can be closely associated with key processes or representative of other suites of taxa. It is for this reason that invertebrates are used extensively in monitoring of biodiversity or ecological processes worldwide (McGeoch 1998; Samways et al. 2010; Gerlach et al. 2013). Particular attributes of invertebrates make them ideal indicators for many ecosystem properties, some of which have been discussed by Samways and colleagues (2010) and Gerlach and colleagues (2013). These include: (1) many taxa are easy to survey, thus providing a cost-effective option to gathering biodiversity data; (2) invertebrates are small in size and occupy a wide range of niches, thus being highly sensitive to environmental change; (3) they are highly mobile and capable of colonising new habitat quickly; (4) their fast generation times and abundance mean they can be ‘numerically responsive’ to habitat change; and (5) they are highly diverse in their biology and ecology, with different species able to be linked to specific environmental parameters. There are, however, some drawbacks to using invertebrate indicators, including: (1) many groups are taxonomically challenging and require specialists for identification; (2) many species are poorly known and their ecological role is not well understood; and (3) laboratory and microscopy equipment are required for processing and sorting. These benefits and drawbacks need to be considered in the context of the questions being asked and the resources available for each restoration program (McGeoch 1998; Gerlach et al. 2013). 2.  Numerous taxa are available as indicators A search of any scientific literature database quickly yields many papers on invertebrate indicators. This vast literature has been subject to several detailed reviews (e.g. McGeoch 1998; Gerlach et al. 2013), but these cover a very diverse range of applications beyond restoration. To gain an understanding of the current state of invertebrate indicators in restoration, we performed a ‘topic’ search in the ISI Web of Science database using the search string ‘indicator* and (insect* OR arthropod* OR invert*) and restoration’ (performed 27 August 2014). After excluding many aquatic and marine studies, as well as review papers, our search yielded 54 empirical studies that used invertebrates as some kind of indicator of restoration success in the last 20 years. Overall, we found 11 different Orders (or higher taxa) of invertebrate had been studied, with 39% of studies examining two or more Orders (Fig. 7.1). The remaining studies focused on only one taxon, and most commonly studied taxa were ants (17%), beetles (11%) and spiders (9%). Only 4% of studies were non-arthropod taxa (one study using nematodes, one using earthworms). Although our search was not comprehensive, it demonstrates that some taxa are favoured over others. For example, ants (Hymenoptera: Formicidae) have been used in mine rehabilitation studies (Majer 1983; Andersen and Sparling 1997), dung beetles (Coleoptera: Scarabaeidae) have been used as indicators of habitat quality (Gollan et al. 2011; Audino et al. 2014), and bugs (Hemiptera) have been used as indicators of vegetation recovery (Moir et al. 2005; Moir et al. 2010). A detailed review of the pros and cons of different invertebrate taxa as indicators of environmental change was conducted by Gerlach and colleagues (2013). They found that the success of particular taxa as indicators depended on what part of the ecosystem was surveyed. Their conclusions were that combinations of isopods, earthworms or mites be used for soil-related systems; ants, millipedes, snails,

7: Invertebrate indicators and ecosystem restoration

Fig. 7.1.  The number of empirical restoration studies that focus on different invertebrate indicators from the last 20 years. (Data sourced from the ISI Web of Science, August 2014.)

ground beetles, harvestmen or gnaphosid spiders be used for the ground layer (using pitfall traps, Fig. 7.2); ants, chrysomelid beetles, theridiid spiders or arctiid moths be used for arboreal habitats (using vacuum suction or branch beating, Fig. 7.2); and ants, orthopter-

Fig. 7.2.  An example of different approaches to sampling invertebrates. Top: Melinda Moir using vacuum suction to actively sample invertebrates present in the shrub layer in heath vegetation, Western Australia. Bottom left: a pitfall trap is an inexpensive and passive technique used to sampling ground-active invertebrates. Bottom right: Melinda using a stick to beat low-hanging tree branches and dislodge foliage invertebrates into a collecting bag.

61

62

Indicators and Surrogates of Biodiversity and Environmental Change

ans or butterflies be used for open habitat (Gerlach et al. 2013). This rough guide suggests that there should be strong functional links between the chosen invertebrate groups and the habitat of interest. .

3.  Invertebrates can indicate a wide range of restoration objectives Restoration programs can have different objectives, depending on the kind of disturbance being studied (e.g. grazing, agriculture, mining or invasive species). These various disturbances could impact on general biodiversity levels or particular ecological processes differently, thus each may require particular indicators to provide information about ecosystem change. For example, reforestation might focus on the return of dominant plant species, whereas mine rehabilitation might focus on the return of soil condition and vegetation communities. In each case, invertebrates can be useful indicators of these different restoration objectives. This is because there are invertebrate taxa associated with nearly every structure or process found in a terrestrial ecosystem. The presence of these taxa in the restored ecosystem might therefore indicate progress towards the return of these structures or processes. For example, the presence of particular pollinating insects might be used as indicators of success of restoration of flowering plant communities (Forup and Memmott 2005). Similarly, the presence of specific insect herbivores might be useful indicators of vegetation restoration (Moir et al. 2005). The diversity of earthworms might also be a useful indicator of improved soil health (Riggins et al. 2009). Current evidence suggests that many invertebrate groups have variable congruence with diversity patterns of other taxa (Westgate et al. 2014), but can be good indicators of more defined ecosystem characteristics or attributes (Gerlach et al. 2013). Therefore, where restoration is occurring to improve biodiversity, invertebrates should only be used in combination with other taxa to improve representation. It is important to note that the absence of invertebrate species is also the objective of some restoration programs, especially where the species is invasive. For example, the control of ants on oceanic islands can be part of restoration programs, with their decline or absence indicators of success (Plentovich et al. 2011). 4.  Invertebrate assemblages rather than single species are typically used as indicators Invertebrate indicators are often represented by assemblages of species or as functional groups of species, such as herbivores (Woodcock et al. 2008). This approach is very different to that for many vertebrate indicators, where single species are commonly used, and often advocated as ‘flagship’ or ‘umbrella’ species (see Chapter 4). For example, the Koala (Phascolarctos cinereus) has been used implicitly as an umbrella for other fauna when rehabilitating habitat damaged by mining in eastern Australia (Cristescu et al. 2013). Of course, single species of invertebrate have also been recommended as indicators of a particular ecosystem state or attribute, but this is usually after a larger assemblage of species has been considered (Fagan et al. 2010). It is much less common for invertebrate species to be used as an umbrella species, except in the case of habitat being restored for some butterfly species (Sands 2008). 5.  Mine rehabilitation leads the way in the application of invertebrate indicators Some of the most innovative and comprehensive restoration programs undertaken have been by mining companies. These may involve the re-construction of soil, vegetation and faunal communities, and the long-term assessment of progress. Numerous fauna groups

7: Invertebrate indicators and ecosystem restoration

have been used to assess mine rehabilitation, including many invertebrates (Majer et al. 2007; Cristescu et al. 2012). Here, we provide a case study of invertebrates used as indicators of bauxite mine restoration. One of the most intensive restoration projects is being undertaken by Alcoa of Australia Ltd in south-west Western Australia. Importantly, this company publishes much of their research (see Cristescu et al. 2012), which allows other organisations to learn from their successes and failures in a form of adaptive management. As Cristescu et al. (2012) commented ‘The literature… [is] representative of mining companies with an interest in research … highly probably linked with high environmental practices’. Alcoa (or researchers that they fund) have examined the return of many different orders of invertebrate, beginning with ants in the mid-1970s (Majer et al. 2013), and including ground-dwelling and arboreal predators (spiders, scorpions), arboreal herbivores (Hemiptera) and various detritivores (termites, worms, wood-beetles) (Majer et al. 2007). Current projects (as of 2014) focus on promoting the return of invertebrates that provide key ecosystem functions, such as litter decomposition and seed dispersal. These different groups have been examined for different reasons, including habitat and multi-trophic linking properties (Majer et al. 2007). When Alcoa decide to alter their restoration techniques, they often assess the impact on invertebrates. For example, when Alcoa wanted to reinstate fire into the restored system, because it is a natural part of the surrounding Jarrah forest, they assessed the double disturbance of restoration and fire on ground-dwelling and arboreal spiders, as well as assessing multiple other factors such as floral diversity, vegetation structure and fuel loads (Brennan et al. 2003). Research findings have led to improved reinstatement of Jarrah forest plant diversity (e.g. for plant-dwelling insects: Moir et al. 2005), returning large rocks and logs (e.g. for ground-dwelling and saproxylic invertebrates), and returning fine woody debris and leaf-litter (for ground-dwelling and saproxylic invertebrates: Lythe 2012 pers. comm.). Finally, and perhaps most importantly, to circumvent annual variation or climatic changes, and assess the success of the restoration over longer time scales, Alcoa have implemented long-term monitoring programs for ground-dwelling invertebrates.

Knowledge gaps 6.  The ‘build it and they will come’ paradigm is flawed Many restoration programs focus on the return of vegetation, with little attention paid to the subsequent return of fauna (Cristescu et al. 2012). This is now recognised as a major impediment to successful restoration, especially where there is a risk of restoring unexpected or undesirable biodiversity (Lindenmayer et al. 2012). Simply adding vegetation and waiting for associated fauna to arrive, including invertebrates, therefore appears to be a naïve assumption. The examination of invertebrate assemblages in revegetated areas has been central to highlighting the flawed logic in expecting the simple planting of shrubs and trees to lead to comprehensive or desired biodiversity gains. Vegetation planted for different restoration purposes, such as erosion control, salinity abatement or habitat connectivity can lead to different biodiversity outcomes, depending on the mix of plant species (Barton et al. 2013). Yet, vegetation planted to mimic pre-existing and natural conditions, and with the goal of improving biodiversity, might still not yield the desired objectives. For example, even strongly linked plant-dwelling herbivores may not return to restored areas when their host plants are present due to a series of filters preventing their return. These include poor dispersal capabilities, presence of competitors, predators or the wrong abiotic conditions (Moir et al. 2005).

63

64

Indicators and Surrogates of Biodiversity and Environmental Change

7.  Multiple indicators are needed for the restoration of multiple ecosystem characteristics The restoration of soils, vegetation, faunal communities and their associated ecosystem services, each require a different set of indicators. This suggests that the more comprehensive a restoration program, the more comprehensive the set of indicators needs to be to provide the information necessary for thorough assessment. Conceptually, the idea of multiple indicators has been advocated over the past decade (Maes and Van Dyck 2005; Samways et al. 2010; Gerlach et al. 2013), and approaches to the selection of indicators are also available (McGeoch 1998; Maes and Van Dyck 2005; Samways et al. 2010). From a restoration perspective, the selection of multiple indicators should consider the following: • What is the restoration intervention? This information is necessary to identify which strata of the ecosystem are being impacted (e.g. soil, leaf-litter, understorey, canopy), and which invertebrate groups will be most sensitive to this impact. See Fig. 7.2 for examples of sampling techniques for different habitat strata. • What is the objective? Is the intervention aimed at returning biodiversity, or a particular ecological process? This information will be useful for selecting the taxa that are most representative of the objective. For example, assessing the biodiversity within restored sites may include examining the diversity of higher taxa (e.g. beetles and ants). By contrast, assessing the restoration of key ecological processes could involve documenting the abundance of several functionally linked taxa (e.g. pollination by different bee species). • What spatial and temporal scales are involved? This information can be used to select which taxa might respond most appropriately over the anticipated scales. For example, short-term studies will require taxa with fast generation times, whereas isolated areas might require highly mobile taxa. The above information can be synthesised to shortlist sets of taxa most appropriate to measuring the anticipated change. All available information on the natural history of the different taxa also should be considered. As noted by Samways and colleagues (2010), however, ecological information at the level of species may be very poorly known, necessitating additional studies or compromises. Such a compromise may be selecting a set of taxa with ecologies that span the greatest range so as to increase their potential representation of the ecosystem properties of interest. Subsequent tests of their performance over time will help to define their usefulness in different circumstances.

8.  Functional approaches show promise Functional traits are characteristics of species, such as dispersal ability or body size, that act as a proxy for how an organism interacts with its environment (Violle et al. 2007). There are two key advantages to using a functional approach to invertebrate indicators. First, it allows a more direct mechanistic link between species and a particular ecological process of interest. For example, long-winged insects or spiders that are capable of ballooning might be expected to be capable of moving large distances to colonise new habitat (Moir et al. 2005; Lambeets et al. 2009). Second, focusing on traits can reveal patterns and responses that transcend the idiosyncrasies of individual species, and provide more general, albeit approximate, prediction of species responses. For example, aspects of the size and shape of ants has been shown to respond similarly to different environments, suggesting morphology might be useful to predict foraging behaviour and habitat use (Gibb and Parr 2013).

7: Invertebrate indicators and ecosystem restoration

Some functional traits are shared by species within the same clade, and this can simplify functional grouping. Ants provide some of the best examples of this, with species in the subfamily Dolichoderinae generally showing competitive dominance behaviour, and species in the genus Camponotus being subdominant within ant assemblages. This broad grouping has been effective at predicting ant assemblage responses to disturbance (Andersen and Sparling 1997; Majer and Nichols 1998; Andersen and Majer 2004). More recently, molecular approaches and isotope studies have been used to determine trophic position, prey items and cryptic species. These methods are more expensive, but are becoming cheaper, and present new ways to identify functional attributes of species potentially useful for indicators of restoration progress. For example, stable isotopes have been used to reveal shifts towards a greater diversity of herbivorous ant species in older eucalypt plantings on farmland (Gibb and Cunningham 2011). Future use of invertebrate indicators in restoration may not be simply in documenting their abundance and composition, but analysing their DNA, trophic position and functional roles to obtain the best information about successful restoration techniques.

9.  Assessment of invertebrate indicators should include control sites and long-term monitoring Gauging restoration success incurs the general problem of how to monitor environmental change. Key to this is the use of control or reference sites, and the need for long-term monitoring. Although environmental monitoring focuses on change over time, restoration success must be measured relative to a control. Despite this, Wortley and colleagues (2013) found that of 301 empirical restoration studies considered, 26% of studies did not use an explicit control in their assessment of restoration success. Further, Wortley and colleagues (2013) reported that ~20% of restoration studies were conducted at sites less than 5 years old and 45% at studies less than 10 years old. Although these data apply to all restoration studies, and not just invertebrates, it demonstrates that these problems are relatively common in the restoration literature. Short-term studies of invertebrate indicators are sensitive to dramatic seasonal or annual changes in species populations and, potentially, whole communities. This can occur due to rare and extreme weather events, for example, that drive large fluxes in the primary production underpinning food webs. Findings from short-term studies must be interpreted with caution. The remedy, of course, is long-term studies over several seasons and years. Such examples include the study of the diversity and composition of ant assemblages over a period of 37 years in restored mine-sites in Western Australia, which may also begin to show the effects of climate change (Majer et al. 2007). Further, sampling across seasons can identify species only present in a single season or outbreaks of one species that may influence diversity indices (Moir et al. 2011). 10.  Choice of invertebrate indicator(s) needs clear conceptual justification The choice of taxon to act as an indicator of ecosystem change can be vague or poorly justified. The reasons for why one taxon is used instead of another is often a result of the researcher’s own expertise or interest, as much as it is due to cost or logistical constraints. Careful consideration needs to be given to: (1) which invertebrate taxa are cost-effective to survey and provide the most data quickly, and (2) the most appropriate taxa to represent the ecological change being measured.

65

66

Indicators and Surrogates of Biodiversity and Environmental Change

The more specific the change being measured, the more specific the relationship needs to be with the species or assemblage. This might require detailed knowledge of the taxonomy and ecology of the potential indicator. For example, assessment of broad changes in biodiversity can be achievable through the sampling of several taxa, and assessment of species richness using the ‘morphospecies’ approach (Oliver and Beattie 1996). This can provide suitable data to quantify coarse changes in species richness. By contrast, the assessment of whether the right kinds of species have returned to a restored area requires greater species-level information, and the morphospecies approach is unlikely to be appropriate. Accurate species identification provides extra information, such as if they are introduced or locally endemic, which can substantially change the measure of restoration success. For example, a high abundance of introduced Portuguese millipedes is not a good indication that restoration of Australian habitats has succeeded in returning native species, but it may indicate that decomposition and nutrient cycling processes are taking place.

Conclusion The restoration of terrestrial ecosystems is a major land management priority, with a strong focus on both biodiversity and ecosystem services (Benayas et al. 2009; Suding 2011). The fact that invertebrates comprise the bulk of biodiversity and deliver many valuable ecosystem services means they will be a necessary part of restoration, both as the restoration goal and as a tool to monitor progress and success. Improving the science behind the use of invertebrate indicators will be critical to the effective restoration of ecosystems.

Acknowledgements We thank zoologist Vicki Stokes for updated information on the activities of Alcoa of Australia Ltd MM was funded by the Australian Research Council’s Centre of Excellence for Environmental Decisions (ARC CEED) and Australian National Environmental Research Program (NERP).

References Andersen AN, Majer JD (2004) Ants show the way Down Under: invertebrates as bioindicators in land management. Frontiers in Ecology and the Environment 2, 291–298. doi:10.1890/1540-9295(2004)002[0292:ASTWDU]2.0.CO;2. Andersen AN, Sparling GP (1997) Ants as indicators of restoration success: relationship with soil microbial biomass in the Australian seasonal tropics. Restoration Ecology 5, 109–114. doi:10.1046/j.1526-100X.1997.09713.x. Audino LD, Louzada J, Comita L (2014) Dung beetles as indicators of tropical forest restoration success: Is it possible to recover species and functional diversity? Biological Conservation 169, 248–257. doi:10.1016/j.biocon.2013.11.023. Barton PS, Colloff MJ, Pullen KR, Cunningham SA (2013) Arthropod assemblages in a focal tree species (Eucalyptus microcarpa) depends on the species mix in restoration plantings. Biodiversity and Conservation 22, 2091–2110. doi:10.1007/s10531-013-0530-4. Benayas JMR, Newton AC, Diaz A, Bullock JM (2009) Enhancement of biodiversity and ecosystem services by ecological restoration: a meta-analysis. Science 325, 1121–1124. doi:10.1126/science.1172460.

7: Invertebrate indicators and ecosystem restoration

Brennan K, Majer J, Koch A (2003) Using fire to facilitate faunal colonization following mining: an assessment using spiders in Western Australian jarrah forest. Ecological Management & Restoration 2, 145–147. Cristescu RH, Frere C, Banks PB (2012) A review of fauna in mine rehabilitation in Australia: current state and future directions. Biological Conservation 149, 60–72. doi:10.1016/j. biocon.2012.02.003. Cristescu RH, Rhodes J, Frere C, Banks PB (2013) Is restoring flora the same as restoring fauna? Lessons learned from koalas and mining rehabilitation. Journal of Applied Ecology 50, 423–431. doi:10.1111/1365-2664.12046. Fagan KC, Pywell RF, Bullock JM, Marrs RH (2010) Are ants useful indicators of restoration success in temperate grasslands? Restoration Ecology 18, 373–379. doi:10.1111/j.1526-100X. 2008.00452.x. Forup ML, Memmott J (2005) The restoration of plant-pollinator interactions in hay meadows. Restoration Ecology 13, 265–274. doi:10.1111/j.1526-100X.2005.00034.x. Gerlach J, Samways M, Pryke J (2013) Terrestrial invertebrates as bioindicators: an overview of available taxonomic groups. Journal of Insect Conservation 17, 831–850. doi:10.1007/ s10841-013-9565-9. Gibb H, Cunningham SA (2011) Habitat contrasts reveal a shift in the trophic position of ant assemblages. Journal of Animal Ecology 80, 119–127. doi:10.1111/j.1365-2656.2010.01747.x. Gibb H, Parr CL (2013) Does structural complexity determine the morphology of assemblages? An experimental test on three continents. PLoS ONE 8, e64005. doi:10.1371/ journal.pone.0064005. Gollan JR, Reid CAM, Barnes PB, Wilkie L (2011) The ratio of exotic-to-native dung beetles can indicate habitat quality in riparian restoration. Insect Conservation and Diversity 4, 123–131. doi:10.1111/j.1752-4598.2010.00115.x. Hobbs RJ, Cramer VA (2008) Restoration ecology: interventionist approaches for restoring and maintaining ecosystem function in the face of rapid environmental change. Annual Review of Environment and Resources 33, 39–61. doi:10.1146/annurev.environ.33. 020107.113631. Lambeets K, Vandegehuchte ML, Maelfait JP, Bonte D (2009) Integrating environmental conditions and functional life-history traits for riparian arthropod conservation planning. Biological Conservation 142, 625–637. doi:10.1016/j.biocon.2008.11.015. Lindenmayer DB, Hulvey KB, Hobbs RJ, Colyvan M, Felton A, Possingham H, et al. (2012) Avoiding bio-perversity from carbon sequestration solutions. Conservation Letters 5, 28–36. doi:10.1111/j.1755-263X.2011.00213.x. Maes D, Van Dyck H (2005) Habitat quality and biodiversity indicator performances of a threatened butterfly versus a multispecies group for wet heathlands in Belgium. Biological Conservation 123, 177–187. doi:10.1016/j.biocon.2004.11.005. Majer JD (1983) Ants: bio-indicators of minesite rehabilitation land-use and land conservation. Environmental Management 7, 375–383. doi:10.1007/BF01866920. Majer JD, Nichols OG (1998) Long-term recolonization patterns of ants in Western Australian rehabilitated bauxite mines with reference to their use as indicators of restoration success. Journal of Applied Ecology 35, 161–182. doi:10.1046/j.1365-2664.1998.00286.x. Majer JD, Brennan KEC, Moir ML (2007) Invertebrates and the restoration of a forest ecosystem: 30 years of research following bauxite mining in Western Australia. Restoration Ecology 15, S104–S115. doi:10.1111/j.1526-100X.2007.00298.x.

67

68

Indicators and Surrogates of Biodiversity and Environmental Change

Majer JD, Heterick B, Gohr T, Hughes E, Mounsher L, Grigg A (2013) Is thirty-seven years sufficient for full return of the ant biota following restoration? Ecological Processes 2, 19. doi:10.1186/2192-1709-2-19. McGeoch MA (1998) The selection, testing and application of terrestrial insects as bioindicators. Biological Reviews of the Cambridge Philosophical Society 73, 181–201. doi:10.1017/S000632319700515X. Moir ML, Brennan KEC, Koch JM, Majer JD, Fletcher MJ (2005) Restoration of a forest ecosystem: the effects of vegetation and dispersal capabilities on the reassembly of plant-dwelling arthropods. Forest Ecology and Management 217, 294–306. doi:10.1016/j. foreco.2005.06.012. Moir ML, Brennan KEC, Majer JD, Koch JM, Fletcher MJ (2010) Plant species redundancy and the restoration of faunal habitat: lessons from plant-dwelling bugs. Restoration Ecology 18, 136–147. doi:10.1111/j.1526-100X.2010.00654.x. Moir ML, Brennan KEC, Fletcher MJ, Majer JD, Koch JM (2011) Multi-scale patterns in the host specificity of plant-dwelling arthropods: the influence of host plant and temporal variation on species richness and assemblage composition of true bugs (Hemiptera). Journal of Natural History 45, 2577–2604. doi:10.1080/00222933.2011.597522. Mora C, Tittensor DP, Adl S, Simpson AGB, Worm B (2011) How many species are there on earth and in the ocean? PLoS Biology 9, e1001127. doi:10.1371/journal.pbio.1001127. Oliver I, Beattie AJ (1996) Invertebrate morphospecies as surrogates for species: a case study. Conservation Biology 10, 99–109. doi:10.1046/j.1523-1739.1996.10010099.x. Plentovich S, Eijzenga J, Eijzenga H, Smith D (2011) Indirect effects of ant eradication efforts on offshore islets in the Hawaiian Archipelago. Biological Invasions 13, 545–557. doi:10.1007/s10530-010-9848-y. Riggins JJ, Davis CA, Hoback WW (2009) Biodiversity of belowground invertebrates as an indicator of wet meadow restoration success (Platte River, Nebraska). Restoration Ecology 17, 495–505. doi:10.1111/j.1526-100X.2008.00394.x. Samways MJ, McGeoch MA, New TR (2010) Insect conservation: a handbook of approaches and methods. Oxford University Press, New York. Sands D (2008) Conserving the Richmond Birdwing butterfly over two decades: where to next? Ecological Management & Restoration 9, 4–16. doi:10.1111/j.1442-8903.2008.00382.x. Suding KN (2011) Toward an era of restoration in ecology: successes, failures and opportunities ahead. Annual Review of Ecology Evolution and Systematics 42, 465–487. doi:10.1146/annurev-ecolsys-102710-145115. Violle C, Navas ML, Vile D, Kazakou E, Fortunel C, Hummel I, et al. (2007) Let the concept of trait be functional! Oikos 116, 882–892. doi:10.1111/j.0030-1299.2007.15559.x. Westgate MJ, Barton PS, Lane PW, Lindenmayer DB (2014) Global meta-analysis reveals low consistency of biodiversity congruence relationships. Nature Communications 5, 3899. doi:10.1038/ncomms4899. Woodcock BA, Edwards AR, Lawson CS, Westbury DB, Brook AJ, Harris SJ, et al. (2008) Contrasting success in the restoration of plant and phytophagous beetle assemblages of species-rich mesotrophic grasslands. Oecologia 154, 773–783. doi:10.1007/s00442-0070872-2. Wortley L, Hero JM, Howes M (2013) Evaluating ecological restoration success: a review of the literature. Restoration Ecology 21, 537–543. doi:10.1111/rec.12028.

8

Mosses as passive and active indicator surrogates for investigations of atmospheric pollution and quality Hanna Salo

Things we know 1 Ubiquitous mosses are excellent bioaccumulators. 2 Inter- and intraspecies variations exist due to different habitat requirements, accumulation properties and uptake efficiencies. 3 Substrate and secondary sources or processes may contribute to element and compound levels in mosses. 4 Sensitivity and varying tolerance make native mosses suitable for qualitative studies. 5 Moss monitoring can enhance the spatial and temporal accuracy of quantitative data. Knowledge gaps 6 The relationship between moss survey data and atmospheric pollution levels is difficult to determine. 7 There is an increasing demand for pollution source identification. 8 The lack of standardisation causes pitfalls for the intercomparison of moss bag surveys. 9 There are challenges for the further development of moss monitoring.

Introduction Monitoring of mosses for atmospheric research was introduced by Rühling and Tyler (1968). They employed mosses as biomonitors for the accumulation of traffic-originated lead in Sweden. The first national-level surveys of mosses were completed in Scandinavia in the late 1970s from where they expanded to all of the Nordic countries and to most of the European countries in the 1980s and 1990s (Poikolainen 2004). Air pollutants, which are defined as ‘any substance present in ambient air and likely to have harmful effects on human health and/or the environment as a whole’ (European Commission 2008), have impacts at all spatial and temporal scales by remaining close to the emission sources, as 69

70

Indicators and Surrogates of Biodiversity and Environmental Change

Table 8.1.  Definitions for key terms Key term

Definition

Biomonitoring

An approach where the state of the environment or part of it is investigated by applying organisms or part of them either as bioindicators or biomonitors.

Bioindicator

The use of organisms or their parts to reflect the qualitative changes in the environment or part of it.

Biomonitor

The use of organisms or their parts for quantitative measurements of accumulated elements and compounds.

Passive monitoring

Application of native mosses both as qualitative indicators and quantitative accumulators of air pollutants. This approach is especially suitable for extensive (inter)national surveys.

Active monitoring

Relocation of moss material from unpolluted to polluted site through the use of moss bags or transplants. This approach is advantageous in small-scale surveys or ‘moss desert’ areas experiencing the lack of native mosses due to high pollution levels.

well as moving thousands of kilometres from emission sources. Thus, moss monitoring has been applied in surveys of local pollution hotspots, regional deposition patterns and continental emission spread, particularly in Europe and North America. The use of mosses in atmospheric monitoring can be categorised as bioindicators and biomonitors (Table 8.1). In the bioindicator approach, qualitative information is based on the effects or impacts of air pollutants (e.g. in species richness, distribution or condition). In the biomonitor approach, mosses are used as accumulators, yielding quantitative data on elements and compounds (Markert et al. 2003). Moss research is typically focused on the present concentrations and spatial distribution of atmospheric heavy metals. However, they are also used in studies of nitrogen, sulphur, radionuclides, organic pollutants and compounds such as polycyclic aromatic hydrocarbons (PAH), and magnetic minerals. Historical pollution levels and changes can be evaluated through comparisons with herbarium collections or peat profiles. In addition, atmospheric monitoring using mosses can be passive or active: the former entails the use of native mosses in situ, and the latter approach involves the moss bags or the transplantation of mosses (Table 8.1).

Things we know 1.  Ubiquitous mosses are excellent bioaccumulators Mosses (Musci or Bryopsida) are bryophytes (Phylum Bryophyta). They rank second in diversity among the land plants: ~8000 of the world’s estimated 14 000 to 15 000 bryophyte species are mosses (Hallingbäck and Hodgetts 2000; Encyclopædia Britannica 2014). Bryophytes are geographically widespread and often found in humid or moist regions growing on soil, rocks, trees and man-made structures. Some species (e.g. Fontinalis) have adapted to aquatic habitats. Bryophytes are generally poikilohydric; that is, they become dormant in dry seasons because they lack structures or mechanisms to prevent water loss (Chopra and Kumra 1988). Nevertheless, they have adapted to dry environments by developing physiological tolerance to desiccation but they can rehydrate rapidly. Two major physiological groups of bryophytes are recognised: endohydric species and ectohydric species, while myxohydric species possess features of both groups. Endohydric

8: Mosses as passive and active indicator surrogates for investigations of atmospheric pollution and quality

mosses have a cuticle-like covering and vascular tissues and they absorb water mainly through rhizoids (Chopra and Kumra 1988). The majority of mosses are ectohydric, which have very thin non-waxy leaves and cuticle, no specialised vascular tissues and weakly developed rhizoids (if rhizoids occur at all). Ectohydric species effectively absorb moisture and dissolved substances, such as nutrients and pollutants, over their entire external surface from dry deposition and precipitation through capillary systems (Chopra and Kumra 1988). The lack of a cuticle enhances uptake from the atmosphere while the lack of true roots limits the direct influence of substrate (Szczepaniak and Biziuk 2003). Slow growth rate and lifespan usually of 2–5 years (but up to 10 years) allows mosses to accumulate air pollutants over a considerable period of time and in larger amounts than that of vascular plants (e.g. Rühling and Tyler 1968). Other advantages of mosses in atmospheric monitoring include a large surface area to weight ratio, a habit of growing in groups, easy and cheap sampling, and the possibility of identifying annual growth segments (Poikolainen 2004). These characteristics mean that mosses (together with lichens) are the most widely used plants in biomonitoring.

2.  Inter- and intraspecies variations exist due to different habitat requirements, accumulation properties and uptake efficiencies Several factors make it challenging to identify common moss species suitable for extensive regional or (inter)national surveys. First, there are interspecies variations in accumulation properties and element uptake efficiencies. For example, Čeburnis and colleagues (1999) reported very similar metal uptake efficiency in Hylocomium splendens and Pleurozium schreberi (see also e.g. Berg and Steinnes 1997), but significantly lower uptake efficiencies for Eurhynchium angustirete. Second, although mosses are omnipresent, most of them are demanding in relation to habitat, moisture and light conditions. Furthermore, species richness, occurrence, distribution and abundance vary significantly between climatic zones, such as between tropical and coniferous forests (Syrjänen 2002). This adds to the difficulty in identifying widespread species suitable for regional or (inter)national surveys. Third, there can be major differences in the chemical composition of individual species, populations of the same species, and between individuals and their separate parts (Onianwa 2001; Poikolainen 2004). For example, younger segments of mosses usually have lower concentrations of some elements than the older parts (e.g. Bargagli et al. 1995) but elements may also be translocated internally between the basal and apical parts of moss shoots (Zechmeister et al. 2003a; Aboal et al. 2010). 3.  Substrate and secondary sources or processes may contribute to element and compound levels in mosses Substrate is generally considered to be an insignificant contributor of elements to mosses. However, correlations between concentrations of metals in mosses and substrate metal levels have been reported and some species (known as copper mosses) are geobotanical indicators of metallic enrichment (Chopra and Kumra 1988; Onianwa 2001). Secondary processes, such as windblown soil dust, leaching or drip from dead or living vegetation, and natural cycling processes, such as the transport of sea salt, also may contribute to the element concentrations in mosses (e.g. Berg et al. 1995; Berg and Steinnes 1997; Čeburnis et al. 1999). In northern countries, significant levels of pollutants appear in run-off water and in particle suspension in the air during the short, but intensive, road dust period after snow melt. Dry deposition of particles can occur on the surface of mosses, especially in

71

72

Indicators and Surrogates of Biodiversity and Environmental Change

arid climates (Poikolainen 2004), but these may be washed away by intensive rain (Čeburnis et al. 1999, Zechmeister et al. 2003b). Thus, sampling after dry periods followed by vigorous rainfall should be avoided.

4.  Sensitivity and varying tolerance make native mosses suitable for qualitative studies Mosses are sensitive to air pollution. However, tolerance varies between species. For example, in Central Europe, Ceratodon purpureus is a resistant to air pollution, Hypnum cupressiforme is insensitive, Leucodon sciuroides is sensitive and Neckera pennata is very sensitive to environmental pollution (Zechmeister et al. 2003a). In plants, including mosses, stress exposure (e.g. by air pollutants) first trigger a decline in physiological function(s) such as impaired growth. Mechanisms for coping with stress, such as morphological adaptations or repair processes, are then activated. When coping mechanisms are exceeded by continuous intensive stress, severe damage, and ultimately death, can occur (Fränzle 2003; Zechmeister et al. 2003a). A complete absence of mosses (i.e. a ‘moss desert’) can result in heavily polluted areas. 5.  Moss monitoring can enhance the spatial and temporal accuracy of quantitative data Air quality data are most often based on measuring pollutants directly in the air using automated monitoring stations, in deposition (e.g. by precipitation studies), or by constructing models of the spread of a given pollutant (Markert et al. 2003). Such data have poor spatial coverage because the number of monitoring stations or sample locations is typically limited and costs are high. Therefore, local spatial variation in pollutants and sources of pollution may remain unknown. The temporal resolution of automated stations is high, but pollutant data can be limited to one or a small number of pollutants: most stations continuously measure levels of such substances as ozone and particulate matter (PM) but ignore heavy metals. Monitoring of mosses, however, enables a large number of pollutants to be investigated simultaneously (Ares et al. 2012), inexpensively, and with a high degree of spatial and temporal representativeness/accuracy. Native mosses have a high spatial representativeness. Hence, passive moss monitoring is appropriate for extensive regional or (inter)national studies, such as the European moss survey, which is conducted every 5 years. The temporal resolution of indigenous species is usually in agreement with the lifespan of mosses, and therefore provides a good reflection of long-term trends in the air pollution and quality. Some species, such as Sphagnum spp. and Hylocomium splendens, grow a new segment annually. This enables precise determination of plant age and exposure time and thus improves the temporal accuracy of results. However, native mosses can adapt to the environment and high pollution levels by developing detoxification mechanisms or by decreasing retention capacity. Mosses can therefore lower heavy metal concentrations leading to under-estimations of the level of deposition (Fernández and Carballeira 2000; Boquete et al. 2013). Active moss monitoring is useful in industrial or urban regions, near local pollution sources and in moss desert areas. It provides: a simple way to establish spatially detailed, dense sampling networks or transects without separate power supplies; temporal accuracy in a sense of a known exposure time; initial concentration levels for the contaminants of interest; and inconspicuous samples less prone to vandalism. Moreover, the lithogenic impact directly from the soil is very small, especially in moss bags because they are placed

8: Mosses as passive and active indicator surrogates for investigations of atmospheric pollution and quality

above the ground. However, the road dust period and windblown soil in the Nordic countries or arid regions, respectively, may still contribute to concentrations of pollutants. Active monitoring techniques help the assessment and identification of emission levels and pollution sources (e.g. Salo and Mäkinen 2014).

Knowledge gaps 6.  The relationship between moss survey data and atmospheric pollution levels is difficult to determine A major challenge is to determine how well moss survey data reflects true levels of atmospheric pollution. According to Steinnes (1995), the concentration data of mosses reflect only relative deposition patterns of pollutants. To estimate absolute deposition rates (e.g. in mg/m2/year), the concentrations of pollutants in mosses needs to be calibrated against wet deposition data for the same sites determined from bulk precipitation analyses. For this, Berg and Steinnes (1997) suggest the use of regression equations. However, after critically evaluating several studies, Aboal and colleagues (2010) argue that bulk deposition cannot be correctly determined from moss concentrations. This is for two key reasons. First, mosses do not integrate atmospheric deposition (e.g. due to other pollutant sources or biological processes). Second, several factors hamper the calibration of data on the accumulation of pollutants in mosses with atmospheric deposition data. 7.  There is an increasing demand for pollution source identification Most air quality research focuses on monitoring the types of pollution, and the concentration and spatial distribution of pollutants. As demands for tighter discharge or emission limits are emerging, the importance of reliable pollution source identification is growing. For this, spatially and temporally representative data on various pollutants are needed: moss survey is an easy and economical sampling approach for unmonitored or heavily polluted areas. However, the recognition of distinct ‘fingerprints’ of pollution sources requires new and multifold methods such as enviromagnetic analyses to be used alongside conventional analyses. Pollutant source and impact identification helps authorities and operators to reduce air pollution by linking the environmental protection policies and actions to the right targets (Salo and Mäkinen 2014). 8.  The lack of standardisation causes pitfalls for the intercomparison of moss bag surveys The moss bag technique (Fig. 8.1) was introduced by Goodman and Roberts (1971), and is more widespread than moss transplants. This technique was standardised in Finland in the 1990s (SFS 5794), allowing comparisons between studies across that nation. However, there is a wide range of different moss species, pre-treatments of moss material and moss bag shapes (reviewed by Ares et al. 2012). Therefore, intergovernmental moss bag studies are not readily comparable and there is a clear need for international standardisation. Conversely, guidelines for the use of native and transplanted mosses exist (e.g. Cenci 2008). The EU funded MOSSclone programme (www.mossclone.eu) aims to develop a protocol for a standardised use of moss bags as air pollution sensors. Moss clone cultivation in the laboratory is proposed as a way to provide homogenous material for moss bags and to enable comparison between different studies and monitoring programs. Ares and colleagues (2012) suggested guidelines for standardised methodology, including cloning of

73

74

Indicators and Surrogates of Biodiversity and Environmental Change

Fig. 8.1.  Active monitoring by moss bags is a practical alternative in small-scale surveys near pollution sources or heavily polluted areas lacking native species (photo H. Salo).

mosses. This idea is interesting and can solve the problems related to the availability of suitable species with constant initial element concentrations. Complex preparation procedures and substantially rising costs due to standardisation should not limit the major advantages of using moss bags.

9.  There are challenges for the further development of moss monitoring Guidelines for passive moss monitoring and sampling are well established (e.g. Cenci 2008), while active moss bag techniques are currently in progress (e.g. the MOSSclone program). Challenges in the use of native mosses include the need to calibrate relative accumulation efficiency among moss species, especially when new species are applied in research (Carballeira et al. 2008). Problems of moss bags relate to insufficiently available specific moss species or the capability of moss bags to endure varying weather conditions during the exposure period. Artificial materials could be used to replace moss bags (e.g. Salo 2014). Research is required to find a substitutive material that is effective and durable accumulator of air pollutants and easier to use than moss bags.

Conclusions Monitoring of mosses can provide accurate and reliable data on the concentration and distribution of atmospheric pollutants. Many properties of mosses, such as good retention and

8: Mosses as passive and active indicator surrogates for investigations of atmospheric pollution and quality

accumulation features, widespread distribution, longevity and simultaneous accumulation of various pollutants, results in them being excellent and usable indicator surrogates for air quality and air pollution. Moss surveys are also spatially and temporally representative, as well as easy to implement. Passive biomonitoring using native mosses is the most advantageous for extensive regional or (inter)national studies when large areas need to be covered. Active biomonitoring – that is, moss bags or transplants – is a recommended alternative for native mosses in small-scale surveys near pollution sources or in moss desert areas. Research is needed to find the best way to determine the true relationship between concentrations of pollutants in mosses and in the atmosphere. Moss surveys provide a practical sampling approach for the identification of pollution sources, but to do this, new multifold methods need to be adopted alongside conventional analyses. While current attempts to create an international moss bag standard are essential, it is important that the fundamental need for simple and affordable active monitoring is not undermined by complex or expensive procedures. Finally, the replacement of moss bags with an artificial material could offer less-laborious way to sample air pollutants.

References Aboal JR, Fernández JA, Boquete T, Carballeira A (2010) Is it possible to estimate atmospheric deposition of heavy metals by analysis of terrestrial mosses? The Science of the Total Environment 408, 6291–6297. doi:10.1016/j.scitotenv.2010.09.013. Ares A, Aboal JR, Carballeira A, Giordano S, Adamo P, Fernández JA (2012) Moss bag biomonitoring: a methodological review. The Science of the Total Environment 432, 143–158. doi:10.1016/j.scitotenv.2012.05.087. Bargagli R, Brown DH, Nelli L (1995) Metal biomonitoring with mosses: procedures for correcting for soil contamination. Environmental Pollution 89(2), 169–175. doi:10.1016/0269-7491(94)00055-I. Berg T, Steinnes E (1997) Use of mosses (Hylocomium splendens and Pleurozium schreberi) as biomonitors of heavy metal deposition: from relative to absolute deposition values. Environmental Pollution 98(1), 61–71. doi:10.1016/S0269-7491(97)00103-6. Berg T, Røyset O, Steinnes E (1995) Moss (Hylocomium splendens) used as biomonitor of atmospheric trace element deposition: estimation of uptake efficiencies. Atmospheric Environment 29(3), 353–360. doi:10.1016/1352-2310(94)00259-N. Boquete MT, Fernández JÁ, Carballeira A, Aboal JR (2013) Assessing the tolerance of the terrestrial moss Pseudoscleropodium purum to high levels of atmospheric heavy metals: a reciprocal transplant study. Science of the Total Environment 461–462, 552–559. Carballeira CB, Aboal JR, Fernández JA, Carballeira A (2008) Comparison of the accumulation of elements in two terrestrial moss species. Atmospheric Environment 42(20), 4904–4917. doi:10.1016/j.atmosenv.2008.02.028. Čeburnis D, Steinnes E, Kvietkus K (1999) Estimation of metal uptake efficiencies from precipitation in mosses in Lithuania. Chemosphere 38(2), 445–455. doi:10.1016/S00456535(98)00183-0. Cenci RM (2008) Guidelines for the Use of Native Mosses, Transplanted Mosses and Soils in Assessing Organic and Inorganic Contaminant Fallout. European Communities, Luxembourg. Chopra RN, Kumra PK (1988) Biology of Bryophytes. Wiley Eastern Limited, New Delhi. Encyclopædia Britannica (2014) Plant. Encyclopaedia Britannica, London, .

75

76

Indicators and Surrogates of Biodiversity and Environmental Change

European Commission (2008) Directive 2008/50/EC of the European Parliament and of the Council of 11 June 2008 on ambient air quality and cleaner air for Europe. Official Journal of the European Union L 152/1, . Fernández JA, Carballeira A (2000) Differences in the responses of native and transplanted mosses to atmospheric pollution: a possible role of selenium. Environmental Pollution 110, 73–78. doi:10.1016/S0269-7491(99)00278-X. Fränzle O (2003) Bioindicators and environmental stress assessment. In Bioindicators and Biomonitors: Principles, Concepts and Applications. (Eds BA Markert, AM Breure and HG Zechmeister) pp. 41–84. Elsevier, Amsterdam. Goodman GT, Roberts TM (1971) Plants and soils as indicators of metals in the air. Nature 231(5301), 287–292. doi:10.1038/231287a0. Hallingbäck T, Hodgetts N (2000) Mosses, Liverworts, and Hornworts. Status Survey and Conservation Action Plan for Bryophytes. IUCN/SSC bryophyte specialist group. IUCN, Gland, Switzerland and Cambridge, UK. Markert BA, Breure AM, Zechmeister HG (2003) Definitions, strategies and principles for Bioindication/biomonitoring of the environment. In Bioindicators and Biomonitors: Principles, Concepts and Applications. (Eds BA Markert, AM Breure and HG Zechmeister) pp. 3–39. Elsevier, Amsterdam. Onianwa PC (2001) Monitoring atmospheric metal pollution: a review of the use of mosses as indicators. Environmental Monitoring and Assessment 71, 13–50. doi:10.1023/A: 1011660727479. Poikolainen J (2004) Mosses, epiphytic lichens and tree bark as biomonitors for air pollutants – specifically for heavy metals in regional surveys. PhD thesis. University of Oulu, Finland. Rühling Å, Tyler G (1968) An ecological approach to the lead problem. Botaniska Notiser 121, 321–342. Salo H (2014) Preliminary enviromagnetic comparison of the moss, lichen, and filter fabric bags to air pollution monitoring. Fennia 192(2), 154–163. doi:10.11143/41354. Salo H, Mäkinen J (2014) Magnetic biomonitoring by moss bags for industry-derived air pollution in SW Finland. Atmospheric Environment 97, 19–27. doi:10.1016/j. atmosenv.2014.08.003. Steinnes E (1995) A critical evaluation of the use of naturally growing moss to monitor the deposition of atmospheric metals. The Science of the Total Environment 160–161, 243–249. doi:10.1016/0048-9697(95)04360-D. Syrjänen K (2002) Sammalten levinneisyys. In Bryophytes of Finland: Distribution, Ecology and Red List Status. (Eds T Ulvinen, K Syrjänen and S Anttila) pp. 25–37. Finnish Environmental Institute, Vammala (in Finnish with English abstract). Szczepaniak K, Biziuk M (2003) Aspects of the biomonitoring studies using mosses and lichens as indicators of metal pollution. Environmental Research 93(3), 221–230. doi:10.1016/S0013-9351(03)00141-5. Zechmeister HG, Grodzińska K, Szarek-Łukasewska G (2003a) Bryophytes. In Bioindicators and Biomonitors: Principles, Concepts and Applications. (Eds BA Markert, AM Breure and HG Zechmeister) pp. 329–375. Elsevier, Amsterdam. Zechmeister HG, Hohenwallner D, Riss A, Hanus-Illnar A (2003b) Variations in heavy metal concentrations in the moss species Abietinella abietina (Hedw.) Fleisch. according to sampling time, within site variability and increase in biomass. The Science of the Total Environment 301, 55–65. doi:10.1016/S0048-9697(02)00296-6.

9

Lichens as ecological indicators to track atmospheric changes: future challenges Cristina Branquinho, Paula Matos and Pedro Pinho

Things we know 1 Metrics for measuring lichen diversity 2 Standardised methods to sample lichen diversity 3 Spatial models obtained from lichen diversity metrics with high spatial resolution 4 Knowledge of the ecological response of lichen diversity to single pollutants 5 An understanding of the integrated responses of lichens to spatial and temporal atmospheric changes Knowledge gaps 6 Comparative studies using lichen diversity in different regions around the world 7 Data on lichen functional diversity 8 An international lichen-trait database 9 Disentangle the effects of multiple pollutants on lichen diversity 10 A definition of lichen thresholds and their inclusion in environmental legislation

Introduction Lichens as ecological indicators Ecological indicators of atmospheric changes can be used as surrogates to describe the effects of atmospheric changes on ecosystem structure and functioning in a simplified, but nonetheless representative, manner. They can also be applied to evaluate the effects of human activities on ecosystem structure and functioning (Box 9.1). This information can then be communicated to environmental stakeholders, including government institutions responsible for creating and implementing environmental legislation. Among the criteria for the selection of ecological indicators of atmospheric changes is a lack of influence of other environment factors, such as soil and/or water. Most vascular plants, for example, are affected by soil proprieties. Epiphytes, which grow non-parasiti77

78

Indicators and Surrogates of Biodiversity and Environmental Change

Box 9.1. Ecological indicators are more than a measure of environmental variables. The same environmental change does not affect all ecosystems in the same way. For example, measuring the change in air temperature provides information on the state of the environment, whereas measuring its impact on the diversity of desert ecosystems or alpine ecosystems, as an ecological indicator, tells us about its effects. A 2°C increase in temperature will have a different impact on desert than on alpine ecosystems. To specifically evaluate the impact of human activities on ecosystems, appropriate ecological indicators are required.

cally on plants, are entirely dependent on the atmosphere for water and nutrients. Some epiphytes, such as bromeliads, orchids and ferns, are able to regulate their uptake of water and nutrients from the atmosphere, through structures such as the cuticle (a protective film covering the epidermis of leaves) and the ability to close their stomata. Consequently, they do not directly reflect atmospheric environmental changes. By contrast, poikilohydric organisms, such as lichens and bryophytes, respond directly to the levels of water and nutrients present in the atmosphere. Because of their dependence on the atmosphere and their inability to control water and nutrient uptake, they are ideally suited as ecological indicators in the evaluation of the effects of atmospheric changes. The use of lichens as ecological indicators has several advantages compared with bryophytes. In particular, lichens are found in almost every biome, including the drier areas from which bryophytes are mostly excluded. Lichens are a symbiosis between two different organisms: a fungus (mycobiont) and a photosynthetic partner (photobiont), which can be a green alga and/or a cyanobacterium. Other features of lichens that account for the interest in their use as ecological indicators of atmospheric changes are listed in Box 9.2 and have been described elsewhere (Branquinho 2001).

Applications of lichens as ecological indicators Lichens have thus far been used as ecological indicators to provide two types of metrics: (1) diversity metrics, based on the differential sensitivity of lichen species to atmospheric changes; and (2) bioaccumulation metrics, based on the capacity of lichens to accumulate chemical elements. In this chapter, we focus on the diversity metrics of lichen communities in response to atmospheric changes. Lichens have been widely used for more than 100 years to study the effects of atmospheric pollution in ecosystems. Lichen diversity can be exploited as an ecological indicator

Box 9.2. Features of lichens that make them reliable ecological indicators of atmospheric changes 1 Lichens species differ in their sensitivities to atmospheric changes. 2 Lichens are slow-growing organisms and do not exhibit seasonal changes in morphology. 3 Lichens have a wide distribution and can be found in almost all terrestrial biomes.

9: Lichens as ecological indicators to track atmospheric changes: future challenges

because different lichen species differ in their tolerance of atmospheric changes: some species are extremely sensitive and will disappear in response to pollution whereas others persist, even in highly polluted air. Identifying the species that occur in a certain area enables us to understand the type and intensity of pollution in a targeted area. Moreover, lichens are responsive to the full range of air quality, from clean air to extremely polluted air. Lichen diversity has been used throughout the world to monitor air quality in urban and industrial areas. In the period beginning with the Industrial Revolution until the 1980s, the loss of lichen diversity in urban and industrial areas could be ascribed to sulphur dioxide pollution. Since then, nitrogen pollution has been the most significant factor in the loss of lichen diversity (Davies et al. 2007). We expect that if current efforts to reduce nitrogen emissions are successful, climate change will emerge as the most important driver of lichen diversity over time. In a recent study, in a city without industrial sources of atmospheric pollution, lichen diversity was shown to more closely reflect the effects of urban heat islands than the effects of nitrogen pollution (Munzi et al. 2014). Based on its successful use as a surrogate of air quality, lichen diversity also has been applied to evaluate the impact of air pollution on human health. In Italy, an association was determined between the air quality measured by a lichen diversity index and mortality due to lung cancer in men (Cislaghi and Nimis 1997). Recently, a study analysed associations between air quality during gestation and birthweight using lichen diversity as a surrogate of air quality in a petro-industrial area in Portugal. A significant association was found between air quality, measured using lichen diversity, and the birthweights of babies exposed during gestation to tobacco smoke (Ribeiro et al. 2014). Lichens are also extremely sensitive to environmental disturbances of natural origin. For example, the extreme sensitivity of lichen diversity to light and water availability has facilitated its use as an indicator of forest integrity (McCune 2000). Lichen diversity also was shown to be sensitive to logging, edge effects, tree density and microclimatic conditions (Pinho et al. 2010), supporting the use of this indicator in forest management and biological conservation. Shifts in lichen species composition due to climate change have been observed within an interval as short as 5 years (van Herk et al. 2002). Thus, lichens may be among the organisms most sensitive to rapid changes in climate (Aptroot and van Herk 2007). Accordingly, there is great interest in developing lichens as a universal indicator of global climate change.

Chapter rationale In this chapter, we review current knowledge on the use of lichens as ecological indicators of atmospheric changes and identify future challenges that must be overcome before they can be used as an ecological indicator of global climate change drivers. We examine what is known thus far and highlight the need for a universal system, in which data obtained from the use of lichens as ecological indicators in specific settings are compared across ecosystems, countries and biomes. An integration of this knowledge can be used to develop a system able to evaluate the impact of global drivers, such as climate change, on ecosystem structure and function. The body of knowledge supporting the use of lichens as ecological indicators has been accumulated from work conducted during the last century and more recently. It has resulted in several metrics that can be used in developing lichens for this purpose (point 1) and in the design and establishment of standard methods that, thus far, have only been applied at local and regional levels in Europe and in the USA (point 2). The ubiquity of lichens is such that there is a wealth of spatial information with high resolution that allows

79

80

Indicators and Surrogates of Biodiversity and Environmental Change

the construction of reliable spatial models (point 3). These ecological models of lichen diversity and other abiotic variables can then be applied to the study of similar geographic areas. The basic ecology and surrogacy relationships underpinning the effects of certain atmospheric changes on lichen diversity (e.g. sulphur and nitrogen pollution) are well understood (point 4). However, lichen diversity responds both spatially and temporally not only to single pollution sources but also to the presence of multiple, simultaneous pollutants and/or disturbances at the same site (point 5). Thus, lichen diversity reflects the integrated effect of those interactions and thereby serves as an end-point in measuring their ecosystem effects. Many studies of lichens have been conducted at the local or regional scale, but only a few at a national scale and almost none on a global scale. Consequently, the indicators confirmed thus far cannot be used at the global level and await confirmation in studies completed across countries and continents (point 6). However, in scaling up from local and regional levels to the global level, it is not possible to use species as indicators, because they are site specific. Other methods to integrate information that can then be used in global comparisons must be developed (point 7). We propose the use of lichen traits, grouped in functional groups as a universal metric, which must be preceded by the creation of an international lichen trait database (point 8). Another point that must be taken into account is that many sites are subject to multiple pollutants and disturbances. Thus, we need to develop methods to disentangle these multiple effects on lichens, to better understand cause-effect relationships (point 9). Application of this new knowledge in management requires that legal thresholds are established that are explicitly designed to avoid the irreversible loss of structure and/or function of ecosystems (point 10).

Things we know 1.  Metrics for measuring lichen diversity The diversity, species richness, coverage and frequency of lichens have all been assessed for their utility as ecological indicators of atmospheric changes. Lichen diversity has been recognised as an ecological indicator since 1866, when Nylander proposed that the absence of lichens along the edge of the Jardins du Luxembourg was caused by air pollution (Gilbert 1973). The identification of sulphur dioxide as the prevailing pollutant influencing lichen diversity in industrialised and urban areas led to the development of a scale to estimate sulphur dioxide concentrations using lichen diversity as an indicator (Hawksworth and Rose 1970). The assumption was made – and subsequently confirmed – that the presence of a certain lichen species would correlate with sulphur dioxide levels within a certain range. Lichen species richness can be used to measure lichen diversity as a surrogate of atmospheric changes (Hawksworth 2002). Species richness is a fundamental measure of biodiversity and is also commonly used in lichen diversity assessments. Previous studies showed that the higher the total species richness, the better the air quality (Pinho et al. 2008a) or the better the state of an ecosystem (McCune 2000). The Index of Atmospheric Purity (IAP) was based on all-species richness determinations and on lichen coverage (LeBlanc and De Sloover 1970). The IAP value for a particular site is calculated by considering the abundance and toxiphobia value of relevant species, reflecting their sensitivity to a given pollutant (Kricke and Loppi 2002). However, the use of this index is problematic, because more than 20 different formulas are available (Conti and Cecchetti 2001). Moreover, the IAP also includes those species that increase in response to pollution, such as the extremely tolerant lichen species Lecanora conizaeoides Cromb.

9: Lichens as ecological indicators to track atmospheric changes: future challenges

(Gilbert 1973). In such cases, calculations of the IAP are biased by the single presence of a tolerant species. An alternative method for evaluating lichen diversity (Kricke and Loppi 2002) was subsequently developed and formed the basis for establishing guidelines in Germany. This method makes use of standard sampling conditions and employs a sampling grid to estimate lichen frequency (Asta et al. 2002). Its advantages are that it greatly reduces ambiguities arising from sampling methods and allows the calculation of lichen species frequency. The same method has been used elsewhere in Europe and has enabled comparisons of results obtained in different European countries. However, the identification of lichen species can be challenging and expertise is not available worldwide. Thus, as a surrogate of total lichen species diversity, richness, coverage or frequency, the identification of only macro-lichen species was suggested, which is not only much easier but also much less time-consuming. But adoption of this method is based on the assumption that macro- and micro-lichen groups respond to the same disturbance in a very similar manner, which has yet to be confirmed for a series of environmental drivers (Bergamini et al. 2005).

2.  Standardised methods to sample lichen diversity The many different methods used to assess lichen diversity, both on a regional and a national level, have hindered comparisons of the data obtained in those studies. In the current context of global change, we need to evaluate both the impact of global change drivers, such as climate change, and the global effect of political measures (the Kyoto protocol, for example) adopted in response. Only by the development of standardised methodologies using well defined metrics can comparisons be made at broad spatial and temporal scales. In Europe and the USA, two methods are routinely used at the respective continental level for determinations of epiphytic lichen diversity: 1 The European standard method, recently adopted under the Comité Européen de Normalization (CEN) framework (Ambient air – Biomonitoring with lichens – Assessing epiphytic lichen diversity. European Standard EN 16413:2014), has gained broad acceptance in Europe (Giordani et al. 2012; Pinho et al. 2012a, 2014). It consists of the placement of a size-standardised grid on the four main exposure sides of a tree trunk (Fig. 9.1). All lichen species (including crustose, fruticose and foliose species) occurring within the grid are recorded and the resulting data are used to calculate the LDV (lichen diversity value) index, which is based on species frequency. The LDV index is usually presented as the mean value for each lichen species on all the trees sampled per sampling site (the number of trees per plot and the size of the plot are decided upon according to the environmental problem being studied). 2 The US Forest Service (USFS) has applied its own method (USDA Forest Service 2011) over the last 20 years to monitor air quality in all states in the country. It is based on the presence and abundance of macro-lichens (foliose and fruticose species; crustose species are excluded) detected on all the trees present in a circular plot with a radius of 36.6 m (total area of the lichen plot is ~4200 m2). The plot is surveyed for a maximum of 2 h and the abundance of each species is estimated according to an established scale. These methods differ in two main aspects: (1) the USFS method includes only macrolichens, while the European method records all lichen species present in the grid; (2) the USFS method is based on an estimation of abundance, whereas the European method

81

82

Indicators and Surrogates of Biodiversity and Environmental Change

Fig. 9.1.  European standard for monitoring lichen diversity. A 50 × 10 cm grid (divided into five 10 × 10 cm grids) is placed on the four main aspects of a tree trunks, chosen based on standard criteria. The estimation of lichen frequency is unambiguous, as all lichens species found within the grid are identified. The number of squares occupied by each species is recorded and represents the frequency of each species.

relies on an unambiguous count of species frequency. Thus, datasets can be compared within Europe and the USA, but not between two continents.

3.  Spatial models obtained from lichen diversity metrics with high spatial resolution As ecological indicators, lichens can provide information about pollution effects at a greater spatial resolution than methods based on monitoring stations or physico-chemical analysis. The low cost of lichen sampling allows the use of a higher number of sites, yielding datasets with high spatial resolution (Pinho et al. 2008a). High-resolution mapping has allowed the detection of areas polluted by unidentified pollution sources, such as at quarries (Branquinho et al. 2008) and resulting from agricultural activities (Pinho et al. 2008b), and has pinpointed sites that merit human health studies. The effects of different environmental factors on lichen diversity metrics may vary with distance. For example, depending on the dispersion range of the specific particle type, the influence of several pollutant sources was shown to range from a few hundred metres to several kilometres (Pinho et al. 2008a). The use of high-spatial-resolution mapping is especially important for evaluating pollutants with short-range dispersion, including those with very local dispersion, such as ammonia (Sutton et al. 1998), and for calculating the critical levels that impact on biodiversity, even when most impacts occur less than 500 m

9: Lichens as ecological indicators to track atmospheric changes: future challenges

from the source (Pinho et al. 2012b). This is the case with dust, which is very common in drylands (Loppi and Pirintsos 2000). Its effects are restricted to a local scale (< 2000 m) and were quantified using lichen diversity (Branquinho et al. 2008; Pinho et al. 2008b). Datasets with high spatial resolution more readily provide reliable statistical relationships (models) between lichen diversity and other environmental variables, such as land use (determined using remote sensing images). Land use, however, is much easier to measure than lichen diversity and can then be applied to larger yet similar areas using models. In urban areas, lichen diversity differs according to land use, which can change radically over very short distances. For example, there is much greater lichen diversity in residential areas and green spaces and very little along roads (Llop et al. 2012).

4.  Knowledge of the ecological response of lichen diversity to single pollutants The atmospheric concentration of sulphur dioxide was estimated using the pattern of lichen diversity. A 10-point scale was constructed and tested in England and Wales before its adoption worldwide (Hawksworth and Rose 1970). At the time, validation of this tool was very important given the effects of sulphur dioxide pollution on human health and the lack of monitoring stations to measure the levels of this pollutant in the environment. The use of lichens in human health research began with the influential work of Cislaghi and Nimis (1997), who correlated the spatial distribution of a lichen diversity index with the spatial distribution of deaths due to respiratory diseases in 662 municipalities of the Veneto region of Italy between 1981 and 1988. The authors found a striking correlation between lichen diversity and lung cancer mortality in males aged 55 years. The results suggested that concentrations of sulphur dioxide and related particles sufficient to alter lichen diversity also have negative effects on human health. Much is known about the effects of ammonia on lichen diversity. Lichen diversity in response to an atmospheric ammonia gradient exhibits two different responses: with increasing atmospheric ammonia levels, species that are oligotrophic (nitrogen intolerant) decrease in number and coverage whereas those that are nitrophytic (nitrogen lover) increase (Fig. 9.2). According to some authors (Pinho et al. 2011), these two lichen groups can be defined as functional response groups (oligotrophic and nitrophytic) whose presence correlates significantly with atmospheric ammonia (Geiser et al. 2010). Measures of functional groups are not site-specific (unlike species) and may provide a universal tool for assessing the atmospheric concentrations of ammonia and perhaps other potential pollutants. 5.  An understanding of the integrated responses of lichens to spatial and temporal atmospheric changes The observed patterns of lichen diversity in a region are spatially and temporally integrated responses to pollutants and other environmental factors (temperature, water availability, light). This is particularly evident in large geographic areas where, for example, a broad spectrum of different sources of atmospheric change occur among different climatic regions. A recent study evaluated the integrated effect of nitrogen and climate change in the Northern Rocky Mountains (Colorado, USA), where over the last century the average annual temperature has increased 0.74°C and annual precipitation has decreased by 80–280 mm (McMurray et al. 2014). The authors of that study constructed a model that tracked changes in climate and nitrogen pollution based on changes in lichen communities over time. The model can be used to identify probable sensitive or impacted habitats, thereby providing key information on natural resource management and conservation.

83

84

Indicators and Surrogates of Biodiversity and Environmental Change

Fig. 9.2.  Effect of atmospheric ammonia on lichen species. The species shown are common on sites with very low to high atmospheric ammonia concentrations (from A, the lowest concentration, to F, the highest concentration). At low concentrations, mostly oligotrophic species are present; at high concentrations, most of the species are nitrophytic.

Because lichen are long-lived organisms, their response to multiple environmental changes occurs over time as well. Thus, lichen diversity measured on a single occasion is also the result of environmental changes that occurred in the past. Consider, for example, a strong atmospheric disturbance (e.g. fire) in a natural area, which causes a decrease in lichen diversity and and/or coverage. If this disturbance is an isolated one, then with time the lichen community will fully recover, thus no legacy effect is observed. However, if the site is permanently subject to one or more disturbances the decrease in the lichen diversity will likewise be permanent (Augusto et al. 2013). If the disturbance has a legacy effect, then recovery at the site may be prolonged. For example, in soil contaminated by heavy metals, the recovery of lichen diversity is likely to be slow and also will depend on whether the climate favours lichen growth. The potential for legacy effects on lichen diversity is an advantage in studies dealing with the turnover of persistent pollutants in the environment.

Knowledge gaps 6.  Comparative studies using lichens in different regions of the world The recognition of generalised relationships between environmental drivers and their effects on ecosystems is the goal of using ecological indicators. However, this can be accomplished only if studies are replicated in different regions of the world. This was the case in the assessment of nitrogen effects, in which the lichen biodiversity data were applied in different regions and under different conditions allowing generalisations to be made about the causal relationships between nitrogen levels and their ecosystem impact. By contrast, the impacts of climate change remain largely unknown, with only limited (mostly European) studies to date (Aptroot and Van Herk 2007; van Herk et al. 2002). Pollution and climate are drivers acting at a global scale and their relationships and effects will be best understood if these methodologies are used in different regions, thereby allowing comparisons at larger geographic scales.

9: Lichens as ecological indicators to track atmospheric changes: future challenges

7.  Data on lichen functional diversity The generalisation of relationships between global change drivers and effects depends not only on standardised sampling procedures but also on a routine, uniform and repeatable method to analyse the results. Species-related indices, such as species richness, have been widely employed as a quantitative method to analyse the response of species diversity to different environmental drivers. However, these indices cannot be used to quantify global change drivers. This is because: (1) species may have limited distributions, impairing comparisons at larger geographic scales; and (2) the indices do not consider the redundancy of species in an ecosystem. Several studies have shown that functional diversity is better than species richness for quantifying ecosystem functioning and/or the response to environmental drivers (Lavorel et al. 2011). A functional trait approach not only accounts for species redundancy but it also has the potential to be both universal and applicable at large spatial scales, because it is not linked to species per se. A functional trait is a characteristic of an organism that is relevant to its response to the environment and/or its effects on ecosystem functioning (Díaz and Cabido 2001). The value and range of a functional trait in a given ecosystem are, by definition, a measure of functional diversity (Díaz and Cabido 2001). Promising results have been achieved using the functional characteristics of plants as ecological indicators (Lavorel 2013). However, the use of lichen traits, rather than species richness, as an ecological indicator of atmospheric changes in pollution and climate is far less developed and thus far consists of only a few studies that have focused on nitrogen deposition effects (Pinho et al. 2012a) and climate (Giordani et al. 2014; Marini et al. 2011; Matos et al. 2015). 8.  An international lichen-trait database Functional diversity assessments can provide a crucial tool to assess the effects of drivers of global change. However, there is no international database covering lichen traits. Lichens are poikilohydric organisms and their interaction with water availability in the environment determines their physiological activity. The photobiont component is particularly responsive to water availability, while green-algae-containing lichens are able to carry out photosynthesis using water vapour alone. Cyan lichens need liquid water and therefore have a greater presence as a functional group in wetter areas. Growth form is also linked to the way in which lichens absorb both water and particles from the atmosphere. The shrublike form of fruticose lichens explains their efficiency in capturing both water vapour and particles from the atmosphere. Thus, this lichen functional group tends to be more frequent in areas with high relative humidity, but it is quite sensitive to high levels of pollution. This combination of physiological features accounts, at least in part, for the high sensitivity and responsiveness of lichens to climate change and pollution and supports the use of lichens to monitor spatial and temporal variation in both. The trait type of algae and growth form (Fig. 9.3) are easily measured and do not require that the lichens also be identified at the species level. Other important traits of lichens are their ability to synthesise numerous chemical compounds, their water-retention capacity, the ratios of photobiont/fungal biomass, and their tolerance of eutrophication or other forms of pollution. However, these traits are less easily identifiable. In southern Europe, most researchers base their trait classification on the Italian database (Nimis and Martellos 2008). However, as a national database, it does not include all European species. This database classifies species into response traits (e.g. eutrophication tolerance or light preferences) according to expert knowledge based on ecological performance. Thus far, it has been successfully applied to monitor responses to nitrogen and

85

86

Indicators and Surrogates of Biodiversity and Environmental Change

Fig. 9.3.  Example of lichen traits. Lichen growth form provide a basis for grouping species (crustose in the upper image, fruticose on the left, and foliose species on the right).

climate (Pinho et al. 2011; Matos et al. 2015). In the USA, a similar eutrophication tolerance index has allowed researchers to compare their results regarding nitrogen deposition. This highlights the importance of collecting data on lichen traits within an international database that can allow a functional trait approach at wider geographic scales.

9.  Disentangling the effects of multiple pollutants on lichen diversity Species are influenced by multiple environmental factors working at different spatial scales. These complex relationships between biodiversity variables and environmental factors are increasingly being assessed from a spatial perspective. The effects on biodiversity of multiple atmospheric pollutants have been examined in only a few studies (Pinho et al. 2014; Ribeiro et al. 2013) that took advantage of the different spatial scales in which pollutants affect biodiversity, which in turn is related to pollutant dispersion ranges. However, when several pollutants or other environmental drivers are present simultaneously, interactions between them are likely. For example, deposition of ammonium in soil results in a decrease in soil pH, which increases aluminium availability to plants. The effects of these types of interactions in lichens remain largely unaccounted for, as do those of interactions between climate drivers and pollution drivers in general (Seed et al. 2013). Yet, as the

9: Lichens as ecological indicators to track atmospheric changes: future challenges

c­limate continues to change, they will become increasingly important at regional and global scales. Future research must include studies on how lichens, as ecological indicators, respond to the combined effects of pollutants and global change drivers. An additional consideration is that mapping of the effects of pollutants at high spatial resolution poses problems with respect to statistical analyses. This is because they must incorporate different spatial scales, such as background pollutants superimposed on local pollutants. Geostatistical analysis can transform this problem into an opportunity, because a multi-scale analysis allows a determination of the range of influence of several sources, thus separating the effects of long-range gaseous pollutants from those of short-range particulate pollutants (Pinho et al. 2008a; Ribeiro et al. 2013).

10.  A definition of lichen thresholds and their inclusion in environmental legislation There are several limitations that hinder the use of lichens as ecological indicators and therefore their use in legislative decision making about environmental protection and for studies on human health. Perhaps the most important is the lack of threshold values for several pollutants. Values exceeding these thresholds result in potentially irreversible damage to the ecosystem, human health or both. Based on data from lichen diversity, the lowest critical levels of atmospheric ammonia have been established in Europe (1 µg NH3/ m3) and the USA (3 kg N/ha), contributing to the protection of ecosystem services and functions in both natural and semi-natural ecosystems (Geiser et al. 2010; Pinho et al. 2012b). Nonetheless, for most pollutants and for other geographic regions, similar assessments are lacking and much more challenging. Without these thresholds that relate lichen patterns to quantitative impacts on ecosystem goods and services or to human health, it will be impossible to broaden the use of lichens as ecological indicators.

Conclusions Lichens can be used to track atmospheric changes. Thus far, much is known about the role of lichens as ecological indicators and their value has been widely demonstrated. The lichen metrics best suited for particular purposes are already recognised (at least in Europe and the USA), and standardised sampling methods have been validated. Lichens integrate atmospheric changes in space and over time and therefore reflect environmental quality. With the appropriate metrics, data at high spatial resolution can be obtained, which in turn enable the construction of reliable spatial models. The responses of lichens to certain pollutants, such as sulphur dioxide and ammonium, are also well known. In the case of nitrogen, this has allowed critical threshold levels of nitrogen to be defined that can provide the basis for legislative decision making aimed at environmental protection at the ecosystem level. Several gaps in our knowledge remain that prevent the broader use of lichens as ecological indicators of atmospheric changes. First, although standard sampling methodologies are available in Europe and in the USA, replicate studies in different regions and under different condition are lacking, but they are essential to identify generalised relationships between global change drivers and effects. Second, although there is broad agreement that functional diversity is a reliable basis for biodiversity related indices, data on lichen functional diversity are still scarce and there is, as yet, no international database of lichen traits. Third, although the individual effects of global change drivers are well known, we still need to understand how they interact, to better interpret their impacts on ecosystems. Finally, information must be synthesised to define thresholds for legislative purposes.

87

88

Indicators and Surrogates of Biodiversity and Environmental Change

References Aptroot A, van Herk CM (2007) Further evidence of the effects of global warming on lichens, particularly those with Trentepohlia phycobionts. Environmental Pollution 146(2), 293– 298. doi:10.1016/j.envpol.2006.03.018. Asta J, Erhardt W, Ferretti M, Fornasier F, Kirschbaum U, Nimis PL, et al. (2002) Mapping lichen diversity as an indicator of environmental quality. In Monitoring with Lichens – Monitoring Lichens Vol. 4. (Eds P Nimis, C Scheidegger and P Wolseley) pp. 273–279. Kluwer Academic, The Netherlands. Augusto S, Maguas C, Branquinho C (2013) Guidelines for biomonitoring persistent organic pollutants (POPs), using lichens and aquatic mosses – a review. Environmental Pollution 180, 330–338. doi:10.1016/j.envpol.2013.05.019. Bergamini A, Scheidegger C, Stofer S, Carvalho P, Davey S, Dietrich M, et al. (2005) Performance of macrolichens and lichen genera as indicators of lichen species richness and composition. Conservation Biology 19(4), 1051–1062. Branquinho C (2001) Lichens. In Metals in the Environment: Analysis by Biodiversity. (Ed. MNV Prasad) pp. 117–158. Marcel Dekker, New York. Branquinho C, Gaio-Oliveira G, Augusto S, Pinho P, Máguas C, Correia O (2008) Biomonitoring spatial and temporal impact of atmospheric dust from a cement industry. Environmental Pollution 151(2), 292–299. doi:10.1016/j.envpol.2007.06.014. Branquinho C, Pinho P, Dias T, Cruz C, Máguas C, Martins-Loução MA (2010) Lichen transplants at our service for atmospheric NH3 deposition assessments. In Biology of Lichens – Symbiosis, Ecology, Environmental Monitoring, Systematics, Cyber Applications. (Eds TH Nash III, L Geiser, B McCune, D Triebel, AM Tomescu and W Sanders) pp. 103­–112. Schweizerbart Science Publishers, Stuttgart, Germany. Cislaghi C, Nimis PL (1997) Lichens, air pollution and lung cancer. Nature 387, 463–464. doi:10.1038/387463a0. Conti ME, Cecchetti G (2001) Biological monitoring: lichens as bioindicators of air pollution assessment - a review. Environmental Pollution 114(3), 471–492. doi:10.1016/S02697491(00)00224-4. Davies L, Bates JW, Bell JNB, James PW, Purvis OW (2007) Diversity and sensitivity of epiphytes to oxides of nitrogen in London. Environmental Pollution 146(2), 299–310. doi:10.1016/j.envpol.2006.03.023. Díaz S, Cabido M (2001) Vive la différence: plant functional diversity matters to ecosystem processes. Trends in Ecology & Evolution 16(11), 646–655. doi:10.1016/S01695347(01)02283-2. Geiser LH, Jovan SE, Glavich DA, Porter MK (2010) Lichen-based critical loads for atmospheric nitrogen deposition in Western Oregon and Washington Forests, USA. Environmental Pollution 158(7), 2412–2421. doi:10.1016/j.envpol.2010.04.001. Gilbert O (1973) Lichens and air pollution. In Lichens. 1st edn. (Eds V Ahmadjian and M Hale) pp. 443–472. Academic Press, New York. Giordani P, Brunialti G, Bacaro G, Nascimbene J (2012) Functional traits of epiphytic lichens as potential indicators of environmental conditions in forest ecosystems. Ecological Indicators 18, 413–420. doi:10.1016/j.ecolind.2011.12.006. Giordani P, Incerti G, Rizzi G, Rellini I, Nimis PL, Modenesi P (2014) Functional traits of cryptogams in Mediterranean ecosystems are driven by water, light and substrate interactions. Journal of Vegetation Science 25(3), 778–792. doi:10.1111/jvs.12119. Hawksworth D (2002) Bioindication: calibrated scales and their utility. In Monitoring with Lichens – Monitoring Lichens. (Eds P Nimis, C Scheidegger and P Wolseley) pp. 11–20. Kluwer Academic, The Netherlands.

9: Lichens as ecological indicators to track atmospheric changes: future challenges

Hawksworth DL, Rose F (1970) Qualitative scale for estimating sulphur dioxide air pollution in England and Wales using epiphytic lichens. Nature 227(5254), 145–148. doi:10.1038/227145a0. Kricke R, Loppi S (2002) Bioindication: the IPA approach. In Monitoring with Lichens – Monitoring Lichens. (Eds P Nimis, C Scheidegger and P Wolseley) pp. 21–37. Kluwer Academic, The Netherlands. Lavorel S (2013) Plant functional effects on ecosystem services. Journal of Ecology 101(1), 4–8. doi:10.1111/1365-2745.12031. Lavorel S, Grigulis K, Lamarque P, Colace MP, Garden D, Girel J, et al. (2011) Using plant functional traits to understand the landscape distribution of multiple ecosystem services. Journal of Ecology 99(1), 135–147. doi:10.1111/j.1365-2745.2010.01753.x. LeBlanc F, De Sloover J (1970) Relation between industrialization and the distribution and growth of epiphytic lichens and mosses in Montreal. Canadian Journal of Botany 48(8), 1485–1496. doi:10.1139/b70-224. Llop E, Pinho P, Matos P, Pereira MJ, Branquinho C (2012) The use of lichen functional groups as indicators of air quality in a Mediterranean urban environment. Ecological Indicators 13(1), 215–221. doi:10.1016/j.ecolind.2011.06.005. Loppi S, Pirintsos SA (2000) Effect of dust on epiphytic lichen vegetation in the Mediterranean area (Italy and Greece). Israel Journal of Plant Sciences 48(2), 91–95. doi:10.1560/EK72-KP5W-U3Q3-AV5Q. Marini L, Nascimbene J, Nimis PL (2011) Large-scale patterns of epiphytic lichen species richness: photobiont-dependent response to climate and forest structure. The Science of the Total Environment 409(20), 4381–4386. doi:10.1016/j.scitotenv.2011.07.010. Matos P, Pinho P, Aragón G, Martínez I, Nunes A, Soares AMVM, Branquinho C (2015) Lichen traits responding to aridity. Journal of Ecology 103(2), 451–458. doi:10.1111/13652745.12364. McCune B (2000) New frontiers in bryology and lichenology – lichen communities as indicators of forest wealth. The Bryologist 103(2), 353–356. doi:10.1639/00072745(2000)103[0353:LCAIOF]2.0.CO;2. McMurray JA, Roberts DW, Geiser LH (2015) Epiphytic lichen indication of nitrogen deposition and climate in the northern Rocky Mountains, USA. Ecological Indicators 49(8), 154–161. doi:10.1016/j.ecolind.2014.10.015. Munzi S, Correia O, Silva P, Lopes N, Freitas C, Branquinho C, et al. (2014) Lichens as ecological indicators in urban areas: beyond the effects of pollutants. Journal of Applied Ecology 51, 1750–1757. doi:10.1111/1365-2664.12304. Nimis P, Martellos S (2008) ITALIC. The Information System on Italian Lichens. Version 4.0. Department of Biology, IN4.0/1, University of Trieste, . Pinho P, Augusto S, Maguas C, Pereira MJ, Soares A, Branquinho C (2008a) Impact of neighbourhood land-cover in epiphytic lichen diversity: analysis of multiple factors working at different spatial scales. Environmental Pollution 151(2), 414–422. doi:10.1016/j. envpol.2007.06.015. Pinho P, Augusto S, Martins-Loucao MA, Pereira MJ, Soares A, Maguas C, et al. (2008b) Causes of change in nitrophytic and oligotrophic lichen species in a Mediterranean climate: impact of land cover and atmospheric pollutants. Environmental Pollution 154(3), 380–389. doi:10.1016/j.envpol.2007.11.028. Pinho P, Branquinho C, Máguas C (2010) Modeling ecology of lichens communities based on photobiont type in response to potential solar radiation and neighborhood land-use. In Biology of Lichens – Symbiosis, Ecology, Environmental Monitoring, Systematics, Cyber Applications. (Eds TH Nash III, L Geiser, B McCune, D Triebel, AM Tomescu and W Sanders) pp. 149–160. Schweizerbart Science Publishers, Stuttgart, Germany.

89

90

Indicators and Surrogates of Biodiversity and Environmental Change

Pinho P, Dias T, Cruz C, Tang YS, Sutton MA, Martins-Loucao MA, et al. (2011) Using lichen functional diversity to assess the effects of atmospheric ammonia in Mediterranean woodlands. Journal of Applied Ecology 48(5), 1107–1116. doi:10.1111/j.1365-2664.2011. 02033.x. Pinho P, Bergamini A, Carvalho P, Branquinho C, Stofer S, Scheidegger C, et al. (2012a) Lichen functional groups as ecological indicators of the effects of land-use in Mediterranean ecosystems. Ecological Indicators 15(1), 36–42. doi:10.1016/j. ecolind.2011.09.022. Pinho P, Theobald MR, Dias T, Tang YS, Cruz C, Martins-Loucao MA, et al. (2012b) Critical loads of nitrogen deposition and critical levels of atmospheric ammonia for semi-natural Mediterranean evergreen woodlands. Biogeosciences 9(3), 1205–1215. doi:10.5194/ bg-9-1205-2012. Pinho P, Llop E, Ribeiro M, Cruz C, Soares A, Pereira M, et al. (2014) Tools for determining critical levels of atmospheric ammonia under the influence of multiple disturbances. Environmental Pollution 188, 88–93. doi:10.1016/j.envpol.2014.01.024. Ribeiro MC, Pinho P, Llop E, Branquinho C, Sousa AJ, Pereira MJ (2013) Multivariate geostatistical methods for analysis of relationships between ecological indicators and environmental factors at multiple spatial scales. Ecological Indicators 29, 339–347. doi:10.1016/j.ecolind.2013.01.011. Ribeiro MC, Pinho P, Llop E, Branquinho C, Soares A, Pereira MJ (2014) Associations between outdoor air quality and birth weight: a geostatistical sequential simulation approach in Coastal Alentejo, Portugal. Stochastic Environmental Research and Risk Assessment 28(3), 527–540. doi:10.1007/s00477-013-0770-6. Seed L, Wolseley P, Gosling L, Davies L, Power SA (2013) Modelling relationships between lichen bioindicators, air quality and climate on a national scale: Results from the UK OPAL air survey. Environmental Pollution 182, 437–447. doi:10.1016/j.envpol.2013.07.045. Sutton MA, Milford C, Dragosits U, Place CJ, Singles RJ, Smith RI, et al. (1998) Dispersion, deposition and impacts of atmospheric ammonia: quantifying local budgets and spatial variability. Environmental Pollution 102, 349–361. doi:10.1016/S0269-7491(98)80054-7. USDA Forest Service (2011) Forest Inventory and Analysis Phase 3 Field Guide, version 5.1. USDA Forest Service, Arlington, VA, . van Herk CM, Aptroot A, Van Dobben H (2002) Long-term monitoring in the Netherlands suggests that lichens respond to global warming. Lichenologist (London, England) 34(2), 141–154. doi:10.1006/lich.2002.0378.

10

Approaches, potential and pitfalls of applying bioindicators in freshwater ecosystems Jani Heino

Things we know 1 2 3 4 5

Different indicators tell us about different things. Indices developed in one region do not necessarily work in another region. Different groups of organisms respond differently to environmental factors. Spatial extent affects the utility of indicators. Most researchers have their own favourite indicators.

Knowledge gaps 6 7 8 9

How to combine information from multiple indicators. How to consider the regional and environmental context. How to use different groups of organisms as indicators at the same time. How to integrate knowledge of ecological processes into the application of indicators. 10 How to think broadly when assessing ecosystem state and measuring biodiversity.

Introduction Freshwater ecosystems harbour high levels of biodiversity compared with their areal extent (Dudgeon et al. 2006; Vörösmarty et al. 2010) and are highly sensitive to multiple anthropogenic stressors (Dudgeon et al. 2006; Ormerod et al. 2010). However, we do not often know enough to direct our limited resources to guide decision making in relation to the state of freshwater ecosystems. A means to achieve a better understanding of management, restoration and conservation is to develop indicators (McGeoch 1998; Caro 2010), which should provide integrated information about ecosystem health, change in biota and state of biodiversity (Table 10.1). In freshwater ecosystems, several indicators and approaches have been used in the past (Friberg et al. 2011), and most of those approaches have focused on determining if a lake or a stream has been degraded by humans (Birk et al. 2012). However, in the past, anthropogenic influences have typically been studied using indices suitable in a specific region only (Karr 1981; Karr and Chu 1999). In addition, most indices are indicators of the ecosystem health (Birk et al. 2012), while fewer indicators specifically try 91

92

Indicators and Surrogates of Biodiversity and Environmental Change

Table 10.1.  A glossary of key terms used throughout this chapter Term

Definition

Biodiversity indicator

A biodiversity indicator is a group of taxa (e.g. genus, family or order) whose diversity reflects some measure of the diversity (e.g. character richness, species richness, level of endemism) of other higher taxa in an ecosystem or set of ecosystems (McGeoch 1998).

Bioindicator

A bioindicator reflects the state of an environment (McGeoch 1998). It represents the effects of environmental changes on a habitat, a community or an ecosystem. It may also act as a surrogate of the diversity of wholesale diversity in an ecosystem or a region.

Diversity index

Incorporates various traditional diversity indices that describe the number of species and the division of individuals among the species present at a site. More recent and sophisticated diversity indices also take into account the phylogenetic relatedness of species or the functional trait differences between species in a community (Magurran 2013). Applications of diversity indices typically assume, other things being equal, that more diversity should be found in a natural ecosystem than in a stressed ecosystem.

Dispersal limitation

Dispersal limitation is manifested if species are absent in environmentally suitable sites because nearest occupied sites are located far away and organism cannot reach those sites (Heino and Peckarsky 2014).

Ecological indicator

An ecological indicator is a characteristic taxon or a set of species that is sensitive to identified environmental stressors (McGeoch 1998). It demonstrates the effect of stressors on biota, and its response is representative of those of at least some other taxa present in the habitat.

Environmental indicator

An environmental indicator is a species or group of multiple species that respond predictably to environmental disturbance or to a change in environmental state (McGeoch 1998).

Indicator and index

An indicator can be defined as a value of an index that can be easily communicated to researchers, policy makers and the public. Bioindicators are often used because indices based on the structure of ecological communities provide integrated information about ecosystem health, change in biota and degree of biodiversity (McGeoch 1998).

Mass effects

In mass effects, species are able to occur in suboptimal environmental conditions owing to high dispersal rates from environmentally more suitable sites (Leibold et al. 2004).

Multimetric index

Multimetric indices or indices of biotic integrity typically combine information from different facets of an ecological community, such as number of species, number of sensitive species, number of different trophic groups and more, leading to a single index value (Karr and Chu 1999). A high index value refers to situations where an ecosystem is not affected by putative anthropogenic stressors.

Predictive modelling

Examines the responses of biological communities to environmental degradation, aiming to reveal the degree to which community structure in a stressed ecosystem differs from that expected if the ecosystem was in a reference condition or close to a natural state (e.g. the ratio of the observed number of taxa to the expected number of taxa or O/E index in RIVPACS-type modelling; Clarke et al. 1996).

10: Approaches, potential and pitfalls of applying bioindicators in freshwater ecosystems

Term

Definition

Species sorting

In species sorting, species are selected to occur in certain environmental conditions according to their abiotic and biotic features (Leibold et al. 2004). Also, the same process is often called environmental filtering (Poff 1997).

Species trait

A species trait can be defined as a feature that reflects a species adaptation to its environment (Menezes et al. 2010). Traits may be biological (e.g. life cycle, maximum body size, lifespan or mobility) or ecological (e.g. acidity tolerance, temperature tolerance or pollution tolerance).

Taxon-specific index

For some groups of organisms, such as diatoms and benthic invertebrates, there are indices constructed using expected responses of species to a specific stressor (e.g. nutrient enrichment, acidification or organic pollution). Those indices use information on certain species, whose presence is known to be related to high nutrient concentrations or low nutrient concentrations (Birk et al. 2012). The resulting index value informs the researcher, for example, if an ecosystem is stressed by excess nutrients.

to estimate the levels of biodiversity, ecosystem functions or ecosystem services in freshwater ecosystems (Feld et al. 2009). The assessment of the state of freshwater ecosystems is typically based on the variety of organisms present (e.g. their numbers, kinds and characteristics). Those assessments have often followed one of five main types of approaches that are partly overlapping (Tables 10.1 and 10.2). First, one commonly used approach relates to the application of predictive modelling, which has been used in various regional and environmental contexts (Hawkins et al. 2000). Second, researchers have used various traditional diversity indices that describe the number of species and the division of individuals among the species present at Table 10.2.  A comparison of five general approaches for assessing the state of freshwater ecosystems. Shown are the potential responses of index values derived from each approach to region specificity, natural environmental gradients and anthropogenic stressors. Bioindication: the approach mainly relates to environmental indication (ENV), ecological indication (ECO) or biodiversity indication (BIO) following the terminology of McGeoch (1998). Means to account for natural gradients

Sensitivity to a single stressor

Sensitivity to multiple stressors

Approach

Region specificity

Response to natural gradients

Predictive modelling

Low to moderate

Moderate to high

A posteriori statistically

Moderate

Moderate to ENV, ECO high

Diversity indices

Moderate

High

A posteriori statistically

Low to moderate

Low to moderate

ECO, BIO

Species traits

Low to moderate

Moderate to high

A posteriori statistically

Moderate

Moderate

ENV, ECO

Multimetric indices

High

Low to moderate

A priori adjustments

Moderate

Moderate

ENV, ECO

Taxonspecific indices

High

Low to moderate

Targeted to specific stressors

Moderate to Low to high moderate

ENV, ECO

Bioindication

93

94

Indicators and Surrogates of Biodiversity and Environmental Change

Bioindicators

Environmental indicator

Ecological indicator

Biodiversity indicator

Indicator taxon FISH, INVERTEBRATES, MACROPHYTES, ALGAE, BACTERIA

Index O/E, DIVERSITY, TRAIT, MULTIMETRIC, TAXON SPECIFIC

Ecosystem health

Change in biota

(e.g. occurrence of nuisance algae)

(e.g. cyprinid fish abundance)

Degree of biodiversity (e.g. species richness of dragonflies)

Management decisions (e.g. A lake should be restored to attain a good ecological status)

Fig. 10.1. A schematic picture of ‘from the indicator to the management decisions’ flow in freshwater ecosystems. Bioindicators can be divided into three types (McGeoch 1998). Within each indicator type, there are potentially several indicator taxa. For each indicator taxon, an appropriate index value can be derived using five approaches, which can be used for assessing ecosystem health, change in biota or degree of biodiversity in an ecosystem. Those assessments will subsequently guide management decisions.

a site (Magurran 2013). Third, an approach focusing directly on species traits has been suggested as an appropriate means for assessing anthropogenic effects (Menezes et al. 2010) because it provides a way to understand mechanistic links between organisms and the environment at various spatial scales (Poff 1997). Fourth, researchers have developed multimetric indices or indices of biotic integrity (see Chapter 11), integrating the expected responses of ecological communities and their biological characteristics to anthropogenic impacts (Karr and Chu 1999). Fifth, taxon-specific indices are commonly used in studies of diatoms and macroinvertebrates in both standing and running waters (Birk et al. 2012). While each of the five approaches may reveal the effects of anthropogenic perturbation on freshwater ecosystems, they should not be applied without considering: (1) natural environmental gradients; (2) ecological processes additional to anthropogenic impacts; and (3) spatial scale (Table 10.2). This chapter will deal with indicators in freshwater ecosystems, focusing specifically on organism-level indicators of ecosystem state and biological diversity (Fig. 10.1).

Things we know 1.  Different indicators tell us about different things There is no universal approach, indicator or index that informs us about the ecosystem health, change in biota and biodiversity at the same time. Hence, the use of predictive

10: Approaches, potential and pitfalls of applying bioindicators in freshwater ecosystems

modelling, diversity indices, species traits, multimetric indices or taxon-specific indices to gauge the state of freshwater ecosystems provide different and complementary information about the ecosystems studied (Hering et al. 2006; Hawkins et al. 2010). Furthermore, while predictive modelling, diversity indices, species traits and multimetric indices may provide more comprehensive information than taxon-specific indices about wholesale ecosystem state, at least the first three are affected by natural environmental gradients, which should be accounted for statistically or focus explicitly on sampling designs minimising natural variability (Hawkins et al. 2010; Heino 2013). In contrast, multimetric indices and taxon-specific indices may be more immune to natural environmental gradients because they are either adjusted to account for natural gradients (Karr and Chu 1999) or targeted to indicate the effects of a certain stressor (Friberg et al. 2011), respectively.

2.  Indices developed in one region do not necessarily work in another region The five approaches outlined above are likely to differ in how well they can be applied in different regions. This is because the regional species pools, gradients in natural environmental conditions and heterogeneity in anthropogenic stressors vary among regions (Heino 2013). For example, while the approaches based on predictive modelling of the responses of whole communities to anthropogenic perturbation can account for natural biological and environmental variation a posteriori in any region (Hawkins et al. 2010), many multimetric indices typically have been a priori adjusted to account for natural environmental variation in a given region (Karr and Chu 1999). Hence, predictive modellingbased approaches can be, and have been, applied in different regions (Hawkins et al. 2000), whereas different multimetric indices and taxon-specific indices should not be applied directly outside of the region in which they were originally developed. This is because both multimetric indices and taxon-specific indices are affected by the species available in the regional species pool and, thus, they are not often transferable among regions (Karr and Chu 1999). In the past, researchers were perhaps too eager to apply various such indices in regions far away from their region of origin. This has led to findings that a certain index does not work at all or, in the worst case, may even have provided wrong information about the ecosystem state. 3.  Different groups of organisms respond differently to environmental factors Due to differences in basic biological characteristics, such as body size, trophic position, life cycle and dispersal ability, different groups of organisms typically respond to different ecological factors and may even show contrasting responses to the same ecological gradient (Heino 2010; Padial et al. 2012). For example, bacteria may be affected by fine-scale variation in sediment conditions on the stream bottom, whereas large fish may respond mainly to large-scale variation in stream size and discharge. Those responses stem from the fact that ecologically disparate organisms typically perceive their environment at different spatial and temporal scales. Hence, for example, indices based on bacteria may be good indicators of environmental changes in situations where change happens rapidly because those organisms can respond almost immediately to altered conditions. In contrast, fish may show a slower response time because they have longer life cycles than bacteria. 4.  Spatial extent affects the utility of indicators The spatial extent of a region where the effects of environmental changes on ecosystems are evaluated may affect our perspectives on how well indicators work. For example, spatial

95

96

Indicators and Surrogates of Biodiversity and Environmental Change

extent has been shown to affect the performance of predictive models used in bioassessment (Ode et al. 2008), which may be related to differences in regional species pools, underlying natural environmental gradients or heterogeneity in anthropogenic impacts (Heino 2013). One could assume that increasing spatial extent increases environmental gradient length and, hence, should lead to stronger environmental filtering (Jackson et al. 2001) or species sorting (Leibold et al. 2004). Actually, the idea of environmental filtering underlies all bioassessment because organisms are supposed to respond to prevailing environmental conditions and changes therein (Friberg et al. 2011). Bioassessment may, however, fall short if other processes that vary with spatial extent are not taken into account (Heino 2013).

5.  Most researchers have their own favourite indicators Researchers use their favourite organismal groups and indices for inferring anthropogenic effects, and typically those indicators appear to be the best available. Hence, researchers may show some subjectivity because they are familiar with a certain organismal group and may easily ignore information provided by other groups. Rather than arguing which group is the best among those available to indicate anthropogenic effects, ichthyologists, entomologists, botanists and phycologists should aim to integrate their findings and try to generate the best information about the state of freshwater ecosystems for environmental managers and policy makers.

Knowledge gaps 6.  How to combine information from multiple indicators Because multiple anthropogenic stressors are often active simultaneously (Ormerod et al. 2010; Friberg et al. 2011), researchers have at least two potential choices. They may examine ecosystem state based on information provided by index values from each of the five approaches outlined above. Alternatively, they may focus on multiple indices within one approach, for example, using information from various diversity indices or various taxonspecific indices. Different indices may not always be strongly correlated. Hence, researchers have problems deciding which index provides the most accurate information about the state of an ecosystem. Further development of indicators should also take into account the simultaneous influences of multiple stressors on freshwater ecosystem. This would entail testing and applying indices sensitive to wholesale changes in freshwater ecosystems, rather than those responding to a specific stressor only. 7.  How to consider the regional and environmental context Researchers should always consider if the index they are going to apply in a region might actually work there. If the index has been developed in another region, it may need to be modified to take account of species available in the regional species pool and the most important natural environmental gradients in the new region (Karr and Chu 1999). Then, the utility of those indices should be tested before their use in guiding management decisions. In contrast, the use of predictive modelling-based approaches may not face the same problems of non-transferability (Hawkins et al. 2000). However, potential differences in the performance and accuracy of those models among regions may still stem from differences in regional species pools and differences in influential environmental gradients (Ode et al. 2008). Those differences should be considered explicitly when comparing large geographical areas, and novel analytical methods should be applied to overcome the limitations of past studies. New methods should be

10: Approaches, potential and pitfalls of applying bioindicators in freshwater ecosystems

able to directly associate ecological processes active in a region (e.g. the interactions between dispersal processes and species sorting) with the assessment of the state of freshwater ecosystems (Heino 2013).

8.  How to use different groups of organisms as indicators at the same time Information from multiple groups of organisms should be used to infer the ecological state of an ecosystem (Heino 2010; Johnson and Hering 2010). Such multi-group assessments of ecosystem state are becoming increasingly common. For example, the member countries of the European Union are required to assess the ecological state of freshwater ecosystems based on multiple groups of organisms (Johnson et al. 2006). These groups include fish, benthic macroinvertebrates, macrophytes and benthic diatoms for running waters (Hering et al. 2006), and fish, macrophytes, benthic macroinvertebrates and phytoplankton for lakes (Brucet et al. 2013). However, although they are based on surveys of multiple groups, some assessment systems rely only on the information from the group of organisms or abiotic parameters giving the lowest status for an ecosystem. It is questionable if such ‘oneout, all-out’ assessments are appropriate because other groups may show a more consistent picture of ecosystem state. Furthermore, if we desire an index of wholesale biodiversity and not only one of ecological state, the assessment should be based on information from multiple groups. There is the issue of cost-efficiency, however, and assessments based on multiple groups may not always be feasible owing to high costs. Future studies should explicitly combine the issue of cost-efficiency with the notion that a single group does not tell us enough about ecosystem state or biodiversity. 9.  How to integrate knowledge of ecological processes into the application of indicators Although environmental filtering or species sorting arguably is the most important force shaping species’ distributions (Friberg et al. 2011), ignoring other forces is questionable and possibly a major pitfall in bioassessment (Heino and Peckarsky 2014). Past bioassessment research paid too little attention to the facts that: (1) a species may be absent in an environmentally suitable site because it has never reached that site owing to dispersal limitation or barriers; (2) it may be present at a unsuitable site (e.g. a anthropogenically stressed site) because high dispersal rates from other sites (e.g. a natural sites) allow it to occur at that site; and (3) its distribution is likely to vary temporally because it may experience local extinctions and colonisations periodically (Leibold et al. 2004). These processes may affect local ecological communities and, hence, the utility of index values used in bioassessment (Hitt and Angermeier 2011). Those notions also relate, at very large spatial scales, to the fact that indices developed in one region may not work in another region because the species pools differ between regions due to unsurmountable barriers for dispersal (Heino 2013). A solution would be to use indicators that are immune to geographical differences in regional species pools. Species traits have been suggested as a solution to this problem (Menezes et al. 2010 and see Chapter 9), although their universal applicability requires further scrutiny (Heino et al. 2013). Another solution would be to use indicator organisms, such as bacteria, whose distributions are presumably little affected by dispersal limitations (Baas-Becking 1934). New developments in microbiology may offer highly useful applications for the bioassessment of freshwater ecosystems, although the utility of microbes for such purposes needs to be assessed in the context of spatial scale and related issues.

97

98

Indicators and Surrogates of Biodiversity and Environmental Change

10.  How to think broadly when assessing ecosystem state and measuring biodiversity Rather than relying on the information provided by a single group of organisms or a single index, researchers should focus on multiple indicators to obtain a more holistic picture of nature. Different groups of organisms typically respond differently to natural environmental gradients and exhibit different responses to anthropogenic stressors (Hering et al. 2006; Heino 2010). The same often applies to different indices based on a single group (Johnson and Hering 2010; Brucet et al. 2013). Although integration of information from multiple groups and various indices is a difficult task, it may be key to the development of indicators that tell us simultaneously more about biodiversity, ecosystem health and potential of ecosystems to provide valuable services for humans. This goal may be reachable through the application of new statistical approaches. Modern multivariate methods allow complex indicator–environment relationships or indicator–spatial location relationships to be revealed, thereby increasing the utility and reliability of indicators in bioassessment. A challenge may be related to communicating those often complex statistical results to the broad audience of environmental managers, policy makers and the public.

Conclusions Using indicators to gauge the state of freshwater ecosystems is not an easy task. Multiple anthropogenic stressors and ecological processes are acting simultaneously, different indicators provide different information about environmental conditions and a reliance on a single indicator is likely to tell us only part of the story about what is going on in nature. The use of multiple group of organisms is a step towards more holistic and, hopefully, more appropriate use of indicators in freshwater ecosystems. Such a multi-group approach is not without problems. This is because using multiple groups means multiple index values, and deciding which index provides the most precise or the most accurate information about the state of freshwater ecosystems is complicated. On the one hand, indicators that are too simple may not provide enough information to guide management decisions. On the other hand, too complex indicators may not be easily communicated to policy makers and the public. There is thus much scope for the further development of indicators. Hence, approaches and associated indices are sorely needed that: (1) distinguish accurately between natural and anthropogenic variation in ecological communities; (2) are immune to chance absences or presences of single species; (3) are applicable in (and comparable among) various geographical regions; and (4) are simple enough to be readily applied. Although a single global indicator of freshwater ecosystem state is not achievable, at least regional- or national-level indicators should be developed to provide better and more reliable information to guide management, restoration and conservation needs.

Acknowledgements The writing of this chapter was supported by a grant from the Academy of Finland.

References Baas Becking L (1934) Geobiologi of inleidning tot de milienkunde. Van Stockum & Zoon, The Hague.

10: Approaches, potential and pitfalls of applying bioindicators in freshwater ecosystems

Birk S, Bonne W, Borja A, Brucet A, Courrat A, Poikane S, et al. (2012) Three hundred ways to assess Europe’s surface waters: an almost complete overview of biological methods to implement the Water Framework Directive. Ecological Indicators 18, 31–41. doi:10.1016/j. ecolind.2011.10.009. Brucet S, Poikane S, Lyche-Solheim A, Birk S (2013) Biological assessment of European lakes: ecological rationale and human impacts. Freshwater Biology 58, 1106–1115. doi:10.1111/ fwb.12111. Caro TM (2010) Conservation by Proxy. Island Press, Washington DC. Clarke RT, Furse MT, Wright JF, Moss D (1996) Derivation of a biological quality index for river sites: comparison of the observed with the expected fauna. Journal of Applied Statistics 23, 311–332. doi:10.1080/02664769624279. Dudgeon D, Arthington AH, Gessner MO, Kawabata Z, Knowler DJ, Lévêque C, et al. (2006) Freshwater biodiversity: importance, threats, status and conservation challenges. Biological Reviews of the Cambridge Philosophical Society 81, 163–182. doi:10.1017/ S1464793105006950. Feld CK, Martins da Silva P, Paulo Sousa J, De Bello F, Bugter B, Grandin U, et al. (2009) Indicators of biodiversity and ecosystem services: a synthesis across ecosystems and spatial scales. Oikos 118, 1862–1871. doi:10.1111/j.1600-0706.2009.17860.x. Friberg N, Bonada N, Bradley DC, Dunbar MJ, Edwards FK, Grey J, et al. (2011) Biomonitoring of human impacts in freshwater ecosystems: the good, the bad and the ugly. Advances in Ecological Research 44, 1–68. doi:10.1016/B978-0-12-374794-5.00001-8. Hawkins CP, Norris RH, Hogue JN, Feminella JW (2000) Development and evaluation of predictive models for measuring the biological integrity of streams. Ecological Applications 10, 1456–1477. doi:10.1890/1051-0761(2000)010[1456:DAEOPM]2.0.CO;2. Hawkins CP, Cao Y, Roper B (2010) Method of predicting reference condition biota affects the performance and interpretation of ecological indices. Freshwater Biology 55, 1066–1085. doi:10.1111/j.1365-2427.2009.02357.x. Heino J (2010) Are indicator groups and cross-taxon congruence useful for predicting biodiversity in aquatic ecosystems? Ecological Indicators 10, 112–117. doi:10.1016/j. ecolind.2009.04.013. Heino J (2013) The importance of metacommunity ecology for environmental assessment research in the freshwater realm. Biological Reviews of the Cambridge Philosophical Society 88, 166–178. doi:10.1111/j.1469-185X.2012.00244.x. Heino J, Peckarsky BL (2014) Integrating behavioral, population and large-scale approaches for understanding stream insect communities. Current Opinion in Insect Science 2, 7–13. doi:10.1016/j.cois.2014.06.002. Heino J, Schmera D, Erős T (2013) A macroecological perspective of trait patterns in stream communities. Freshwater Biology 58, 1539–1555. doi:10.1111/fwb.12164. Hering D, Johnson RK, Kramm S, Schmutz S, Szoskiewitz K, Verdonschot PFM (2006) Assessment of European streams with diatoms, macrophytes, macroinvertebrates and fish: a comparative metric-based analysis of organism response to stress. Freshwater Biology 51, 1757–1785. doi:10.1111/j.1365-2427.2006.01610.x. Hitt NP, Angermeier PL (2011) Fish community and bioassessment responses to stream network location. Journal of the North American Benthological Society 30, 296–309. doi:10.1899/09-155.1. Jackson DA, Peres-Neto PR, Olden JD (2001) What controls who is where in freshwater fish communities – the roles of biotic, abiotic, and spatial factors. Canadian Journal of Fisheries and Aquatic Sciences 58, 157–170.

99

100

Indicators and Surrogates of Biodiversity and Environmental Change

Johnson RK, Hering D (2010) Spatial congruency of benthic diatom, invertebrate, macrophyte, and fish assemblages in European streams. Ecological Applications 20, 978–992. doi:10.1890/08-1153.1. Johnson RK, Hering D, Furse MT, Clarke RT (2006) Detection of ecological change using multiple organism groups: metrics and uncertainty. Hydrobiologia 566, 115–137. doi:10.1007/s10750-006-0101-8. Karr JR (1981) Assessment of biotic integrity using fish communities. Fisheries (Bethesda, Md.) 6, 21–27. doi:10.1577/1548-8446(1981)0062.0.CO;2. Karr JR, Chu EW (1999) Restoring Life in Running Waters. Better Biological Monitoring. Island Press, Covelo, CA. Leibold MA, Holyoak M, Mouquet N, Amarasekare P, Chase JM, Hoopes MF, et al. (2004) The metacommunity concept: a framework for multi-scale community ecology. Ecology Letters 7, 601–613. doi:10.1111/j.1461-0248.2004.00608.x. Magurran AE (2013) Measuring Biological Diversity. Wiley-Blackwell, Oxford. McGeoch MA (1998) The selection, testing and application of terrestrial insects as bioindicators. Biological Reviews 73, 181–201. doi:10.1017/S000632319700515X. Menezes S, Baird DJ, Soares AMVM (2010) Beyond taxonomy: a review of macroinvertebrate trait-based community descriptors as tools for freshwater biomonitoring. Journal of Applied Ecology 47, 711–719. doi:10.1111/j.1365-2664.2010.01819.x. Ode PR, Hawkins CP, Mazor RD (2008) Comparability of biological assessments derived from predictive models and multimetric indices of increasing geographic scope. Journal of the North American Benthological Society 27, 967–985. doi:10.1899/08-051.1. Ormerod SJ, Dobson M, Hildrew AH, Townsend CR (2010) Multiple stressors in freshwater ecosystems. Freshwater Biology 55, 1–4. doi:10.1111/j.1365-2427.2009.02395.x. Padial AA, Declerck SAJ, De Meester L, Bonecker CC, Lansac-Toha FA, Rodrigues LC, et al. (2012) Evidence against the use of surrogates for biomonitoring of Neotropical floodplains. Freshwater Biology 57, 2411–2423. doi:10.1111/fwb.12008. Poff NL (1997) Landscape filters and species traits: towards mechanistic understanding and prediction in stream ecology. Journal of the North American Benthological Society 16, 391–409. doi:10.2307/1468026. Vörösmarty C, McIntyre PB, Gessner MO, Dudgeon D, Prusevich A, Green P, et al. (2010) Global threats to human water security and river biodiversity. Nature 467, 555–561. doi:10.1038/nature09440.

11

Searching for the holy grail of wetland integrity: are surrogates still relevant in conservation planning? Hamish Greig and Aram Calhoun

What we know 1 2 3 4 5

IBIs are confounded by spatio-temporal processes. IBIs measure structure but infer function. Development of IBIs requires capacity and motivation. Developed IBIs are broadly (and perhaps blindly) applied to a range of goals. IBIs have often been developed without community consultation.

Knowledge gaps 6 Acknowledge the spatio-temporal limitations of IBIs and develop protocols that incorporate these processes. 7 Evaluate the use of functional surrogates (diversity, food webs or direct measures) in IBIs. 8 Broadly assess the value of ecosystem services and risk of decline. Use traditional knowledge and global resources in areas of limited capacity. 9 Develop clear goals and tailor assessments to meet those goals. Key factors to consider: spatial, temporal and cultural context, wetland type, dynamic nature of wetlands and climate. 10 Solicit stakeholder input for conservation of natural resources and in finding solutions to complex coupled natural human systems. Provide outreach and educational opportunities.

Introduction Imagine a small, temporary pond in a verdant forest. It is teeming with life. The forest is cleared and replaced by a housing development. The pool becomes eutrophic and the forest-dependent amphibians are replaced with pollution-tolerant generalist frogs and weedy plants. You are alarmed, as you knew the pool when it was a pristine. Your new neighbours are pleased with the pool as it has pretty, flowering plants and frogs (does it really matter what kind?) their children can catch. Step back and consider this from the perspective of a 101

102

Indicators and Surrogates of Biodiversity and Environmental Change

21st century ecologist. What is ‘natural’ and desirable? What is the baseline or reference for conservation? Biological assessment techniques were developed by scientists to attempt to objectively evaluate the biotic integrity of aquatic resources before and after a human perturbation using surrogates. Embedded in the debate between the neighbours (one horrified, one pleased) over our imagined pool, and in the response of scientists tasked to assessing the pool, are the issues of: acceptable reference conditions; cultural values (including a range of goals for appropriate outcomes); geographic context; extent of degradation; and the feasibility of restoration. Water quality is likely to be the focus of this debate as it is the most noticeably changed component of the pool ecosystem. At first glance, the effect of the absent adjacent forest on the complex aquatic community and ecosystem-level interactions, that encompass both wetland and upland realms, are invisible. How will the scientists know when the pool returns to a ‘healthier’ state? Water, fundamental to life, predictably became a focus of national environmental laws during the age of environmental awareness in developed countries during the last decades of the 20th century (e.g. US Clean Water Act 1972; European Union Water Framework Directive 2003). Biological assessments (bioassessments) are tools used to evaluate the biological condition of a water body or aquatic resource that use surveys and other direct measurements of resident biota in surface waters as indicator surrogates of environmental conditions, as compared with a relatively undisturbed system, or reference system (Cohen et al. 2005; Brooks and Gebo 2013). The first bioassessments, known as indices of biological integrity (IBIs), were developed by Karr (1981) for monitoring stream health in the Midwestern US (for a short review of IBIs and bioassessment development see Euliss and Mushet 2011; Wilson and Bayley 2012). IBIs typically combine indices from multiple indicator surrogates (either from multiple taxa or multiple response variables from the same taxa) into a single metric of habitat condition (Whittier et al. 2007). Fish, for example, were used as indicator species and assumed to be indicative of the condition of other biotic communities. Since then, the methods have been extended to other aquatic resources globally and bioassessments have been regionalised (e.g. the Ohio wetland IBI, Mack 2007; Florida Wetland Condition Index, Deimeke et al. 2013) and tailored to specific taxa (e.g. macroinvertebrates (Kerans and Karr 1994; Barbour et al. 1999), microbes (Sim et al. 2013), benthic diatoms (Lane et al. 2009) and microcrustaceans (Boix et al. 2005)). Scientists use indicator or management surrogates in IBIs to evaluate degraded aquatic resources, to define acceptable goals for ecosystem restoration, or to identify key aquatic resources for conservation (Wilson and Bayley 2012). The efficacy or reliability of using indicator or management surrogates for evaluating ecosystem integrity, alone or in combination, has been questioned. This is because bioassessments using the same indicator species or assemblages are increasingly applied to different aquatic resources influenced by varying spatial and temporal scales and dynamic climatic conditions (Wilcox et al. 2002). In this paper, we use examples largely from temporary wetlands to explore five foundational things we know about the use of indicator and/or management surrogates in the development of IBIs. For each of the five sections on what we know, we incorporate a discussion of how to use the shortcomings in the current application to guide a way forward to ensure more effective conservation decisions for freshwater wetlands.

Things we know 1.  IBIs are confounded by spatio-temporal context Bioassessments are designed to assess biological responses to changes in ecosystem condition. A key assumption in the application of IBIs is that biological responses are deter-

11: Searching for the holy grail of wetland integrity

Fig. 11.1.  The proportion of total variation in an IBI (the rectangular box) explained by spatial, temporal, abiotic and biotic processes (circles), and their interactions (areas of overlap). Areas outside circles indicate the role of unmeasured variables or stochastic processes. The use of surrogates often assumes that abiotic processes are the primary driver of responses (box a). However, in many cases, spatial processes, such as dispersal and habitat location (in the case of box b), biotic interactions such as predation, or temporal factors such as season or decade, have an overriding influence on the dynamics of an indicator. Similarly, random or stochastic processes (box c), such as genetic drift in a threatened population, may also override the influence of other processes, and thereby disconnect indicators from measurable variables.

mined by environmental conditions. Thus, indicator surrogates (either species or species assemblages) that are absent or in decline indicate a shortcoming of the local habitat. Communities are governed by complex interactions among local abiotic conditions and species, and by spatial and temporal context (Fig. 11.1). Unexplained variation arising from spatiotemporal factors can cloud the ability of IBIs to detect and quantify environmental change. For example, dispersal processes linked to spatial isolation can influence populations of wetland biota (Richter-Boix et al. 2007). Similarly, seasonal variation in wetland communities is common and can overwhelm the detection of environmental differences (Miller et al. 2008; Culler et al. 2014). Evolutionary and biogeographic processes that govern the traits of species at the broadest scales can modify processes at finer spatial and temporal scales (Fig. 11.2). For example,

103

104

Indicators and Surrogates of Biodiversity and Environmental Change

Fig. 11.2.  Hierarchical controls of local communities highlighting the filters (grey boxes) that influence the assembly of taxa (arrows) from a regional species pool into local communities (after Poff 1997). Species inhabiting a local community must be in the regional species pool at a given point in time, and must then be able to disperse to a habitat, and tolerate local abiotic conditions under the pressures of biotic interactions, such as predation and competition. Importantly, the traits of species within a regional pool modify the extent to which filters constrain local persistence. For example, dispersal can maintain populations of mobile taxa under strong predation pressure, thereby reducing the impact of biotic interactions on a community. Despite this tangle of complex filters and feedbacks, IBIs often consider that the biota observed within a local habitat is a product of the abiotic conditions of that environment, and thus focuses on the link outlined in dashed lines.

Batzer (2013) argued that invertebrates in seasonal wetlands are generalists adapted to environmental variability and are therefore relatively unresponsive to local habitat conditions. Similarly, adaptation of fauna to naturally acidic waterways may confer resilience to anthropogenic acidification (Greig et al. 2010). These examples highlight potential limitations of applying indicator or management surrogates across regions, habitat types and taxonomic groups.

2.  Bioassessments measure structure but infer function IBIs have predominately focused on abundance of specific taxa, or metrics of community diversity and composition that describe the structure of ecological communities. Changes in these structural indicators are then used to infer potential changes in ecosystem function; that is, structure becomes an indicator surrogate for function. The functional inference of IBIs can be improved with attention to the processes that link structure to function, or by directly incorporating functional metrics into IBIs. 3.  Development of IBIs requires capacity and motivation IBI development requires logistical and financial capacity, legislated environmental targets, technical expertise and motivation of stakeholders. Therefore, it is not surprising that

11: Searching for the holy grail of wetland integrity

IBIs are most commonly used in developed countries with strong environmental regulations (i.e. motivation) to assess the integrity of highly valued ecosystems. For example, the first IBIs developed in the USA were in response to the Clean Water Act 1972 (Karr 1981) and focused on rivers that provided drinking water and fisheries resources. Similar flurries of IBI development occurred following the European Water Framework Directive (2000/60/CE) (Boix et al. 2005 and references therein). Although freshwater IBIs have been implemented in developing countries, their application can be limited by political will and financial and personnel resources (Wood 2003). Furthermore, a lack of scientific data and taxonomic information limits the resolution of metrics and the identification of reference sites and environmental baselines (Jones 2008). Although indigenous knowledge can assist in providing information on environmental baselines, the loss of experiential knowledge is reducing this capacity (Wood 2003).

4.  Developed IBIs are broadly applied to a range of goals The concept of IBIs and rapid assessments has been broadly embraced by regulators and scientists as a relatively simple way to implement mandates to remediate increasingly impaired aquatic ecosystems and to meet mitigation requirements (Euliss and Mushet 2011; Wilson and Bayley 2012; Brooks and Gebo 2013). Because developing IBIs is expensive and time consuming, it is common for existing IBIs to be blindly and uncritically adopted in new geographic regions or across wetland types. This blanket application can lead to an erroneous assessment of ecosystem condition and inferred function, and, hence, mismanagement. To remedy these shortcomings while building on the existing knowledge base, the scientific and regulatory communities must tackle the three key tasks outlined in point 9. 5.  IBIs have often been developed without user-community consultation The ecological integrity of freshwater ecosystems is increasingly impaired by multiple stressors that are often the product of complex interactions between socioeconomic and biophysical conditions (Hart and Calhoun 2010). Scientists developed IBIs for use by environmental regulators, but consideration of the other social factors, beyond the need for something ‘quick and dirty’ (i.e. a ‘rapid’ assessment), was lacking. The current application of IBIs is hampered by a lack of focus on the strong coupling between natural and human systems.

Knowledge gaps 6.  Acknowledge the spatio-temporal limitations of IBIs and develop protocols that incorporate these processes In practical terms, the community within one temporary woodland pool, for example, cannot be considered in isolation from the other pools around it, nor from the season in which it was sampled. Sampling programs for IBIs need to account for these factors by stratifying sampling in space and time, accommodating the phenology of indicator organisms, or by incorporating spatio-temporal data as covariates in IBI analyses (Miller et al. 2008). One approach to overcoming broad spatio-temporal gradients is the predictive modelling of reference conditions in IBIs. Predictive modelling can explain variation in reference communities by accounting for a suite of spatial and environment variables. Such procedures have overcome spatial differences in river fauna (e.g. constraints to fish dispersal in river networks; Joy and Death 2002) and have shown promise in wetland assessment (Davis et al. 2006). However, these models must still be tailored to specific regions because they are influenced by temporal changes in the accuracy of reference predictions (Davis et al. 2006).

105

106

Indicators and Surrogates of Biodiversity and Environmental Change

7.  Evaluate the use of functional surrogates (diversity, food webs or direct measures) in IBIs Incorporate functional traits into assessment

Employing trait-based measures of ecological diversity is a promising approach to bridge structural and functional measures of biological integrity, especially when traits are closely related to ecosystem functions (de Bello et al. 2010). For example, some freshwater IBIs use information on the diversity and relative abundance of functional feeding groups of macroinvertebrates (Barbour et al. 1999). Similar trait-based approaches have been used to assess the success of wetland restoration (D’Astous et al. 2013). These biological metrics are likely to provide a more appropriate measure of changes in function than taxonomic indicators because of their mechanistic linkages to ecosystem processes (de Bello et al. 2010). Functional metrics also reduce the need for detailed taxonomic information. This may reduce costs and enable the development of IBIs in regions where taxonomy is poorly developed or resources are scarce. When specific information is lacking on links between species traits and ecosystem functions, using structural traits as indicator surrogates may be appropriate; for example, body size is strongly linked to trophic position, life history and metabolic traits (Woodward et al. 2005). Develop food-web approaches to merge the structural and functional aspects of communities

Food webs integrate structural and functional aspects of communities by describing the trophic interactions among species. Food webs also integrate the responses of organisms performing different functional roles, thereby providing a holistic view of ecological responses to environmental change (Thompson et al. 2012). Broadening IBIs to include multiple trophic levels, or organisms from different habitat zones (e.g. the open water and substrate of a wetland basin), enables the detection of a wider range of impacts while retaining the ability to use individual responses to assess the mechanisms of change. Rapid bioassessment of rivers, for example, combines metrics from periphyton, macroinvertebrates and fish (Barbour et al. 1999). Similarly, Boix and colleagues (2005) found that metrics that included both zooplankton and benthic insects provided the most robust assessment of water quality in a variety of coastal wetlands. Advances in the use of stable isotopes to describe food-web structure have enabled more complex analyses, but their use in IBIs is only emerging and outcomes are equivocal. For example, Hogsden and Harding (2014) compared the use of food webs to traditional structural and function metrics in assessing coal-mining impacts on streams. Although traditional metrics were more sensitive to mining stress, food-web metrics revealed the mechanisms behind structural and functional changes. Clearly this is an area for further research. Incorporate direct measures of functional processes in IBIs

The most direct way to incorporate functional attributes into IBIs is to measure ecosystem functions. For example, rates of organic matter breakdown measured by the addition of litter bags is a simple, cost-effective metric that integrates the structure and functions of fungal, bacterial and macroinvertebrate communities (Young et al. 2008). Similarly, measuring respiration as an indicator of whole-ecosystem metabolism quantifies an overarching ecosystem property that is indicative of a variety of natural and anthropogenic gradients in freshwaters (Young et al. 2008; Yates et al. 2014). Although these functional measures have been applied mostly in flowing water, they are likely to be transferrable to wetlands. For example, forested vernal pool ecosystems in Maine are subsidised by carbon

11: Searching for the holy grail of wetland integrity

from terrestrial detritus (Capps et al. 2014). Therefore measuring leaf breakdown in their bioassessment would incorporate a critical functional process. Despite their advantages, functional metrics are not silver bullets for IBIs, and in many cases are likely to add complementary information to structural metrics serving as indicator surrogates, rather than replace them. For example, in a comparison of metrics for river assessment, functional approaches were preferable in detecting broad land-use changes, but structural metrics showed greater sensitivity to specific impacts or pollutants (Yates et al. 2014). Similarly, functional measures in wetlands perform best when integrated with structural metrics, such as taxonomic distinctness, and when applied across taxonomic groups (Gallardo et al. 2011).

8.  Broadly assess the value of ecosystem services and risk of decline. Use traditional knowledge and global resources in areas of limited capacity Foster international collaboration to enhance capacity

Success of recent programs in Thailand (Boonsoong et al. 2009) and Ethiopia (Mereta et al. 2013) highlights the benefits of international collaboration with scientists from areas with established indicator or management surrogates used in IBIs. Similarly, international agreements for the protection of freshwaters, such as the 2002 Ramsar Convention, have provided broad motivation for developing countries to develop such assessments. Value a diversity of freshwater resources

There is a considerable disparity in the development of IBIs among different freshwater types. For example, despite similar legislative protection, the development of IBIs in wetlands has lagged behind riverine and lacustrine assessments. This disparity may be caused by the perception that wetland resources are less valuable than rivers and lakes. In the USA, the coordinated development of wetland IBIs did not begin until 1990 (Adamus and Brandt 1990) and similar delays in development of national wetland IBIs have occurred elsewhere (e.g. Australia, Davis et al. 2006). However, the critical role of wetlands in regional carbon and nutrient budgets (Capps et al. 2014) and contributions to landscape biodiversity suggests the value of ecosystems and their services needs to be more broadly defined and communicated to both the public and legislators (see point 5 for more detail).

9.  Develop clear goals and tailor assessments to meet those goals. Key factors to consider: spatial, temporal and cultural context, wetland type, dynamic nature of wetlands and climate Tailor indicator and management surrogates used for IBIs to be goal and context specific

Choosing indicator or management surrogates for assessing biotic integrity will be very goal dependent, and practitioners should be clear on desired outcomes. For example, in the glaciated north-eastern USA, fairy shrimp (Order Anostraca) are a reliable indicator surrogate for assessing short temporary pond hydroperiods, but they are an inadequate indicator of overall pond ecological condition. Temporary ponds (or vernal pools) in this region support diverse species that are adapted to a range of hydroperiods, from extremely ephemeral to semi-permanent water bodies that can remain inundated through much of the summer. Thus, a landscape that will capture the wide array of vernal pool functions is one in which variable hydroperiods are represented (Semlitsch 2002). Focusing on one indicator surrogate for one hydroperiod will not achieve this goal.

107

108

Indicators and Surrogates of Biodiversity and Environmental Change

Critically evaluate past foundational work and recalibrate when necessary

In complex systems, it may be hard to distinguish changes based on natural variability from those reflecting anthropogenic stressors (Euliss and Mushet 2011; Bried et al. 2013). DeKeyser and colleagues (2003) developed an Index of Plant Community Integrity (IPCI) for quantitatively assessing the quality of seasonal wetland plant communities along a disturbance gradient in the Prairie Pothole Region (PPR) of the USA. Later, Euliss and Mushet (2011) re-evaluated use of this IPCI using a long-term dataset and demonstrated that the IPCI varies so greatly with climate dynamics and anthropogenic stressors in this region that natural climate variability may mask anthropogenic disturbance signals. Misinterpretation of this IBI could lead to mismanagement of temporary wetlands in the PPR. Ultimately, the IPCI is applicable depending on goals, but should be fine-tuned. IBIs can be successfully applied to multiple goals by recalibrating metrics to different types of human impacts. For example, the widely applied Macroinvertebrate Community Index in New Zealand assesses the impacts of agriculture on stream ecosystems by weighting taxa presence by their relative tolerance of organic pollution (Stark 1993). The success of this simple, numerical indicator motivated the development of similar indices that recalibrated tolerance scores of taxa to gradients of urbanisation (Suren et al. 1998) and mining (Gray and Harding 2012). These examples highlight that the continual assessment, refinement and recalibration of IBIs is an efficient way to both maintain accuracy and broaden the reach of bioassessments. Continue to develop new surrogates and approaches for assessing ecosystem condition

Researchers and practitioners should continue to re-evaluate existing IBIs, but also explore use of other surrogates and technologies. For example, Lane and colleagues (2009) explored the use of benthic diatom composition in forested wetlands in the Florida, USA, and have found them to be robust indicators whether sampled in wet or dry phases. Similarly, the use of eDNA to determine species presence in aquatic systems is a promising tool for future inventories (Pilliod et al. 2013), especially for detecting rare taxa (Jerde et al. 2011) or recent invaders (Takahara et al. 2013).

10.  Solicit stakeholder input for conservation of natural resources and in finding solutions to complex coupled natural human systems. Provide outreach and educational opportunities. Stakeholder-driven research and co-production of knowledge by scientists and stakeholders can lead to more enduring, adaptive solutions to complex environmental challenges (Hart and Calhoun 2010), especially in development of IBIs (and hence selection of indicator and management surrogates) where local knowledge may be key. For example, a collaborative approach to vernal pool conservation in Maine, USA, was developed through stakeholder engagement. A vernal pool IBI was developed with regulators from all levels of government, town officials, university ecologists, and the development community to engage in local conservation planning using place-specific conservation tools and pool assessment data collected by trained citizen scientists (Calhoun et al. 2014). The work was guided by an understanding that the biophysical causes and consequences of wetland impairment alone will rarely be sufficient for achieving sustainable management policies and practices (Hart and Calhoun 2010), particularly at local levels of government and when citizen behaviour and attitude is so closely linked to the conservation outcome. Choices of indicator and management surrogates must be tightly linked with stakeholder goals and ultimately lead to closer relationships between local communities and wetland conservation.

11: Searching for the holy grail of wetland integrity

Conclusions We see the use of any surrogates (indicator or management), like any tool, as being valuable if the right ones are chosen to meet the task. Context is everything. If we can avoid taking out splinters with chisels, the use of IBIs may facilitate strong conservation outcomes. Use of assessment tools will need to be tested in various comparable contexts before they are used to dictate mitigation strategies to avoid mismanagement (DeKeyser et al. 2003; Euliss and Mushet 2011). Practitioners need to critically assess the balance of accuracy versus logistical and financial constraints: a process akin to model selection in statistics, where the most parsimonious combination of approaches (i.e. the most reward for the lowest effort) is selected from a range of metrics of differing complexity and intensity. If these filters are applied with specific goals in mind, then managers are more likely to achieve optimal balance of costs and benefits. However, continued re-assessment of these decisions will be necessary as environmental baselines change, technological advances provide new methods, and as shifts in social-economic conditions modify the goals of assessment. Finally, all parties must be open to paradigm shifts. Since the 1970s and the development of the first IBIs, the challenges to conservation professionals have become more complex than considering within-ecosystem water quality issues. The ecological surrogates that have typically served as indicator surrogates (species-specific or assemblages) or management surrogates in wetlands may be shifting rapidly, with varying sensitivities to anthropogenic and natural disturbance factors (Hunter et al. 1988). Other approaches to making conservation decisions related to wetland mitigation or restoration will also need to be considered. For example, Hunter and colleagues (1988) and Hjort and colleagues (2015) argue for the conservation of physical environments (i.e. geodiversity) that encompasses the ‘stage’ upon which communities change and develop. Conservation of geodiversity requires an integrated approach to nature conservation planning and management that considers both biological and geological components at all scales from small sites to whole landscapes. This holistic approach may require less tedious biotic inventories, especially in countries with limited natural resources datasets or financial resources.

Glossary Biological assessments  assessments of habitat or ecosystem state that are based on qualitative or quantitative properties of biota. Ecosystem function  the rates, relative importance and other properties of ecosystem processes. Functional metrics  metrics that are based on a directly quantified ecosystem function; that is, the rates, properties and relative importance of ecosystem processes. IBI (index of biological integrity)  an index that combines multiple biological indicator surrogates into a single metric that describes the similarity of structural and functional aspects of an ecosystem to its natural state.

References Adamus PR, Brandt KH (1990) Impacts on Quality of Inland Wetlands of the United States: A Survey of Indicators, Techniques, and Applications of Community-Level Biomonitoring Data. EPA/600/3–90/.073. United States Environmental Protection Agency Environmental Research Laboratory, Corvalis, OR.

109

110

Indicators and Surrogates of Biodiversity and Environmental Change

Barbour MT, Gerritsen J, Snyder B, Stribling J (1999) Rapid bioassessment protocols for use in streams and wadeable rivers. United States Environmental Protection Agency, Washington DC. Batzer DP (2013) The seemingly intractable ecological responses of invertebrates in North American wetlands: a review. Wetlands 33, 1–15. doi:10.1007/s13157-012-0360-2. Boix D, Gascón S, Sala J, Martinoy M, Gifre J, Quintana XD (2005) A new index of water quality assessment in Mediterranean wetlands based on crustacean and insect assemblages: the case of Catalunya (NE Iberian peninsula). Aquatic Conservation: Marine and Freshwater Ecosystems 15, 635–651. doi:10.1002/aqc.750. Boonsoong B, Sangpradub N, Barbour MT (2009) Development of rapid bioassessment approaches using benthic macroinvertebrates for Thai streams. Environmental Monitoring and Assessment 155, 129–147. doi:10.1007/s10661-008-0423-2. Bried JT, Jog SK, Matthews JW (2013) Floristic quality assessment signals human disturbance over natural variability in a wetland system. Ecological Indicators 34, 260–267. doi:10.1016/j.ecolind.2013.05.012. Brooks RP, Gebo NA (2013) Wetlands restoration and mitigation. In Mid-Atlantic Freshwater Wetlands: Advances in Wetlands Science, Management, Policy, and Practice. (Eds RP Brooks and DH Wardrop) pp. 421–440. Springer, New York. Calhoun AJK, Jansujwicz JS, Bell KP, Hunter ML, Jr (2014) Improving management of small natural features on private lands by negotiating the science-policy boundary. Proceedings of the National Academy of Sciences of the United States of America 111, 11002–11006. doi:10.1073/pnas.1323606111. Capps KA, Rancatti R, Tomczyk N, Parr T, Calhoun AJK, Hunter MD (2014) Biogeochemical hotspots in forested landscapes: the role of vernal pools in denitrification and organic matter processing. Ecosystems 17, 1455–1468. Cohen MJ, Lane CR, Reiss KC, Surdick JA, Bardi E, Brown MT (2005) Vegetation based classification trees for rapid assessment of isolated wetland condition. Ecological Indicators 5, 189–206. doi:10.1016/j.ecolind.2005.01.002. Culler LE, Smith RF, Lamp WO (2014) Weak relationships between environmental factors and invertebrate communities in constructed wetlands. Wetlands 34, 351–361. doi:10.1007/ s13157-013-0502-1. D’Astous A, Poulin M, Aubin I, Rochefort L (2013) Using functional diversity as an indicator of restoration success of a cut-over bog. Ecological Engineering 61, 519–526. doi:10.1016/j. ecoleng.2012.09.002. Davis J, Horwitz P, Norris R, Chessman B, McGuire M, Sommer B (2006) Are river bioassessment methods using macroinvertebrates applicable to wetlands? Hydrobiologia 572, 115–128. doi:10.1007/s10750-005-1033-4. de Bello F, Lavorel S, Díaz S, Harrington R, Cornelissen JH, Bardgett RD, et al. (2010) Towards an assessment of multiple ecosystem processes and services via functional traits. Biodiversity and Conservation 19, 2873–2893. doi:10.1007/s10531-010-9850-9. Deimeke E, Cohen MJ, Reiss KC (2013) Temporal stability of vegetation indicators of wetland condition. Ecological Indicators 34, 69–75. doi:10.1016/j.ecolind.2013.04.022. DeKeyser ES, Kirby DR, Ell MJ (2003) An index of plant community integrity: development of the methodology for assessing prairie wetland plant communities. Ecological Indicators 3, 119–133. doi:10.1016/S1470-160X(03)00015-3. Euliss NH, Mushet DM (2011) A multi-year comparison of IPCI scores for prairie pothole wetlands: implications of temporal and spatial variation. Wetlands 31, 713–723. doi:10.1007/s13157-011-0187-2.

11: Searching for the holy grail of wetland integrity

Gallardo B, Gascón S, Quintana X, Comín FA (2011) How to choose a biodiversity indicator – redundancy and complementarity of biodiversity metrics in a freshwater ecosystem. Ecological Indicators 11, 1177–1184. doi:10.1016/j.ecolind.2010.12.019. Gray D, Harding J (2012) Acid Mine Drainage Index (AMDI): a benthic invertebrate biotic index for assessing coal mining impacts in New Zealand streams. New Zealand Journal of Marine and Freshwater Research 46, 335–352. doi:10.1080/00288330.2012.663764. Greig HS, Niyogi DK, Hogsden KL, Jellyman PG, Harding JS (2010) Heavy metals: confounding factors in the response of New Zealand freshwater fish assemblages to natural and anthropogenic acidity. The Science of the Total Environment 408, 3240–3250. doi:10.1016/j.scitotenv.2010.04.006. Hart DD, Calhoun AJ (2010) Rethinking the role of ecological research in the sustainable management of freshwater ecosystems. Freshwater Biology 55, 258–269. doi:10.1111/j.1365-2427.2009.02370.x. Hjort J, Gordon JE, Gray M, Hunter ML, Jr (2015) Valuing the stage: why geodiversity matters. Conservation Biology 29, 630–639. doi:10.1111/cobi.12510. Hogsden KL, Harding JS (2014) Isotopic metrics as a tool for assessing the effects of mine pollution on stream food webs. Ecological Indicators 36, 339–347. doi:10.1016/j. ecolind.2013.08.003. Hunter ML Jr, Jacobson GL Jr, Webb T, III (1988) Paleoecology and the coarse filter approach to maintaining biological diversity. Conservation Biology 2, 375–385. doi:10.1111/j.15231739.1988.tb00202.x. Jerde CL, Mahon AR, Chadderton WL, Lodge DM (2011) “Sight-unseen” detection of rare aquatic species using environmental DNA. Conservation Letters 4, 150–157. doi:10.1111/j.1755-263X.2010.00158.x. Jones FC (2008) Taxonomic sufficiency: the influence of taxonomic resolution on freshwater bioassessments using benthic macroinvertebrates. Environmental Reviews 16, 45–69. doi:10.1139/A07-010. Joy MK, Death RG (2002) Predictive modelling of freshwater fish as a biomonitoring tool in New Zealand. Freshwater Biology 47, 2261–2275. doi:10.1046/j.1365-2427.2002.00954.x. Karr JR (1981) Assessment of biotic integrity using fish communities. Fisheries (Bethesda, Md.) 6, 21–27. doi:10.1577/1548-8446(1981)0062.0.CO;2. Kerans BL, Karr JR (1994) A benthic index of biotic integrity (B-IBI) for rivers of the Tennessee Valley. Ecological Applications 4, 768–785. doi:10.2307/1942007. Lane CR, Reiss KC, DeCelles S, Brown MT (2009) Benthic diatom composition in isolated forested wetlands subject to drying: implications for monitoring and assessment. Ecological Indicators 9, 1121–1128. doi:10.1016/j.ecolind.2008.12.010. Mack JJ (2007) Developing a wetland IBI with statewide application after multiple testing iterations. Ecological Indicators 7, 864–881. doi:10.1016/j.ecolind.2006.11.002. Mereta ST, Boets P, De Meester L, Goethals PL (2013) Development of a multimetric index based on benthic macroinvertebrates for the assessment of natural wetlands in Southwest Ethiopia. Ecological Indicators 29, 510–521. doi:10.1016/j.ecolind.2013.01.026. Miller AT, Hanson MA, Church JO, Palik B, Bowe SE, Butler MG (2008) Invertebrate community variation in seasonal forest wetlands: implications for sampling and analyses. Wetlands 28, 874–881. doi:10.1672/07-58.1. Pilliod DS, Goldberg CS, Arkle RS, Waits LP, Richardson J (2013) Estimating occupancy and abundance of stream amphibians using environmental DNA from filtered water samples. Canadian Journal of Fisheries and Aquatic Sciences 70, 1123–1130. doi:10.1139/ cjfas-2013-0047.

111

112

Indicators and Surrogates of Biodiversity and Environmental Change

Poff NL (1997) Landscape filters and species traits: towards mechanistic understanding and prediction in stream ecology. Journal of the North American Benthological Society 16, 391–409. doi:10.2307/1468026. Richter-Boix A, Llorente GA, Montori A (2007) Structure and dynamics of an amphibian metacommunity in two regions. Journal of Animal Ecology 76, 607–618. doi:10.1111/j.1365-2656.2007.01232.x. Semlitsch RD (2002) Critical elements for biologically based recovery plans of aquaticbreeding amphibians. Conservation Biology 16, 619–629. doi:10.1046/j.1523-1739.2002. 00512.x. Sim L, Davis J, Strehlow K, McGuire M, Trayler K, Wild S, et al. (2013) The influence of changing hydroregime on the invertebrate communities of temporary seasonal wetlands. Freshwater Science 32, 327–342. doi:10.1899/12-024.1. Stark JD (1993) Performance of the Macroinvertebrate Community Index: effects of sampling method, sample replication, water depth, current velocity, and substratum on index values. New Zealand Journal of Marine and Freshwater Research 27, 463–478. doi:10.1080/ 00288330.1993.9516588. Suren AM, Snelder T, Scarsbrook M (1998) Urban Stream Habitat Assessment Method (USHA). National Institute of Water and Atmospheric Research, Christchurch, New Zealand. Takahara T, Minamoto T, Doi H (2013) Using environmental DNA to estimate the distribution of an invasive fish species in ponds. PLoS ONE 8, e56584. doi:10.1371/journal. pone.0056584. Thompson RM, Brose U, Dunne JA, Hall RO, Hladyz S, Kitching RL, et al. (2012) Food webs: reconciling the structure and function of biodiversity. Trends in Ecology & Evolution 27, 689–697. doi:10.1016/j.tree.2012.08.005. Whittier TR, Hughes RM, Stoddard JL, Lomnicky GA, Peck DV, Herlihy AT (2007) A structured approach for developing indices of biotic integrity: three examples from streams and rivers in the western USA. Transactions of the American Fisheries Society 136, 718–735. doi:10.1577/T06-128.1. Wilcox DA, Meeker JE, Hudson PL, Armitage BJ, Black MG, Uzarski DG (2002) Hydrologic variability and the application of index of biotic integrity metrics to wetlands: a Great Lakes evaluation Wetlands 22, 588–615. doi:10.1672/0277-5212(2002)022[0588:HVATAO]2 .0.CO;2. Wilson MJ, Bayley SE (2012) Use of single versus multiple biotic communities as indicators of biological integrity in northern prairie wetlands. Ecological Indicators 20, 187–195. doi:10.1016/j.ecolind.2012.02.009. Wood C (2003) Environmental impact assessment in developing countries. International Development Planning Review 25, 301–321. doi:10.3828/idpr.25.3.5. Woodward G, Ebenman B, Ernmerson M, Montoya JM, Olesen JM, Valido A, et al. (2005) Body size in ecological networks. Trends in Ecology & Evolution 20, 402–409. doi:10.1016/j. tree.2005.04.005. Yates AG, Brua RB, Culp JM, Chambers PA, Wassenaar LI (2014) Sensitivity of structural and functional indicators depends on type and resolution of anthropogenic activities. Ecological Indicators 45, 274–284. doi:10.1016/j.ecolind.2014.02.014. Young RG, Matthaei CD, Townsend CR (2008) Organic matter breakdown and ecosystem metabolism: functional indicators for assessing river ecosystem health. Journal of the North American Benthological Society 27, 605–625. doi:10.1899/07-121.1.

12

Surrogates for coral reef ecosystem health and evaluating management success Maria Beger

What we know 1 Coral reef management implements indicators of success that are broad measures requiring a relatively low level of specialisation to measure. 2 Long-term measurements are extremely valuable and unlikely to be discontinued, so the choice of the appropriate surrogates at the outset is crucial. 3 Coral reef monitoring encapsulates consideration of unexpected processes or serendipitous findings. 4 Surrogates and indicators are rarely picked with explicit objectives articulated or indicators are not linked to specific actions, threats and expected outcomes in reef ecosystem quality. 5 Process-based surrogates, including functional groups, growth rates and fecundity of species, may be most effective. Knowledge gaps 6 Which types of surrogates and indicators can best minimise uncertainty about management success? 7 How can decision-theoretic approaches improve surrogate choices by linking objectives, actions, costs, indicators and thresholds for adapting coral reef management? 8 How much decline is OK? What are thresholds in magnitude and temporal trajectory that relate to specific management objectives? 9 How effective are of novel surrogates? 10 Single species indicators versus threat-specific community surrogates – which are better when?

Introduction Marine protected areas (MPAs) are the primary management instruments employed in globally declining coral reef ecosystems (Selig and Bruno 2010; De’ath et al. 2012). Monitoring programs are key to test the effectiveness of such management, but also to 113

114

Indicators and Surrogates of Biodiversity and Environmental Change

improve our understanding of the responses of coral reefs to stressors (White et al. 2011; Castillo et al. 2012; Olds et al. 2014). The idea of using surrogates to evaluate MPAs is well established. Surrogates at the community and species levels are cost-effective ways to deal with some of these monitoring objectives when resources and expertise are a constraint (Valavi et al. 2009; McDonald-Madden et al. 2010; Rouphael et al. 2011; Olds et al. 2014). Despite some correlative tests of indicators, their reliability is rarely quantified. In reality, monitoring programs on coral reefs typically employ broad community-level surrogates such as coral cover and the abundance of fished species. Management success on coral reefs relates to ecosystem health (Selig and Bruno 2010), (functional) diversity (Russ and Alcala 2011), fish abundance (Abesamis and Russ 2005), fisheries sustainability (Abesamis and Russ 2005; Kronen et al. 2012), invasive (native/ introduced) species (De’ath et al. 2012), social acceptance (Fox et al. 2014) and indirect effects (Babcock et al. 2010). Indirect effects of management actions could be viewed as perverse outcomes, depending on how management success is measured. For example, where a reef is protected in a MPA to manage fished lobster populations – these could be monitored and success could be ascertained. However, if abalone was selected as a surrogate for success, populations may decline because lobsters eat abalone. This is a perverse outcome for fisheries, because abalone is also valuable to them. Many long-term monitoring programs aim to assess the ecological integrity of reefs over the time of management. Their designs typically capture inside-outside-far away contrasts to evaluate relative differences in reef health or fish stock trends (Abesamis and Russ 2005; Babcock et al. 2010). On the other hand, measuring the success of management can be extremely expensive for remote reefs, and management organisations sometimes opt for monitoring inside the managed areas only to report trends over time (Ceccarelli et al. 2011). The application of surrogates in monitoring and management of coral reefs, as well as the potential use of indicator species to evaluate ecosystem health (Olds et al. 2014), is widespread, and based on well-supported foundational ideas. New paradigms in reef ecology and decision science and novel technologies, however, challenge the existing practices beyond this comfort zone. This chapter aims to highlight five known things, and five key challenges for the future of surrogate use for coral reefs and MPA establishment, management and monitoring.

Things we know 1.  Coral reef management implements indicators of success that are broad measures requiring a relatively low level of specialisation to measure The high species richness of coral reefs hinders the use of diversity indices as indicators, and even single species indicators beyond commonly known species (i.e. big fishes). Monitoring programs therefore most often focus on community-level surrogates such as coral cover, benthic cover of grouped categories such as algae, sponges and corals, often with a functional typology (e.g. fleshy macro-algae and encrusting red algae that play competitive versus structural roles in reef ecosystems) (Table 12.1). Coral cover is by far the most widely used indicator of MPA effectiveness (Selig and Bruno 2010; De’ath et al. 2012; Richards 2013), even though it is an indirect measure of the fishing closure, which does not directly impacts corals. Recent research points towards the use of fish families (Heenan and Williams 2013) and coral genera (Richards 2013) as surrogates to represent species diversity or assemblage type.

12: Surrogates for coral reef ecosystem health and evaluating management success

Table 12.1.  Examples of surrogates and indicators in coral reef monitoring Objective

Indicator

Value

Problems

Reference

Community-level surrogates MPA effectiveness

Coral cover

Commonly measured, effective in showing general trends, easy to measure

Weak representation of coral diversity or function, ignores dependent taxa

(Selig and Bruno 2010; Richards 2013; Olds et al. 2014)

MPA effectiveness

Total fish abundance or biomass

Effective for temporal comparisons

Recorded assemblage may differ in tropics

(Babcock et al. 2010; Weijerman et al. 2013; Guillemot et al. 2014)

MPA effectiveness

Functional group biomass (B)

Effective for temporal comparisons, herbivore B predicts coral cover

Potential MPA effectiveness

Socio-economic pressure

Productivity of reef relative to artisanal fisher numbers

Potentially biased responses

(Kritzer 2004; Kronen et al. 2012)

Total species richness

Coral genus richness

Effective surrogate for species richness

Specialised observer required

(Richards 2013)

Climate effects of MPA

Total fish biomass, functional group biomass

Effective in measuring climate trends over time, temperate system

Specialised observer and analytical skills required

(Bates et al. 2014)

Patchy distributions, often locally extinct

(Olds et al. 2014)

(Abesamis and Russ 2005; Russ and Alcala 2011; Heenan and Williams 2013; Weijerman et al. 2013; Bates et al. 2014; Guillemot et al. 2014)

Single species indicators MPA effectiveness

Focal species, (e.g. Bumphead Parrotfish, Bolbometopon muricatum)

Effects of MPA proximity to supporting ecosystems (mangroves) on fish communities

Level of impact

Sea Urchin (Echinometra mathaei)

Positive correlation with commercial fishing = high indicator value for commercial fishing areas

(Valavi et al. 2009)

Level of impact

Arabian Butterflyfish (Chaetodon melapterus)

Significant indicator for none to low fishing

(Valavi et al. 2009)

115

116

Indicators and Surrogates of Biodiversity and Environmental Change

Often, MPA success is measured directly using the species directly affected by the cessation of fishing, such as the fishes and invertebrates caught by the surrounding fisheries. While monitoring the recovery of fish stocks unnecessarily proves that fishing kills fish (Gerber et al. 2005), it also allows an assessment of times to recovery and potential poaching (Kritzer 2004). It quantifies the relative change compared with reefs that continue to be fished. These changes usually rely on abundance or biomass measures of fished taxa, or in rare circumstances indicator species. For example, the Bumphead Parrotfish (Bolbometopon muricatum) was found to be an indicator for functional intactness (bio-erosion function), fishing pressure and overall reef condition in the Solomon Islands (Olds et al. 2014).

2.  Long-term measurements are extremely valuable and unlikely to be discontinued, so the choice of the appropriate surrogates at the outset is crucial Adaptation of surrogates and indicators in monitoring programs may improve their effectiveness as our knowledge of the processes measured consolidates (McDonald-Madden et al. 2010). However, this is rarely feasible for adapting long-term monitoring programs on coral reefs. Long-term monitoring projects that evaluate coral reef management success are an exception, with few programs currently running for more than 10 years. It is unlikely to ever be advantageous to discontinue these existing measurements in favour of new indicators. This emphasises the importance of choosing surrogates at the start of coral reef monitoring programs. Too few parameters measured could mean that, after several years of monitoring, it is found that they are wrong, capture only part of the process, hide perverse outcomes or are ineffective in representing ecosystem state. Too many parameters may render a program too expensive and/or collecting some data that may not be useful. In this case, it will also be more difficult to add new parameters in the future, because the opportunity to collect additional data is constrained (e.g. boat space, available expertise). Neither choosing too few nor too many surrogates is likely to be beneficial. Monitoring should relate to the surrogates used to prioritise management, as well as picking parameters that are informative across the specific objectives of management. 3.  Coral reef monitoring encapsulates consideration of unexpected processes or serendipitous findings Surrogates such as such as coral cover, algal cover and the abundance of fished species allow for a broad scope of coral reef management evaluation. This allows for unexpected results, but often limits the exploration of novel questions or the impacts of new threats because broad community indicators are irrelevant to these questions. Conversely, recording species-level abundance and biomass data allows the exploration of emerging research questions, such as recent explorations of functional diversity (Stuart-Smith et al. 2013; Sommer et al. 2014). Long-term full community surveys, while expensive and logistically challenging, have the capacity to discover unanticipated relationships, such as the discovery that MPAs halt the effects of climate change on temperate reef tropicalisation (Bates et al. 2014), when the monitoring program was designed at a time when climate change not yet known. 4.  Surrogates and indicators are rarely picked with explicit objectives articulated or indicators are not linked to specific actions threats, and expected outcomes in reef ecosystem quality Coral reef monitoring programs generally fulfil multiple objectives, but often the objectives are not clearly defined a priori in an outcome-specific actionable context. For example, while these reef monitoring programs are about measuring change, they do not

0.0

0.4

0.6

0.8

1.0

0.0

0.2

0.4

0.6

0.8

1.0

0.0

0.2

800

450

700

400

Total Biomass [kg]

Total Abundance

b)

0.2

Coral vs Echinoderms

600 500 400 300 200 100 0

0.4

0.6

0.8

0.0 0.2 0.4 0.6 0.8 1.0

0.2

0.4 0.2 0.0

0.0

0.0 0.2 0.4 0.6 0.8 1.0

1.0

Coral vs Fishes

0.4

0.6

0.6 0.8

Coral vs Molluscs

0.8

a)

1.0

12: Surrogates for coral reef ecosystem health and evaluating management success

Coral vs Crustacea

1.0

0.0

C

D

0.2

0.4

0.6

0.8

1.0

350 300 250 200 150 100 50

A

B

C

D

E

F

G

0

A

B

E

F

G

Fig. 12.1.  The lack of congruence in potential coral reef monitoring surrogates necessitates careful objective-driven indicator choices: (a) relationships of site species richness at Ashmore Reef, Australia, for five taxa (coral on x-axis); and (b) diagrams showing divergent patterns for fish abundance and biomass measures for the same seven sites.

envisage how much change is permissible nor what would be the best parameters to measure it. Most of these programs aim to assess whether there is a detectable difference between management zones and adjacent fisheries (Abesamis and Russ 2005; Babcock et al. 2010; Russ and Alcala 2011). However, it is unclear which outcome is desired, at which point management will be changed, what are possible alternative management actions and how much sampling is required to minimise uncertainty. Many monitoring programs assess the success of MPAs based on coral and other benthic cover (Ceccarelli et al. 2011) or coral diversity (Richards 2013). Yet, there is a disconnect between the fishing threats potentially acting on fishes, and their often negligible direct effect on coral cover. Given the incongruent distributions of corals and fishes, and other associated fauna (Beger et al. 2007) (Fig. 12.1a), monitoring of fishes and invertebrates is also required. This is common practice in coral reef monitoring. The abundance, and occasionally biomass, of fishes is regularly monitored, often targeting specific fisheries target species, but patterns may differ substantially between metrics (Abesamis and Russ 2005; Guillemot et al. 2014) (Fig. 12.1b). Many coral reef monitoring programs measure ecological surrogates, but fail to record water-quality parameters, and socio-economic surrogates such as fish markets, incomes or support for continued management in the population. Adapting coral reef monitoring programs towards clearer objectives, and linking them to relevant and powerful surrogates could increase efficiency in many cases (Fox et al. 2014).

5.  Process-based surrogates, including functional groups, growth rates and fecundity of species, may be most effective The distributions of the main coral reef taxa are not congruent (Beger et al. 2003) (Fig. 12.1a). When designing marine protected areas, therefore, surrogate taxa are limited in

117

118

Indicators and Surrogates of Biodiversity and Environmental Change

representing each other (Beger et al. 2003, 2007). Species within each taxon are likely to exhibit differential responses to environmental and management changes, and non-congruent patterns among taxa may mean that cross-taxon surrogacy is ineffective, particularly with respect to measuring environmental stress effects. Environmental stress affects most vulnerable species first, and some effects will be sub-lethal. Unless physiological parameters are measured, these first warning signals will be missed by coral reef monitoring programs that measure broad community surrogates. Process-based indicators mostly are values associated with single species or functional groups, and apply to specific taxa (e.g. corals, fish, echinoderms). For all taxa, indicators for single species that generally relate inversely to stress include growth rates (Hughes et al. 2000; Feary et al. 2009; Castillo et al. 2012) and, more tentatively, fecundity (Hughes et al. 2000; Bauman et al. 2011). For example, coral calcification rates are higher in the more sheltered lagoon environments, compared with exposed reef crests (Castillo et al. 2012). Recruitment failure in corals indicates imminent threat to populations, but its detection depends on monitoring recruit abundances and size structure distributions that assess population dynamics of key species (Guzner et al. 2007). Single fish species indicators for ontogenetic connectivity between coral reef and mangrove ecosystems show promise in representing and determining if these functional links are intact in MPA networks (Olds et al. 2014). Functional groups are often excellent indicators for both stressor impacts and management outcomes. For example, the ecological effect of fishing is well captured in functional group summary metrics, targeting relevant groups such as large species or carnivores (Weijerman et al. 2013; Guillemot et al. 2014). Indirect effects of fishing and management on benthic assemblages also can be measured by functional groups in fishes, such as increases in parrotfish biomass (herbivores) that predict the increase of coral cover (Heenan and Williams 2013). Certain combinations of species traits, or functional types, can relate to the stress tolerance or environmental responses of species and communities (Bates et al. 2014; Sommer et al. 2014). Some trait combinations shared by species groups may be more vulnerable to stress, such as competitive and brooding coral species (Sommer et al. 2014). Similarly, certain functional types, such as herbivores, can indicate community level temporal trends of tropicalisation (Bates et al. 2014). Although these issues are increasingly discussed in the scientific literature, they are only beginning to emerge among practitioners.

Knowledge gaps 6.  Which types of surrogates can best minimise uncertainty about management success? Monitoring is inevitably restricted to a small number of system characteristics, and recording can be intermittent. In the absence of effective design, there is a high risk that monitoring will lack the power to detect relevant changes in the target environment in a timely fashion (Legg and Nagy 2006; White et al. 2011), or will measure the wrong parameters. Surrogates typically used to monitor the fate of coral reefs, such as coral cover or fish abundance, are prone to measurement variability that can overshadow the trends to be monitored. In typical long-term monitoring projects, the main – and often limiting – expenses are boat time and expertise. Therefore, it is reasonably similar in effort to record the abundance of one surrogate species versus a larger suite of species. Only if we can find parameters that can be easily recorded by relative non-specialists, yet be more specific to monitoring objectives than broad indicators, uncertainty and time-lags in recording responses can be reduced.

12: Surrogates for coral reef ecosystem health and evaluating management success

7.  How can decision-theoretic approaches improve surrogate choices by linking objectives, actions, costs, indicators and thresholds for adapting coral reef management? Ecological coral reef monitoring practices tend to be poorly connected with decision making (Rouphael et al. 2011; White et al. 2011; Fox et al. 2014). Some consider the lack of consistent monitoring protocols and a framework to convert data into actionable guidelines as the main challenge in adaptive coral reef management (Fox et al. 2014). Fuelled by decades of monitoring studies that have reported population declines with no apparent link to management objectives (Babcock et al. 2010; Hughes et al. 2011), or without any responsive action being taken, some argue that monitoring is a waste of management resources (Legg and Nagy 2006). Furthermore, disagreement about the causes of population or ecosystem decline (Hughes et al. 2011) hampers reactive or adaptive management. Yet, as monitoring time series grow, analyses can link declines even in broad surrogates such as coral cover to causes such as Crown of Thorns Starfish outbreaks (De’ath et al. 2012), which have clear management responses. However, most studies concerned with possible declines in species or ecosystems lack a clear plan of action if a decline is detected (Lindenmayer et al. 2013). Often different institutions are charged with monitoring and management, so that any gains of optimising the allocation of resources for monitoring within adaptive management frameworks cannot be realised. An unsolved question is whether surrogates and indicators specifically linked to manageable threats will enable us to more efficiently couple coral reef management and monitoring. 8.  How much decline is OK? What are thresholds in magnitude and temporal trajectory that relate to specific management objectives? Linking monitoring to specific management objectives and actions includes setting decision thresholds in a measured environmental or biological variable that trigger a management response. For this, a program must pre-determine the limits of acceptable change in the most valued elements of the system (Gerber et al. 2005; White et al. 2011). This includes actual trigger levels in monitored parameters and elements such as required data accuracy (White et al. 2011), tolerable uncertainty in data (McDonald-Madden et al. 2010), and the availability of monitoring targets that are both feasible and useful. Most coral reef monitoring is poorly linked to management, particularly with respect to pre-determined action thresholds. Coral reef communities exhibit considerable variability in any single ecological parameter measured locally and regionally (Hughes et al. 2000). Good trigger values will be associated with parameters that are easily measurable, and that respond rapidly and specifically to a manageable impact or threat. For example, hard coral cover provides a measure of the impact of a tropical storm or bleaching, and could trigger immediate action in protecting herbivores that support recovery processes (Heenan and Williams 2013). However, the time-lag associated with recovering populations of long-lived herbivores such as fishes (Wilson et al. 2006) could mean a significant extinction debt arises by the time the monitoring measures a response. Whole community assessments and indices might allow the evaluation of management in the light of measurable benchmarks. For example, an annual reef productivity of 5 Mt/ km2 reef fish could be a threshold for unsustainable fisheries in the Pacific (Kronen et al. 2012), translating to actionable management, such as closures. However, the diversity and multi-facetted interactions of reef organisms complicate monitoring triggers for adaptive management, as these could relate to potentially interacting factors such as diversity, community structure, trophic structure, size structure of corals and fishes, and indirect effects

119

120

Indicators and Surrogates of Biodiversity and Environmental Change

(Babcock et al. 2010; Hughes et al. 2011). For instance, the size structure of coral colonies is an important indicator for the stage of development of coral communities, irrespective of overall coral cover (Guzner et al. 2007).

9.  How effective are novel surrogates? The extent and spatial scale of coral reef monitoring is hampered by operational limitations associated with marine fieldwork. Emerging novel approaches to monitoring ecosystem states promise cost reductions, higher levels of accuracy, and applicability to broader spatial scales. On the largest spatial scales of operation, remote sensing is applied to monitor threatening levels of thermal stress (NOAA Coral Reef Watch, http://coralreefwatch. noaa.gov/), and can record the extent of threats such as large-scale bleaching (Mumby et al. 2001). Remote satellite imagery also serves to discern habitat types and, to some extent, monitor their intactness (Kachelriess et al. 2014). Remote sensing using high resolution satellites or aerial images may provide opportunities to evaluate fishing pressure on predatory fishes over wide spatial scales through the use of grazing halos (Madin et al. 2011). The size of grazed areas around coral heads is larger where fewer predators are present, because herbivores venture farther from shelter – this method relies on reef morphologies where grazing halos can be discerned, and needs to be tested in different regions. Highly structured reef habitats provide enhanced shelter for associated fauna, so rugosity is a good surrogate for reef condition and fisheries productivity (Friedman et al. 2012). Automated image recognition techniques are poised to evaluate rugosity from stereo images (Friedman et al. 2012; Gonzalez-Rivero et al. 2014), but these techniques are still being developed. Novel use and capabilities of genetic tools are also new in quantitative monitoring. Environmental DNA shows particular promise in the field of detecting rare and threatened species, but methods are in their infancy (Veldhoen et al. 2012 and see Chapter 15). 10.  Single species indicators versus threat-specific community surrogates – which are better when? Coral reefs are highly complex systems where monitoring could easily miss a change when limited numbers of parameters are measured (Hughes et al. 2011). Both single species indicators and broad community surrogates limit the representation of species interactions, whether they be trophic interactions or inter-species indirect effects of management. Such indirect effects of management are widespread in marine systems (Babcock et al. 2010), and sometimes may represent perverse outcomes. Community surrogates such as coral cover, fish biomass and functional diversity are relatively easy to collect and non-lethal to organisms (Table 12.1). Yet, they tend to be slow to reliably show trends, which often are outside management cycles and are difficult to confirm and test. Single species indicators, on the other hand, may be more suitable to measure specific threats (Valavi et al. 2009). Single species indicators also may be more powerful to examine specific ecological processes and physiological responses (point 5). However, many techniques to collect information such as fecundity data, growth rates, stress hormones, and physiological tolerances may harm the organisms measured, are costly, and require specialised expertise. Therefore, the choice of indicators and surrogates on coral reefs is complicated by uncertainty in species interactions, and often disagreement regarding accepted definitions and benchmarks of coral reef condition, including which are the key parameters that indicate changes in reef systems.

12: Surrogates for coral reef ecosystem health and evaluating management success

Conclusion Coral reef monitoring parameters generally include community surrogates such as coral cover and the abundance of fishes. In many cases, monitoring seeks to document broadscale changes in biodiversity and community structure, with no specific threat in mind (Gerber et al. 2005; Rouphael et al. 2011). For example, a monitoring program might seek to record the success of restoration activities, to document local socio-economics of reef resource use (Fox et al. 2014), to assess the effectiveness of Marine Protected Areas (MPAs) (Abesamis and Russ 2005) or to document recovery from disturbance (Babcock et al. 2010). Accordingly, most often the parameters measured on behalf of overall reef status pertain to community surrogates, such as benthic cover, fisheries target species/ families abundance, and biomass. These parameters respond slowly to pressures, and more specific process-based indicators such as growth rates and functional group biomass or composition increasingly contribute towards monitoring ecosystem health. Single-species indicators are rarer, but can be used for water quality monitoring as well as indicator species for reef condition (Olds et al. 2014). Coral reef indicator species are often charismatic megafauna (turtles, Dugong Dugong dugon, Bumphead Parrotfish Bolbometopon muricatum) or highly valued fisheries species (coral trout – Plectropomus spp., rabbitfishes – Siganus spp.), but the purpose of these focal species indicators as social tools to elicit stakeholder support is generally not acknowledged. Where is the field of coral reef monitoring going? There are many new tools becoming available (compare points 4 and 9), and new surrogates are being tested (see Chapter 13) (compare points 4, 5 and 9). Like in the monitoring of other ecosystems (McDonald-Madden et al. 2010), most coral reef monitoring processes and indicators are not linked to actions, nor are the data often analysed to produce actionable recommendations (Fox et al. 2014). There still is a discrepancy between scientific knowledge about monitoring design, indicators and surrogates, and the implementation in the field by managers, and much can be gained from cross-disciplinary learning.

References Abesamis RA, Russ GR (2005) Density-dependent spillover from a marine reserve: long-term evidence. Ecological Applications 15, 1798–1812. doi:10.1890/05-0174. Babcock RC, Shears NT, Alcala AC, Barrett NS, Edgar GJ, Lafferty KD, et al. (2010) Decadal trends in marine reserves reveal differential rates of change in direct and indirect effects. Proceedings of the National Academy of Sciences of the United States of America 107, 18256–18261. doi:10.1073/pnas.0908012107. Bates AE, Barrett NS, Stuart-Smith RD, Holbrook NJ, Thompson PA, Edgar GJ (2014) Resilience and signatures of tropicalization in protected reef fish communities. Nature Climate Change 4, 62–67. doi:10.1038/nclimate2062. Bauman AG, Baird AH, Cavalcante GH (2011) Coral reproduction in the world’s warmest reefs: southern Persian Gulf (Dubai, United Arab Emirates). Coral Reefs 30, 405–413. doi:10.1007/s00338-010-0711-5. Beger M, Jones GP, Munday PL (2003) Conservation of coral reef biodiversity: a comparison of reserve selection procedures for corals and fishes. Biological Conservation 111, 53–62. doi:10.1016/S0006-3207(02)00249-5. Beger M, McKenna SA, Possingham HP (2007) Effectiveness of surrogate taxa in the design of coral reef reserve systems in the Indo-Pacific. Conservation Biology 21, 1584–1593.

121

122

Indicators and Surrogates of Biodiversity and Environmental Change

Castillo KD, Ries JB, Weiss JM, Lima FP (2012) Decline of forereef corals in response to recent warming linked to history of thermal exposure. Nature Climate Change 2, 756–760. doi:10.1038/nclimate1577. Ceccarelli DM, Richards ZT, Pratchett MS, Cvitanovic C (2011) Rapid increase in coral cover on an isolated coral reef, the Ashmore Reef National Nature Reserve, north-western Australia. Marine and Freshwater Research 62, 1214–1220. doi:10.1071/MF11013. De’ath G, Fabricius KE, Sweatman H, Puotinen M (2012) The 27-year decline of coral cover on the Great Barrier Reef and its causes. Proceedings of the National Academy of Sciences of the United States of America 109, 17995–17999. doi:10.1073/pnas.1208909109. Feary DA, McCormick MI, Jones GP (2009) Growth of reef fishes in response to live coral cover. Journal of Experimental Marine Biology and Ecology 373, 45–49. doi:10.1016/j. jembe.2009.03.002. Fox HE, Holtzman JL, Haisfield KM, McNally CG, Cid GA, Mascia MB, et al. (2014) How are our MPAs doing? Challenges in assessing global patterns in marine protected area performance. Coastal Management 42, 207–226. doi:10.1080/08920753.2014.904178. Friedman A, Pizarro O, Williams SB, Johnson-Roberson M (2012) Multi-scale measures of rugosity, slope and aspect from benthic stereo image reconstructions. PLoS ONE 7, e50440. doi:10.1371/journal.pone.0050440. Gerber LR, Beger M, McCarthy MA, Possingham HP (2005) A theory for optimal monitoring of marine reserves. Ecology Letters 8, 829–837. doi:10.1111/j.1461-0248.2005.00784.x. Gonzalez-Rivero M, Bongaerts P, Beijbom O, Pizarro O, Friedman A, Rodriguez-Ramirez A, et al. (2014) The Catlin Seaview Survey – kilometre-scale assessment and monitoring of coral reef ecosystems. Aquatic Conservation: Marine and Freshwater Ecosystems 24, 184–198. doi:10.1002/aqc.2505. Guillemot N, Chabanet P, Kulbicki M, Vigliola L, Leopold M, Jollit I, et al. (2014) Effects of fishing on fish assemblages in a coral reef ecosystem: from functional response to potential indicators. Ecological Indicators 43, 227–235. doi:10.1016/j.ecolind.2014.02.015. Guzner B, Novoplansky A, Chadwick NE (2007) Population dynamics of the reef-building coral Acropora hemprichii as an indicator of reef condition. Marine Ecology Progress Series 333, 143–150. doi:10.3354/meps333143. Heenan A, Williams ID (2013) Monitoring herbivorous fishes as indicators of coral reef resilience in American Samoa. PLoS ONE 8, e12345. Hughes TP, Baird AH, Dinsdale EA, Moltschaniwskyj NA, Pratchett MS, Tanner JE, et al. (2000) Supply-side ecology works both ways: the link between benthic adults, fecundity, and larval recruits. Ecology 81, 2241–2249. doi:10.1890/0012-9658(2000)081[2241:SSEWB W]2.0.CO;2. Hughes TP, Bellwood DR, Baird AH, Brodie J, Bruno JF, Pandolfi JM (2011) Shifting baselines, declining coral cover, and the erosion of reef resilience: comment on Sweatman et al. (2011). Coral Reefs 30, 653–660. doi:10.1007/s00338-011-0787-6. Kachelriess D, Wegmann M, Gollock M, Pettorelli N (2014) The application of remote sensing for marine protected area management. Ecological Indicators 36, 169–177. doi:10.1016/j. ecolind.2013.07.003. Kritzer JP (2004) Effects of noncompliance on the success of alternative designs of marine protected-area networks for conservation and fisheries management. Conservation Biology 18, 1021–1031. doi:10.1111/j.1523-1739.2004.00022.x. Kronen M, Pinca S, Magron F, McArdle B, Vunisea A, Vigliola L, et al. (2012) Socio-economic and fishery indicators to identify and monitor artisanal finfishing pressure in Pacific

12: Surrogates for coral reef ecosystem health and evaluating management success

Island countries and territories. Ocean and Coastal Management 55, 63–73. doi:10.1016/j. ocecoaman.2011.07.011. Legg CJ, Nagy L (2006) Why most conservation monitoring is, but need not be, a waste of time. Journal of Environmental Management 78, 194–199. doi:10.1016/j. jenvman.2005.04.016. Lindenmayer DB, Piggott MP, Wintle BA (2013) Counting the books while the library burns: why conservation monitoring programs need a plan for action. Frontiers in Ecology and the Environment 11, 549–555. Madin EMP, Madin JS, Booth DJ (2011) Landscape of fear visible from space. Scientific Reports 1, 14. doi:10.1038/srep00014. McDonald-Madden E, Baxter PWJ, Fuller RA, Martin TG, Game ET, Montambault J, et al. (2010) Monitoring does not always count. Trends in Ecology & Evolution 25, 547–550. doi:10.1016/j.tree.2010.07.002. Mumby PJ, Chisholm JRM, Clark CD, Hedley JD, Jaubert J (2001) A bird’s-eye view of the health of coral reef. Nature 413, 36. doi:10.1038/35092617. Olds AD, Connolly RM, Pitt KA, Maxwell PS, Aswani S, Albert S (2014) Incorporating surrogate species and seascape connectivity to improve marine conservation outcomes. Conservation Biology 28, 982–991. doi:10.1111/cobi.12242. Richards ZT (2013) A comparison of proxy performance in coral biodiversity monitoring. Coral Reefs 32, 287–292. doi:10.1007/s00338-012-0963-3. Rouphael AB, Abdulla A, Said Y (2011) A framework for practical and rigorous impact monitoring by field managers of marine protected areas. Environmental Monitoring and Assessment 180, 557–572. doi:10.1007/s10661-010-1805-9. Russ GR, Alcala AC (2011) Enhanced biodiversity beyond marine reserve boundaries: the cup spillith over. Ecological Applications 21, 241–250. doi:10.1890/09-1197.1. Selig ER, Bruno JF (2010) A global analysis of the effectiveness of marine protected areas in preventing coral loss. PLoS ONE 5(2), e9278. doi:10.1371/journal.pone.0009278. Sommer B, Harrison PL, Beger M, Pandolfi JM (2014) Trait-mediated environmental filtering drives assembly at biogeographic transition zones. Ecology 95, 1000–1009. doi:10.1890/131445.1. Stuart-Smith RD, Bates AE, Lefcheck JS, Duffy JE, Baker SC, Thomson RJ, et al. (2013) Integrating abundance and functional traits reveals new global hotspots of fish diversity. Nature 501, 539–544. Valavi H, Savari A, Yavari V, Kochanian P, Safahieh A, Savadkuhi OS (2009) Coral reef anthropogenic impact bio-indicators in the northern part of the Persian Gulf. Applied Ecology and Environmental Research 7, 215–227. doi:10.15666/aeer/0703_215227. Veldhoen N, Ikonomou MG, Helbing CC (2012) Molecular profiling of marine fauna: integration of omics with environmental assessment of the world’s oceans. Ecotoxicology and Environmental Safety 76, 23–38. doi:10.1016/j.ecoenv.2011.10.005. Weijerman M, Fulton EA, Parrish FA (2013) Comparison of coral reef ecosystems along a fishing pressure gradient. PLoS ONE 8(5), e63797. doi:10.1371/journal.pone.0063797. White JW, Botsford LW, Baskett ML, Barnett LAK, Barr RJ, Hastings A (2011) Linking models with monitoring data for assessing performance of no-take marine reserves. Frontiers in Ecology and the Environment 9, 390–399. doi:10.1890/100138. Wilson SK, Graham NAJ, Pratchett MS, Jones GP, Polunin NVC (2006) Multiple disturbances and the global degradation of coral reefs: are reef fishes at risk or resilient? Global Change Biology 12, 2220–2234. doi:10.1111/j.1365-2486.2006.01252.x.

123

This page intentionally left blank

‡7HQ&RPPLWPHQWVSSLQGG

$0

13

Abiotic surrogates in support of marine biodiversity conservation Camille Mellin

Things we know 1 Two major types of abiotic surrogates exist, based on the methodology used to construct them: pattern-based and selection-based surrogates. 2 Pattern-based surrogates can be constructed based on three distinct approaches: surrogate mapping; assemble first, predict later; and predict first, assemble later. 3 Some abiotic surrogates perform consistently well across marine ecosystems, others are ecosystem- or region-dependent. 4 The effectiveness of abiotic surrogates in support of marine conservation is scale-dependent. 5 The effectiveness of abiotic surrogates is constrained by statistical and sampling considerations. Knowledge gaps 6 The extent to which the effectiveness of abiotic surrogates is stationary over space and time remains unknown. 7 How much natural and human-induced disturbance can influence the robustness of abiotic surrogates is unclear. 8 The cost-effectiveness of different surrogates across spatial scales is yet to be quantified. 9 The impact of limited data availability, sampling selectivity and coarse spatial resolution remain unassessed. 10 Optimised ways of using abiotic surrogates need to be developed for supporting regionalisation approaches that incorporate socio-ecological considerations.

Introduction In line with the Convention of Biological Diversity, there is an increasing need to implement more and better-designed no-take marine protected areas (MPAs) for the protection of 125

126

Indicators and Surrogates of Biodiversity and Environmental Change

marine biodiversity, which is currently undergoing unprecedented decline (Halpern et al. 2012). MPAs currently represent only 3.9% of the marine environment, yet a 10% coverage target by 2020 has been agreed on internationally (Kachelriess et al. 2014). This implies the need to protect not only coastal but also pelagic environments and, importantly, to strengthen the scientific basis behind MPA design (Moilanen 2012). In this regard, systematic conservation planning is a powerful prioritisation approach that uses biological data (e.g. species distributions) as typical targets. In the marine realm, however, exhaustive biological surveys are typically expensive and time consuming, restricted to small spatial scales, and often taxonomically or geographically biased. The subsequent sparseness of biological data is a major impediment to designing systematic conservation planning strategies based on spatially explicit species distributions at large scales. As a result, there is an increasing use of surrogates that act as indicators for marine biodiversity patterns (Mellin et al. 2011). Among all possible surrogates for marine biodiversity, abiotic surrogates (including habitats or environmental characteristics) are becoming increasingly available through the rapid development of high-resolution and cost-effective technologies such as remote sensing imagery or acoustic techniques (Mellin et al. 2009; McArthur et al. 2010; Brown et al. 2011; Mellin et al. 2012). Although such technological development has happened only over the last two decades, studies assessing the effectiveness of abiotic surrogates are now flourishing. This allows for the first generalisations across ecosystems, regions and spatial scales. Here, I summarise some key patterns that emerge from such generalisations, as well as knowledge gaps on which future research efforts should focus.

1.  Two major types of abiotic surrogates exist, based on the methodology used to construct them: pattern-based and selection-based surrogates Abiotic surrogates (sometimes referred to as indicators) are physical entities such as environmental variables or habitats that are used to represent a target entity such as genes, species or ecosystem metrics. ‘Pattern-based surrogacy’ identifies effective surrogates based on their statistical congruence (e.g. goodness of fit) with the target. Conversely, ‘selectionbased surrogacy’ identifies effective surrogates through systematic conservation planning or some prioritisation analysis. Prioritisation analyses typically rely on an algorithm that iteratively selects and adds sites to a set of protected areas to optimise surrogate complementarity (Mellin et al. 2011; Andréfouët et al. 2012). Most marine studies of abiotic surrogates are in fact pattern-based. They rely on the use of correlative techniques such as generalised linear (or additive) models, or machine learning techniques such as boosted regression trees and random forests, to predict a biological response variable (e.g. number of species) as a function of some environmental or spatial predictors (the abiotic surrogates). By contrast, very few marine studies have considered selection-based surrogates, such as habitat maps or substrate data (Dalleau et al. 2010; Van Wynsberge et al. 2012; Hamel et al. 2013; Peckett et al. 2014). The effectiveness of selection-based surrogates is typically assessed based on rarity-complementarity algorithms (Sarkar et al. 2006), the MARXAN site-selection optimisation software (Ball et al. 2009) or spatial conservation prioritisation (Moilanen 2012). Evidence is emerging that pattern-based surrogates that perform consistently well across regions and ecosystems (see point 3), such as sediment covers (McArthur et al. 2010), can sometimes perform poorly as selection-based surrogates, to the point that using no data at all can be equally effective (Peckett et al. 2014). Both approaches might have valid and sensible rationales, yet their distinction in conservation strategies is essential.

13: Abiotic surrogates in support of marine biodiversity conservation

Indeed, although effective pattern-based surrogates are often recommended for conservation planning, only selection-based approaches can identify useful surrogates in a systematic conservation planning approach (Andréfouët et al. 2012). Due to their overwhelming majority in the marine realm, pattern-based surrogates will be the focus of most of this chapter, unless otherwise specified.

2.  Pattern-based surrogates can be constructed based on three distinct approaches: surrogate mapping; assemble first, predict later; and predict first, assemble later A review of the literature on abiotic surrogates from the past decade (Brown et al. 2011) identified three main methods of map production using pattern-based abiotic surrogates: (1) abiotic surrogate mapping (Fig. 13.1a); (2) assemble first, predict later (Fig. 13.1b); and (3) predict first, assemble later (Fig. 13.1c). While methods included in strategies (2) and (3) describe the majority of studies and reflect the rapid development of species distribution modelling and community modelling, the first strategy (1), which uses physical data only, remains a significant component of surrogacy studies that also deserves attention. 1 Abiotic surrogate mapping (Fig. 13.1a) This strategy adopts unsupervised classification methods to infer environmental ‘patterns’, and usually involves little or no in situ ground-truthing. The output maps (often termed seascapes) may be useful for identification of broad geomorphological features (e.g. rocky reefs, canyons) that can be of ecological or biological significance. For example, Huang and colleagues (2012) predicted a range of abiotic surrogates including sediment covers from multibeam acoustic data, which were later used to predict benthic infauna diversity indices (Huang et al. 2014). Przeslawski and colleagues (2011) showed that the ability of seascapes (defined using physical data) to predict benthic diversity patterns was strongly dependent on seascapes, regions and spatial scales. Nevertheless, a continental-scale system of seascapes is one of the most appropriate surrogates for broad-scale benthic community patterns when biological data are limited. 2 Assemble first, predict later (Fig. 13.1b) This strategy is the most commonly used in studies of abiotic surrogates and takes a top-down approach, whereby environmental and biological data are organised or classified separately before being combined (unsupervised classification) (Brown et al. 2011). Environmental data are typically classified into spatial units, which are then used to predict the occurrence of single species or biological communities across space (see Brown et al. 2011 for further detail and examples). 3 Predict first, assemble later (Fig. 13.1c) This strategy takes a bottom-up approach whereby the biological data are used to inform the classification of environmental data (supervised classification). It can be applied from a single-species or a community stand point. A typical example is the definition of bioregions based on the simultaneous consideration of bathymetry and benthic fauna (McArthur et al. 2010).

3.  Some abiotic surrogates perform consistently well across marine ecosystems, others are ecosystem- or region-dependent Abiotic surrogates are often linked to their biological target through a particular ecological process or mechanism, implying that the context-specificity of surrogate performance

127

128

Indicators and Surrogates of Biodiversity and Environmental Change

a)

b)

c)

Fig. 13.1.  Three basic strategies for the production of benthic habitat maps using abiotic surrogates. (Adapted from Brown et al. 2011.)

strongly depends on that of the underlying ecological process (McArthur et al. 2010). Likewise, surrogates can be classified according to three categories based on the type of ecological process linking the surrogate and the target (Meynard and Quinn 2007): (1) gradients of resources consumed by species; (2) direct gradients (exerting a physiological influence on species, such as temperature) or (3) indirect gradients (not necessarily biologically relevant but correlated with other gradients, such as depth or latitude).

13: Abiotic surrogates in support of marine biodiversity conservation

While resource gradients are by definition species-specific, direct gradients such as temperature, productivity or substrate characteristics are often relevant across ecosystems, as are indirect gradients – although the latter often perform at regional to global scales only (see point 4). Less commonly considered (but potentially useful) surrogates include connectivity (Beger et al. 2010a, b); habitat area and isolation (Mellin et al. 2010b); and habitat heterogeneity, thought to promote species richness and population stability (Oliver et al. 2010; Mellin et al. 2014a). An exhaustive review of the different types of surrogates in each category and their associated performance in different ecosystems is beyond the scope of this chapter. However, there have been several detailed reviews that the reader may wish to consult (Mellin et al. 2009; McArthur et al. 2010; Brown et al. 2011; Kachelriess et al. 2014). One important aspect of the transferability of abiotic surrogates (and of their performance across ecosystems) relies on their applicability across distinct technologies that are often restricted to particular habitats or depth ranges. Likewise, a major constraint of remote-sensing techniques is that they can derive information only from the upper layer of the ocean. Space-borne optical sensors penetrate the water to a maximum of only 27 m under the best conditions (Mellin et al. 2009); and air-borne sensors such as the bathymetric light detection and ranging (LIDAR) up to only 46 m (Kachelriess et al. 2014). These remote-sensing technologies can provide abiotic surrogates such as geomorphology, benthic cover, primary productivity and ocean colour, temperature and wind. Beyond these depths, acoustic technologies (e.g. single beam acoustic ground discrimination systems, video scan sonars, multibeam echo sounders) are the only option to document bathymetric or backscatter characteristics of the seafloor such as depth, slope, aspect, curvature or roughness (Brown et al. 2011). These technical constraints mean that, by construction, most abiotic surrogates will be applicable only to a particular type of marine habitat, ecosystem or depth range.

4.  The effectiveness of abiotic surrogates in support of marine conservation is scale-dependent The ecological processes that generate patterns in α- (local species richness), β- (species turnover), and γ-diversity (regional diversity) are scale-dependent, and so are the abiotic surrogates that are linked to such processes through some causative or correlative relationship (Mellin et al. 2009). Spatial gradients such as distance to coast or latitude can be useful at regional scales (Mellin et al. 2010a) and even more so at global scales, which is likely due to local–regional diversity relationships (Caley and Schluter 1997), geometric or geographical constraints on species distributions such as the mid-domain effect (Colwell and Hurtt 1994) or Rapoport’s rule (Rapoport 1982). The mid-domain effect postulates that species’ geographical ranges randomly placed between the bounds of a domain (i.e. ocean basins) have a higher probability of overlapping in the centre of the domain, where species richness is therefore likely to be greatest. Rapoport’s rule postulates that gradients in species richness indirectly arise from the tendency of latitudinal range size to decrease towards the poles in response to greater environmental variability – although many exceptions to this rule exist (Colwell and Hurtt 1994). Based on the former, distance to land mass can be expected to provide a useful surrogate at a global scale (but see Connolly et al. 2003), whereas the latter suggests that latitude can additionally inform us about species distributions based on their geographical ranges. The spatial scale at which abiotic surrogates perform best also depends on the study species, based on their geographical and home ranges (Rees et al. 2014) – a consideration

129

130

Indicators and Surrogates of Biodiversity and Environmental Change

that can become dauntingly complex in multispecies assemblages. To deal with this difficulty, Mellin and colleagues (2014b) proposed a generalisation of geographically weighted regression to compare the performance of different surrogates as the spatial scale (or area of influence) is altered.

5.  The effectiveness of abiotic surrogates is constrained by statistical and sampling considerations Any biological sampling technique is selective to a degree, which can strongly influence our understanding of species–surrogate relationships (McArthur et al. 2010). It is thus recommended, where possible, to compare diversity estimates among sampling techniques and sampled habitats, such as using species accumulation curves. These curves display the number of species recorded with increasing sampling effort: an area is considered well sampled when the curve asymptotes. For selection-based surrogates, the effectiveness of habitat surrogates also appeared highly sensitive to the fish sampling design, in particular, the quantity and location of fish surveys (Van Wynsberge et al. 2012) and to the number of rare species included (Hermoso et al. 2013). From a statistical perspective, the effectiveness of abiotic surrogates can be strongly constrained by the modelling technique and underlying assumptions, such as its ability to handle non-linear species­environment relationships (Pitcher et al. 2012), multicolinearity among predictors (Dormann et al. 2013) or non-stationarity, i.e. the variation of species– environment relationships over space (Fotheringham et al. 2002; Mellin et al. 2014b). Spatial autocorrelation is also a common issue in ecology that, if not properly accounted for, can result in inflated probabilities of Type-I errors and biased estimation of model coefficients (Legendre 1993).

Knowledge gaps 6.  The extent to which the effectiveness of abiotic surrogates is stationary over space and time remains unknown Although substantial knowledge has been gathered on abiotic surrogates of marine biodiversity, we still know very little about the degree to which such relationships are stable over space and time (e.g. seasonal variation in species–environment relationships). Yet, this degree of stationarity is crucial not only in evaluating surrogate utility, but also in formulating management strategies (McArthur et al. 2010). Temporal variation in surrogate effectiveness over time could indeed introduce bias in evaluation due to frequent temporal mismatch between the collection of biological and physical data (Mora 2014). In terms of applications for management, spatial variation in surrogate effectiveness results in limited opportunity for predictive mapping and extrapolation (or interpolation to a lesser extent) to unsampled areas. Interpolation or extrapolation can become particularly risky in the case of non-linear species–surrogate relationships that are indicative of step changes in species composition (Pitcher et al. 2012), for which we might not have an accurate picture in unsampled areas. 7.  How much natural and human-induced disturbance can influence the robustness of abiotic surrogates is unclear Patterns of marine biodiversity observed in many regions today may not reflect the distribution and diversity of species of just a few decades ago, as over fishing, urban development, introduced species and climate change have altered habitats. Therefore,

13: Abiotic surrogates in support of marine biodiversity conservation

physical surrogates of present-day patterns of biodiversity may carry some uncertainty about the degree of impact study areas may have suffered (McArthur et al. 2010). As a result, natural and human-induced disturbance have likely caused a shift in the best surrogates of marine biodiversity. Likewise, fishing effort outperformed physical surrogates for predicting benthic assemblage structure on continental shelf rocky reefs (30–70 m) off South-East Queensland (Richmond and Stevens 2014). Reef protection status (MPA versus non MPA) also outperformed habitat composition and hydrodynamic conditions while predicting tropical invertebrate distribution on New Caledonia reef flats (Jimenez et al. 2012). In addition to influencing surrogate effectiveness, the degree of human alteration of marine systems also could bias our appreciation of desirable surrogates in systematic conservation planning. For example, habitat heterogeneity (promoting species richness and ecosystem stability; see point 3) can be a desirable feature when natural, but undesirable when induced by human activities (e.g. habitat fragmentation resulting from urban development) (Moilanen 2012). Further research is thus required to assess the extent to which disturbance can influence both the effectiveness and the utility of abiotic surrogates of marine biodiversity ­– a necessary but tricky exercise, because virtually no marine habitat remains totally undisturbed.

8.  The cost-effectiveness of different surrogates across spatial scales is yet to be quantified Cost-efficient management solutions are routinely sought in systematic conservation planning, whereby optimisation algorithms are applied to satisfy conservation targets with minimum costs (Moilanen 2012). However, the actual cost of the various surrogates used in such algorithms is yet to be quantified and can imply substantial variation in the total cost of the decision-making process, once surrogate estimation is scaled up to entire regions or ecosystems. This cost will determine the applicability of surrogates at varying spatial scales, which is crucial for developing management strategies (McArthur et al. 2010), and should thus be accounted for in the decision-making process. Estimation of surrogate cost should: (1) include the cost of data collection, treatment, analysis and ground-truthing, both in terms of time, budget and expertise required; and (2) be compared with that of sampling species directly. In some situations, it might be preferable to opt for slightly less effective surrogates if their estimation over large spatial scales is much more affordable. 9.  The impact of limited data availability, sampling selectivity and coarse spatial resolution remain unassessed There has been limited research to date investigating whether different sampling methods (implying variable selectivity) vary in their suitability for constructing surrogates (McArthur et al. 2010). Furthermore, evidence suggests that, for a given method, the sampling design (i.e. number and location of samples) critically affects our perception of surrogate effectiveness (Van Wynsberge et al. 2012). However, how many samples of a given sampling technique are required to obtain an accurate picture of surrogate effectiveness has, to my knowledge, never been explicitly considered. In data-sparse situations, this would at least partially answer the question of whether conservation efforts should be initiated immediately or if more data should be collected first (Moilanen 2012). Data paucity is typical in the marine realm, in particular in many developing countries, where expertise is lacking, ecosystems are under high pressure and conservation action is urgent. This raises the need to develop methods that are less data-demanding and can, for

131

132

Indicators and Surrogates of Biodiversity and Environmental Change

example, be used as priors in more specific analyses (Moilanen 2012). Indeed, expert knowledge or alternative data such as fishing pressure could provide an alternative to poorly performing surrogates (Peckett et al. 2014). This suggests the potential for Bayesian techniques to circumvent these shortcomings in an adaptive learning framework (Caley et al. 2014): that is, a conceptual model that is progressively updated as we learn more about the study system.

10.  Optimised ways of using abiotic surrogates need to be developed for supporting regionalisation approaches that incorporate socioecological considerations Among the many numerically derived regionalisation attempts based on abiotic surrogates, none has gained universal acclaim or has been widely adopted for biodiversity management (McArthur et al. 2010). This is partly due to potential inadequacy of the proposed regionalisation with respect to socio-ecological constraints. For selection-based surrogates, a few studies have incorporated such constraints in the reserve selection process, for example, by comparing output maps of the best conservation network with and without an existing protected area being maintained (Peckett et al. 2014). In another example, Hamel and colleagues (2013) showed that both ecological and socio-economic objectives (i.e. 20% of habitats protected and all existing fishing grounds maintained) could not be met simultaneously, highlighting the shortcoming of fine-resolution coral reef habitat maps as surrogates used in this objective. Such studies remain rare, and optimised ways need to be developed for using surrogates in regionalisation approaches that incorporate economical, societal or governmental considerations.

Conclusion The global conservation network is currently biased and inadequate, and small additions might not prevent species extinctions (Moilanen 2012). This underpins the need to improve the scientific basis behind conservation planning, for which the use of biodiversity surrogates is becoming inevitable (Hermoso et al. 2013). Surrogate-based management is, in fact, the only option in many situations where data are lacking such as in high-sea, pelagic environments (McArthur et al. 2010). Among all possible biodiversity surrogates, abiotic surrogates are cheaper to resolve than species-based surrogates (which sometimes fail to provide effective surrogates; Mellin et al. 2011; Sutcliffe et al. 2014). Furthermore, abiotic surrogates can be mapped at larger scales with the use of space- or ship-borne imagery that limits spatial gaps (Dalleau et al. 2010). These technological developments have thus propelled the construction and application of abiotic surrogates in marine environments, mostly using pattern-based (correlative) approaches whereby various biodiversity metrics can be predicted as a function of abiotic surrogates. This increasing number of studies now allows the identification of surrogates that consistently perform across ecosystems, regions and spatial scales, and the methodological and analytical considerations that constrain surrogate effectiveness. In addition to the identification of key general patterns, the rapid development of surrogate-based marine research has highlighted a strong bias towards pattern- versus selection-based surrogates. I advocate the development of selection-based surrogates in future studies, because these are the only option to effectively support conservation design and management planning (Andréfouët et al. 2012). A better assessment of their cost-effectiveness across spatial scales needs to be conducted, as does an evaluation of how natural and

13: Abiotic surrogates in support of marine biodiversity conservation

human-induced disturbance, or data paucity, can influence surrogate effectiveness. Finally, surrogate-based reserve design procedures need to better incorporate socio-economic constraints, which should ultimately optimise their uptake by end users such as stakeholders and governments.

Glossary Abiotic surrogates  physical entities such as environmental variables or habitats that are used to represent a target entity such as genes, species, or ecosystem metrics.

Acknowledgements Julian Caley and two anonymous reviewers provided helpful comments. CM was funded by an ARC grant (DE140100701).

References Andréfouët S, Hamel MA, Dalleau M (2012) Distinction between effective pattern-based and selection-based biodiversity surrogates is essential: caveats for managers. Marine Ecology Progress Series 452, 287–295. doi:10.3354/meps09655. Ball IR, Possingham HP, Watts M (2009) Marxan and relatives: software for spatial conservation prioritisation. In Spatial Conservation Prioritisation: Quantitative Methods and Computational Tools. (Eds A Moilanen, KAWilson and HP Possingham) pp. 185–195. Oxford University Press, Oxford, UK. Beger M, Linke S, Watts M, Game E, Treml E, Ball I, et al. (2010a) Incorporating asymmetric connectivity into spatial decision making for conservation. Conservation Letters 3(5), 359–368. doi:10.1111/j.1755-263X.2010.00123.x. Beger M, Grantham HS, Pressey RL, Wilson KA, Peterson EL, Dorfman D, et al. (2010b) Conservation planning for connectivity across marine, freshwater, and terrestrial realms. Biological Conservation 143, 565–575. doi:10.1016/j.biocon.2009.11.006. Brown CJ, Smith SJ, Lawton P, Anderson JT (2011) Benthic habitat mapping: a review of progress towards improved understanding of the spatial ecology of the seafloor using acoustic techniques. Estuarine, Coastal and Shelf Science 92, 502–520. doi:10.1016/j. ecss.2011.02.007. Caley MJ, Schluter D (1997) The relationship between local and regional diversity. Ecology 78, 70–80. doi:10.1890/0012-9658(1997)078[0070:TRBLAR]2.0.CO;2. Caley MJ, Fisher R, Mengersen K (2014) Global species richness estimates have not converged. Trends in Ecology & Evolution 29, 187–188. doi:10.1016/j.tree.2014.02.002. Colwell RK, Hurtt GC (1994) Nonbiological gradients in species richness and a spurious Rapoport effect. American Naturalist 144, 570–595. doi:10.1086/285695. Connolly SR, Bellwood DR, Hughes TP (2003) Indo-Pacific biodiversity of coral reefs: deviations from a mid-domain model. Ecology 84, 2178–2190. doi:10.1890/02-0254. Dalleau M, Andréfouët S, Wabnitz CCC, Payri C, Wantiez L, Pichon M, et al. (2010) Use of habitats as surrogates of biodiversity for efficient coral reef conservation planning in Pacific Ocean islands. Conservation Biology 24, 541–552. doi:10.1111/j.1523-1739.2009.01394.x. Dormann CF, Elith J, Bacher S, Buchmann C, Carl G, Carré G, et al. (2013) Collinearity: a review of methods to deal with it and a simulation study evaluating their performance. Ecography 36, 27–46. doi:10.1111/j.1600-0587.2012.07348.x.

133

134

Indicators and Surrogates of Biodiversity and Environmental Change

Fotheringham AS, Brunsdon C, Charlton M (2002) Geographically Weighted Regression: the Analysis of Spatially Varying Relationships. John Wiley & Sons, Chichester, UK. Halpern BS, Longo C, Hardy D, McLeod KL, Samhouri JF, Katona SK, et al. (2012) An index to assess the health and benefits of the global ocean. Nature 488, 615–620. doi:10.1038/ nature11397. Hamel MA, Andréfouët S, Pressey RL (2013) Compromises between international habitat conservation guidelines and small-scale fisheries in Pacific island countries. Conservation Letters 6, 46–57. doi:10.1111/j.1755-263X.2012.00285.x. Hermoso V, Januchowski-Hartley SR, Pressey RL (2013) When the suit does not fit biodiversity: loose surrogates compromise the achievement of conservation goals. Biological Conservation 159, 197–205. doi:10.1016/j.biocon.2012.11.026. Huang Z, Nichol SL, Siwabessy JP, Daniell J, Brooke BP (2012) Predictive modelling of seabed sediment parameters using multibeam acoustic data: a case study on the Carnarvon Shelf, Western Australia. International Journal of Geographical Information Science 26, 283–307. doi:10.1080/13658816.2011.590139. Huang Z, McArthur M, Przeslawski R, Siwabessy J, Nichol S, Brooke B (2014) Predictive mapping of soft-bottom benthic biodiversity using a surrogacy approach. Marine and Freshwater Research 65, 409–424. doi:10.1071/MF13104. Jimenez H, Dumas P, Ponton D, Ferraris J (2012) Predicting invertebrate assemblage composition from harvesting pressure and environmental characteristics on tropical reef flats. Coral Reefs 31, 89–100. doi:10.1007/s00338-011-0820-9. Kachelriess D, Wegmann M, Gollock M, Pettorelli N (2014) The application of remote sensing for marine protected area management. Ecological Indicators 36, 169–177. doi:10.1016/j. ecolind.2013.07.003. Legendre P (1993) Spatial autocorrelation: trouble or new paradigm? Ecology 74, 1659–1673. doi:10.2307/1939924. McArthur MA, Brooke BP, Przeslawski R, Ryan DA, Lucieer VL, Nichol S, et al. (2010) On the use of abiotic surrogates to describe marine benthic biodiversity. Estuarine, Coastal and Shelf Science 88, 21–32. doi:10.1016/j.ecss.2010.03.003. Mellin C, Andréfouët S, Kulbicki M, Dalleau M, Vigliola L (2009) Remote sensing and fish-habitat relationships in coral reef ecosystems: review and pathways for systematic multi-scale hierarchical research. Marine Pollution Bulletin 58, 11–19. doi:10.1016/j. marpolbul.2008.10.010. Mellin C, Bradshaw CJA, Meekan MG, Caley MJ (2010a) Environmental and spatial predictors of species richness and abundance in coral reef fishes. Global Ecology and Biogeography 19, 212–222. doi:10.1111/j.1466-8238.2009.00513.x. Mellin C, Huchery C, Caley MJ, Meekan MG, Bradshaw CJA (2010b) Reef size and isolation determine the temporal stability of coral reef fish populations. Ecology 91, 3138–3145. doi:10.1890/10-0267.1. Mellin C, Delean S, Caley J, Edgar G, Meekan M, Pitcher R, et al. (2011) Effectiveness of biological surrogates for predicting patterns of marine biodiversity: a global meta-analysis. PLoS ONE 6, e20141. doi:10.1371/journal.pone.0020141. Mellin C, Parrott L, Andréfouët S, Bradshaw CJ, MacNeil M, Caley M (2012) Multi-scale marine biodiversity patterns inferred efficiently from habitat image processing. Ecological Applications 22, 792–803. doi:10.1890/11-2105.1. Mellin C, Bradshaw CJA, Fordham DA, Caley MJ (2014a) Strong but opposing beta-diversitystability relationships in coral reef fish communities. Proceedings. Biological Sciences 281, 20131993. doi:10.1098/rspb.2013.1993.

13: Abiotic surrogates in support of marine biodiversity conservation

Mellin C, Mengersen K, Bradshaw CJA, Caley MJ (2014b) Generalizing the use of geographical weights in biodiversity modelling. Global Ecology and Biogeography 23, 1314–1323. doi:10.1111/geb.12203. Meynard CN, Quinn JF (2007) Predicting species distributions: a critical comparison of the most common statistical models using artificial species. Journal of Biogeography 34, 1455–1469. doi:10.1111/j.1365-2699.2007.01720.x. Moilanen A (2012) Spatial conservation prioritization in data-poor areas of the world. Natureza & Conservacao 10, 12–19. doi:10.4322/natcon.2012.003. Mora C (2014) Patterns and processes in reef fish richness. In The ecology of Fishes on Coral Reefs. (Ed. C Mora) pp. 88–96. Academic Press, London. Oliver T, Roy DB, Hill JK, Brereton T, Thomas CD (2010) Heterogeneous landscapes promote population stability. Ecology Letters 13, 473–484. doi:10.1111/j.1461-0248.2010.01441.x. Peckett FJ, Glegg GA, Rodwell LD (2014) Assessing the quality of data required to identify effective marine protected areas. Marine Policy 45, 333–341. doi:10.1016/j. marpol.2013.09.013. Pitcher C, Lawton P, Ellis N, Smith SJ, Incze LS, Wei C-L, et al. (2012) Exploring the role of environmental variables in shaping patterns of seabed biodiversity composition in regional-scale ecosystems. Journal of Applied Ecology 49, 670–679. doi:10.1111/j.1365-2664. 2012.02148.x. Przeslawski R, Currie DR, Sorokin SJ, Ward TM, Althaus F, Williams A (2011) Utility of a spatial habitat classification system as a surrogate of marine benthic community structure for the Australian margin. ICES Journal of Marine Science 68, 1954–1962. doi:10.1093/ icesjms/fsr106. Rapoport EH (1982) Areography: Geographical Strategies of Species. Pergamon, Oxford. Rees MJ, Jordan A, Price OF, Coleman MA, Davis AR (2014) Abiotic surrogates for temperate rocky reef biodiversity: implications for marine protected areas. Diversity & Distributions 20, 284–296. doi:10.1111/ddi.12134. Richmond S, Stevens T (2014) Classifying benthic biotopes on sub-tropical continental shelf reefs: how useful are abiotic surrogates? Estuarine, Coastal and Shelf Science 138, 79–89. doi:10.1016/j.ecss.2013.12.012. Sarkar S, Pressey RL, Faith DP, Margules CR, Fuller T, Stoms DM, et al. (2006) Biodiversity conservation planning tools: present status and challenges for the future. Annual Review of Environment and Resources 31, 123–159. doi:10.1146/annurev.energy.31.042606.085844. Sutcliffe P, Mellin C, Pitcher C, Possingham H, Caley M (2014) Regional-scale patterns and predictors of species richness and abundance across twelve major tropical inter-reef taxa. Ecography 37, 162–171. doi:10.1111/j.1600-0587.2013.00102.x. Van Wynsberge S, Andréfouët S, Hamel MA, Kulbicki M (2012) Habitats as surrogates of taxonomic and functional fish assemblages in coral reef ecosystems: a critical analysis of factors driving effectiveness. PLoS ONE 7, e40997.

135

This page intentionally left blank

‡7HQ&RPPLWPHQWVSSLQGG

$0

14

Building indicators for coupled marine ­ socio-ecological systems Catherine Longo and Benjamin Halpern

Things we know 1 Indicators are useful tools for assessments when they’re relevant, easily measured, updated, interpreted and communicated. 2 Suites of indicators can be used to increase assessment robustness. 3 Suites of complementary indicators, leveraging socio-ecological conceptual frameworks, can be used for multi-dimensional, coupled system assessments. 4 Multi-dimensional assessments are more easily summarised through qualitative narratives, but more rigorously synthesised through objective, quantitative procedures. 5 Despite the demand for integrated assessments, social and ecological dimensions are often evaluated through different methods that are difficult to combine. 6 Setting goals and reference points is a key step for any assessment. Knowledge gaps 7 Societal goals for healthy oceans are not always well defined, and thus hard to track. 8 We need to improve our understanding of functional relationships and nonlinearities within socio-ecological systems. 9 There is still a lot to learn about how people value and prioritise benefits from marine ecosystems. 10 Methods to measure uncertainty for composite socio-ecological indicators still need to be developed.

Introduction In the past decade or so, the need to adopt an ecosystem-based approach to managing marine and coastal resources, rather than single-species and single-sector approaches, has been widely acknowledged (e.g. ecosystem-based management (EBM), Arkema et al. (2006); marine spatial planning (MSP), Crowder et al. (2006); integrated ecosystem assess137

138

Indicators and Surrogates of Biodiversity and Environmental Change

Fig. 14.1. Global score of the Ocean Health Index in 2014. The number shown at the centre of the flower plot is the unweighted average of the 10 scores that comprise the index (for methods see Halpern et al. 2012). Individual goal scores and sub-goal scores and their names are shown on the outside labels and represented by the length of the shaded portion of each of the flower ‘petals’. (Modified from Halpern et al. 2015.)

ments (IEAs), Levin et al. (2009)). The holistic perspective of these approaches acknowledges the complex trophic, biophysical and socioeconomic interactions within ecosystems, and focuses on sustainably delivering benefits to society. Indeed, effective marine management requires satisfying multiple, interacting objectives, including maintaining fisheries yields, conserving biodiversity, maintaining thriving coastal economies, and preserving places of special significance (Levin et al. 2009). To inform such decisions, an assessment of ocean health must incorporate human dimensions (cultural and economic) as well as biophysical aspects (food webs and, environmental forcing) (Mace 2014), and integrate them into tractable, simplified measures such as indicators. Indicators of ecosystem health that integrate social and biological aspects into a single, repeatable, standardised assessment benefit from being comprehensive, but also face novel challenges. These challenges stem mainly from the high number of system attributes that must be evaluated simultaneously, and the fact that heterogeneous measures need to be combined. We illustrate these challenges through example applications of the Ocean Health Index (OHI; Halpern et al. 2012; www.oceanhealthindex.org). The OHI is a composite index designed to assess the health of coupled socio-ecological marine systems by measuring 10 broadly held public goals (herein referred to as ‘10 goals’) for healthy oceans, namely: food provision from mariculture and fisheries; clean water; sense of place; marinerelated livelihoods and economies; biodiversity; carbon storage; coastal protection; artisanal fishing opportunities; coastal tourism and recreation opportunities; and natural products (see Fig. 14.1). We share some lessons learned in the process of developing and

14: Building indicators for coupled marine ­socio-ecological systems

using the OHI at a variety of scales and contexts (Elfes et al. 2014, Halpern et al. 2014, Selig et al. 2014), and propose future research priorities.

1.  Indicators are useful tools for assessments when they’re relevant, easily measured, updated, interpreted and communicated Indicators have the ability to capture the state of complex environmental attributes in a simplified way. Best practices have been extensively described for optimal indicator selection and development in marine management (UNESCO 2003; Rice and Rochet 2005; Rochet and Rice 2005), combining multiple indicators into a single assessment (Rochet et al. 2005; Shin et al. 2005), and setting management goals and reference points (UNESCO 2003; Samhouri et al. 2012; Levin et al. 2013). This literature provides applied examples of well-established principles in indicator theory, making them accessible to marine management technical advisors and practitioners (Rochet and Rice 2005). Good indicators to advise managers must take into account the practical constraints of calculating them (e.g. data availability, frequency of data updates, cost-effectiveness, ease of computation), indicator properties (e.g. directionality, specificity), management needs (e.g. easily communicated to non-technical audiences, tracks an actionable target, timely, responsive to management actions), and public needs (e.g. concreteness, public awareness) (Rice and Rochet 2005, Rochet and Rice 2005). 2.  Suites of indicators can be used to increase assessment robustness Indicators are frequently used as suites of complementary measures, rather than in isolation. There are two main categories of indicator suites: those designed to measure the same attribute, and those designed to measure multiple attributes of the system (covered in the next section). The first type is motivated by the fact that combining different indicators – all intended to capture related aspects of the same thing – provides more robust assessments than relying on a single measure (Rochet and Rice 2005). For example, mean trophic level of fisheries catch (MTLc) has been proposed as a measure of fishing impacts on food-web integrity. It relies on the observation that high commercial fishing pressure has been found to cause bigger, higher trophic level species to decline (Pauly et al. 1998). Nonetheless, MTLc may change due to changes in the species targeted by fishermen, but not mirror changes happening in the food web (e.g. Hornborg et al. 2013). In other words, data quality or incorrect assumptions may cause the indicator to fail to perform as expected. A suite of different indicators (e.g. Shannon et al. 2014), all suggesting similar patterns, increases confidence in the results. When several indicators show diverging patterns, their interpretation is not necessarily straightforward. A common strategy is to rely on the pattern conveyed by the most number of indicators (e.g. Greenstreet and Rogers 2006). If one or more indicators in the suite do not vary independently, however, this gives a false sense of confidence in their results (i.e. double counting). Indeed, the main rule for selecting appropriate suites of indicators is orthogonality (Niemeijer and de Groot 2008); that is, each indicator responds differently (if at all) to potentially confounding variables. Shin and colleagues (2005) identified a set of indicators of community effects of fishing such that each possible combination of their trends unambiguously indicates whether there is a fishing impact (or instead other drivers are at work, such as food availability or recruitment). This requires a thorough understanding of how each indicator responds to the drivers in the system (Link et al. 2010), so that they may be organised through a diagnostic tree (Rochet et al. 2005; Dambacher et al. 2009). These examples highlight the importance of rules for selecting sets of indicators with known and desirable complementary

139

140

Indicators and Surrogates of Biodiversity and Environmental Change

characteristics, in addition to them performing well individually (Kang 2002; Niemeijer and de Groot 2008; OECD 2008).

3.  Suites of complementary indicators, leveraging socio-ecological conceptual frameworks, can be used for multi-dimensional, coupled system assessments The second type of indicator suites are developed to capture multiple different attributes of a system. For example, to track progress towards the Convention for Biological Diversity objectives to halt biodiversity loss, a broad set of indicators has been chosen, each tracking a slightly different aspect of the ecosystem (Secretariat of the Convention on Biological Diversity 2014). Conceptual frameworks, such as the driver–pressure–state– impact–response (DPSIR), have been used to help guide the selection of essential attributes that indicators should capture (Binder et al. 2013). For example, we may be interested in understanding the current state of the system, but we may also want to track whether certain drivers or pressures are changing in response to management actions. These may be used as early warning signals of whether the actions are causing desirable changes (response), before their effects on the system (state) are observed. Many other frameworks have been proposed, designed to identify key interactions to consider when evaluating ecosystem services, but also stressors derived from human interactions with ecosystems (Binder et al. 2013). 4.  Multi-dimensional assessments are more easily summarised through qualitative narratives, but more rigorously synthesised through objective, quantitative procedures As an increasing number of indicators are combined, it becomes difficult to interpret and synthesise the results without structured guidance. Generally, the results from a suite of separate measures are reported in a narrative way (e.g. CBD Aichi target indicators, Millennium Ecosystem Assessment, OECD wellbeing indicators). The advantage of using expert judgment to draw conclusions is that it helps appreciate nuances that could not be captured analytically, or tease out effects of variables that could not be included in the assessment due to modelling constraints (e.g. lack of information, or excessive complexity) (Carpenter et al. 2009). Nevertheless, a narrative synthesis is inherently subjective, and generally not transparent. What constitutes a pattern, and which patterns are deemed to be the most important, is the result of the author(s)’ personal insight, and the implicit assumptions and valuations made are rarely documented. Composite indices help solve these problems by offering a transparent, mathematical way of aggregating indicators into a single index. This helps to distil and communicate results through a simple message, such as the Environmental Performance Index (EPI) or World Bank Governance Index (WGI) (OECD 2008). At the same time, mathematically combining indicators helps to document decisions made, ensuring transparency, reproducibility and accountability. For example, if a set of measures is combined through a weighted average, the relative importance attributed to the different indicators is made explicit by the choice of weights. All the indicators comprising the OHI are combined mathematically to produce a measure of ocean health. While this is a very powerful communication tool, it can be very challenging because it requires quantifying diverse attributes on the same comparable scale, and defining functional relationships among all the indicators that are being combined.

14: Building indicators for coupled marine ­socio-ecological systems

5.  Despite the demand for integrated assessments, social and ecological dimensions are often evaluated through different methods that are difficult to combine Calculating a single score for ocean health requires quantifying the full range of goals assessed, from thriving livelihoods and economies, to maintaining a sense of place, to sustainable seafood provision. Narrative descriptions are often preferred to quantitative measures in evaluating ‘intangible’ values such as cultural, aesthetic, spiritual or moral values (Chan et al. 2012). Bioeconomic approaches to evaluating ecosystem services assign economic value to intangibles by measuring ‘the willingness to pay’ (de Groot et al. 2002), for example, for the enjoyment of special places or preserving iconic species. These techniques find a powerful application in quantitative dynamic models capturing multiple types of socio-ecological interactions. For example, White et al. (2012) were able to identify the range of optimal solutions for having productive fisheries, conserving marine mammals (quantified through whale-watching) and producing energy through wind farms. Nonetheless, they do not always capture the full value and breadth of interactions in a satisfactory way, and are very demanding in terms of modelling complexity and parameterisation, becoming intractable if too many benefits need to be assessed simultaneously. In the OHI, rather than using economic value as the common measurement unit, the status of each goal is expressed as a dimensionless score representing where we are relative to where we should be (i.e. the reference point). In some cases, this is achieved simply by dividing the current state by the reference point (e.g. current sustainable mariculture harvest divided by the optimal mariculture harvest), but it may also be obtained through a more complex function (e.g. mean distance from optimal harvest of fisheries stocks). In this way, the scores for the 10 goals comprising the index (Fig. 14.1) are all expressed on a comparable scale that is driven by societal preferences. This approach affords flexibility in quantifying social values, as it is not constrained by the ability to monetise them. Nonetheless, it remains a simplified and thus imperfect representation. For example, to assess the OHI sub-goal of maintaining livelihoods in marine sectors, we use indicators based on number of jobs and average salaries to capture the level of employment and job quality afforded by marine-related activities. These simple metrics cannot assess certain aspects of quality of life linked to jobs, such as the level of social or gender equity in the labour market. 6.  Setting goals and reference points is a key step for any assessment The first key step towards designing assessment and monitoring plans is to define clear societal goals (Levin et al. 2013). The 10 OHI goals were defined after an extensive literature review, and have stood the test of use across several spatial scales, data availability, geographies and cultures (Elfes et al. 2014; Halpern et al. 2014; Selig et al. 2014). The framework allows dropping any goal(s) that do not apply to a given region (e.g. livelihoods in uninhabited islands) and to customise the way goals are assessed (i.e. indicated) based on different or better data available for a region. Once indicators to track these goals are selected, it is possible to identify what value of the indicator corresponds to the state when the goal is fully satisfied, and setting this as a target or reference point (Samhouri et al. 2012). Determining reference points is ultimately both a scientific and subjective process. Ideally, reference points are established based on a functional relationship between human use of, or interaction with, the ecosystem and the amount of the goal (benefit) achieved (Samhouri et al. 2012). For example, fisheries management commonly adopts quantitative

141

142

Indicators and Surrogates of Biodiversity and Environmental Change

benchmarks of fishing effort that will produce sustainable fisheries (e.g. maximum sustainable yield; MSY), based on stock assessment bioeconomic models. Even when a functional relationship is known, local societal preferences may determine slightly different choices of reference points. When we analysed food provision at a global scale, the chosen target was to maximise sustainable seafood harvest because global food security is an important concern (Halpern et al. 2012). Thus, areas that were under-fished (i.e. where the biomass was above that ensuring MSY) were penalised for not fully using sustainable food resources. However, while in Fiji, which is an island nation where fishing plays an important part in food security, the penalty was applied to any stock with biomass above the biomass that produces MSY (BMSY) (Selig et al. 2014). We applied a more precautionary buffer around the reference point in the US West coast (i.e. under-fishing was gradually penalised only starting from population biomass levels 1.5 times BMSY). In the US West coast, in fact, local seafood is not the main source of nutrition, and local governing authorities put greater emphasis on conserving and rebuilding over maximising yields (Halpern et al. 2014). In summary, we tailored a target that was the result of a bioeconomic model in consideration of regional societal needs.

Knowledge gaps 7.  Societal goals for healthy oceans are not always well -defined, and thus hard to track When goals are poorly understood in a location, the choice of reference points may not reflect societal needs, or may be controversial. For example, in the US west coast there is no clearly defined societal preference for optimal levels of mariculture harvest, and so we used a target set by NOAA (Halpern et al. 2014). However, that target (Nash 2004) was not based on ecological modelling of feasibility, or sustainability, or a trade-off analysis on preferred locations and species. Such potential differences in reference points highlight the need for a better understanding of what local communities want from their region. In the interim, our estimates are a first attempt at envisioning what that may be, and revealed that current harvests represent a very small percentage of potential sustainable production. Finally, even when goals are clear, setting appropriate reference points may be challenging when appropriate data are missing or the processes involved are poorly understood. Alternative approaches include using temporal reference points (i.e. choosing a previous state in time) or spatial reference points (i.e. using the best performing region as reference) (Samhouri et al. 2012). For example, the condition or extent of habitats at some point in the past is often used as a benchmark for what represents habitat health. In the US West coast, a well-studied marine region, high-resolution habitat data are available only for the last decade or two. However, most habitat loss due to anthropogenic activity occurred before then. Using recent habitat extent as a reference point for the optimal condition of salt marshes, sand dunes or sea grass beds would suggest that the habitat conservation goal in the US West coast is fully achieved. On the other hand, using reconstructions of habitat extent in pre-colonial times would suggest a very low goal score with little likelihood of improvement because it is unrealistic to restore most lost habitat. In our study, we arbitrarily set a reference point of recovering 50% of habitat extent lost since the pre-colonial era for salt marshes and 100% of extent loss since the 1950s and 1960s for sand-dunes. This yielded unexpected results because greater historical losses occurred in Northern California, compared with the southern part of the state, so that the former had a lower habitat conservation score than the latter. This is contrary to the common perception, based on current levels of coastal development in the two regions. While on the one hand

14: Building indicators for coupled marine ­socio-ecological systems

it is a useful lesson in how shifting baselines affect people’s perceptions, it also inspires caution in making these decisions. Indeed, better information on what is a realistic and achievable habitat restoration objective may yield a different answer.

8.  We need to improve our understanding of functional relationships and non-linearities within socio-ecological systems Once the range of indicator values corresponding to the minimum and maximum goal scores is established, the next step is to rescale all values. The best approach for this depends on how the indicator changes as a function of the goal it is intended to track. The simplest assumption is that this relationship is linear. For example, for the OHI, where indicators range from 0 to 100, if the current extent of mangroves in a region of interest is 80% of the reference extent, then assuming a linear relationship means a habitat health score of 80. It becomes more challenging when these relationships are non-linear or the relationship is poorly understood. For example, the species biodiversity sub-goal in the OHI is measured as the average extinction risk status across all species assessed in the region (using IUCN Red List assessments, see Selig et al. 2013), with a target for all species to be in ‘least concern’ state. The goal is assigned a score of 0 when 75% of all species are extinct, because this roughly corresponds to the species loss during the five mass extinction events, and 100 if all species are at no risk of extinction. But how fast should the indicator approach 0 as more species become vulnerable or threatened? Does people’s sense of biodiversity loss decrease linearly as more species become threatened or vulnerable to extinction, or is there a threshold number of species that need to decline before people start becoming concerned (e.g. Chapin et al. 2000; Selig et al. 2013)? Because no data exist to determine this relationship, we assumed a linear response (Selig et al. 2013), based on the belief that the average across different people with different value sets is likely to be linear. The actual shape of this relationship is a fundamental knowledge gap. Such potential non-linear relationships between change in the natural system and human values related to those changes are likely to be common, but currently poorly understood and thus difficult to incorporate into indicators such as the OHI. A better understanding of those relationships would make assessments more accurate and, more importantly, may help identify approaching tipping points. 9.  There is still a lot to learn about how people value and prioritise benefits from marine ecosystems Another challenge with composite indicators is how best to combine, or weight, the component parts to produce a single, synthetic assessment. It is not always possible to satisfy all societal goals at a given time. For example, preserving lasting special places by creating a zone where public access is forbidden may reduce tourism and recreation opportunities. The index enables managers and stakeholders to explore the effect of alternate interventions, so as to identify the most acceptable trade-offs (i.e. those that satisfy the priority goals), even though it is at the cost of other goals. In order to appropriately capture these trade-offs, we need to know which components of ocean health people value more than others (Halpern et al. 2013). In the OHI, the 10 goal scores are combined as a weighted average. Societal values can thus be captured by assigning a higher weight to priority goals. Assigning weights based on societal preferences means that the health score is higher when the most valued goals are the ones scoring best, rather than when the lower priority goals are scoring best. Incorporating these unequal weights ensures that the change in the index score is commensurate to how that change affects people, and can help identify when management action is more urgently

143

144

Indicators and Surrogates of Biodiversity and Environmental Change

needed. Importantly, it ensures that the values used to make an overall assessment of health are stated explicitly, making decision makers accountable. Finding consensus priorities across goals can be difficult. Several analytical methods to quantify and model people’s values and preferences have been developed (e.g. Cooke and Goossens 2004; Guo and Sanner 2010; O’Connor et al. 2010), but can be cumbersome to apply. For example, we leveraged decision theory techniques to elicit preferences across the 10 OHI goals among a small, yet representative, group of stakeholders from the US West coast (Halpern et al. 2013). This specific approach, requiring an expert-mediated in-person meeting, is not reproducible at the scale that would be relevant for regional management decisions. However, it was a helpful test of the heuristic value of the OHI framework as a platform for facilitated group discussions about shared objectives among people with different, potentially conflicting, values. Subjective weights enter the calculation of the OHI in several other places. For example, the relative importance of seafood provision from mariculture versus fishing is based on relative yield (but instead could be done based on nutrition quality, for example), and of different marine economic sectors are considered all equally valuable to coastal economies (while, in reality, maintaining certain sectors may be preferred over others). In the absence of well-established knowledge, an advantage of a quantitative framework such as the OHI is that any choices, including the most subjective ones, are fully documented and tracked through the mathematical implementation. In addition to ensuring greater transparency than expert-based synthesis, it also draws attention to knowledge gaps in societal priorities that may be otherwise overlooked. Additionally, it is possible to design sensitivity analyses of the effects of setting different priorities, and of parameter uncertainty.

10.  Methods to measure uncertainty for composite socio-ecological indicators still need to be developed Uncertainty of single indicators or models is calculated in several ways (from frequentist to Bayesian approaches, to model-averaging techniques). Reporting uncertainty along with indicator estimates has become routine practice, at least in some cases (e.g. model averaging for maximum sustainable yield in fisheries, or confidence intervals for threshold values for safe levels of contaminants in the water). As the number of indicators assessed within a composite index increases, existing techniques become increasingly difficult to apply, particularly when mixing very different types of data and models. Uncertainty stems from several aspects of the calculation process, from the accuracy with which the indicator tracks changes in the measure of interest – that is, whether it is a good proxy (e.g. ‘pedigree’ in Christensen and Walters 2004) – to patchy data quality resulting in certain parts of the region or certain aspects of the system being assessed more reliably than others, to heterogeneous data types requiring different methods to measure the variance around estimates (e.g. expert opinionbased categorical measures, versus spatially interpolated environmental data). Capturing all of these aspects of uncertainty is extremely valuable when managers need to decide whether to take action, or to prioritise data collection. Synthesising these diverse sources of error into a confidence envelope will require developing a framework that harmonises expert knowledge of data and model quality with quantitative measures of component uncertainty.

Conclusions A lot of progress has been made in understanding how to best design and communicate scientific information to managers. Tools to provide scientific advice from a broader

14: Building indicators for coupled marine ­socio-ecological systems

ecosystem perspective are also improving (e.g. Kareiva et al. 2011). Integrated assessments ideally track all societally relevant goals, incorporate stakeholder preferences, and appropriately integrate ecological, social and economic dimensions so that they are comparable within a single framework. Such assessments help managers and the general public to understand trade-offs across conflicting objectives and make informed decisions. In order to synthesise this information and transparently communicate it to managers and the public, a quantitative approach is often preferable to an exclusively narrative one. The challenge of aggregating all indicators into a single number lies mainly in accurately representing reference points, interactions, non-linearities and relative weights through mathematical relationships. Future research should aim towards refining methods for combining social and ecological information, understanding utility functions for ecosystem benefits to people, and measuring uncertainty. However, it is important to recognise that this effort needs to be combined with institutional practices on the ground that support interdisciplinary scientific collaborations, and coordinated governance across multiple sectors. The number of people on the planet and their demands on the oceans for material, spiritual and cultural benefits are expected to grow. So too will the potential for overlap and conflicting interests. To assess and track the sustainability of such complex systems, composite measures will become increasingly valuable tools. While a lot of progress has already been made, overcoming the remaining research gaps to incorporate coupled systems assessments in management practice remains an important priority.

Glossary Coupled systems  integrated systems in which people interact with nature, characterised by non-linear interactions and feedbacks. They are the object of interdisciplinary studies combining social and ecological sciences. EBM (ecosystem based management)  an integrated approach to environmental management that incorporates broad ecological principles, so that single sectors or objectives are not considered in isolation, but as part of an interconnected system, and where human and ecological wellbeing are tightly coupled. IEAs (integrated ecosystem assessments)  a formal synthesis and quantitative analysis that integrates social, economic and ecological information, to assess and monitor EBM goals, based on the principle that human and natural factors are intimately entwined. MSP (marine spatial planning)  ‘a public process of analysing and allocating the spatial and temporal distribution of human activities in marine areas to achieve ecological economic, and social objectives that usually have been specified through a political process. Characteristics of marine spatial planning include ecosystem-based, areabased, integrated, adaptive, strategic and participatory.’ (UNESCO MSP Initiative, http://www.unesco-ioc-marinesp.be/marine_spatial_planning_msp)

References Arkema KK, Abramson SC, Dewsbury BM (2006) Marine ecosystem-based management: from characterization to implementation. Frontiers in Ecology and the Environment 4, 525–532. doi:10.1890/1540-9295(2006)4[525:MEMFCT]2.0.CO;2. Binder CR, Hinkel J, Bots PW, Pahl-Wostl C (2013) Comparison of frameworks for analyzing social-ecological systems. Ecology and Society 18(4), 26–44. doi:10.5751/ES-05551-180426.

145

146

Indicators and Surrogates of Biodiversity and Environmental Change

Carpenter SR, Folke C, Scheffer M, Westley M (2009) Resilience: accounting for the noncomputable. Ecology and Society 14(1), 13. Chan KMA, Guerry AD, Balvanera P, Klain S, Satterfield T, Basurto X, et al. (2012) Where are cultural and social in ecosystem services? A framework for constructive engagement. Bioscience 62(8), 744–756. doi:10.1525/bio.2012.62.8.7. Chapin SFIII, Zavaleta ES, Eviner VT, Naylor RL, Vitousek PM, Reynolds HL, et al. (2000) Consequences of changing biodiversity. Nature 405, 234–242. doi:10.1038/35012241. Christensen V, Walters CJ (2004) Ecopath with Ecosim: methods, capabilities and limitations. Ecological Modelling 172, 109–139. doi:10.1016/j.ecolmodel.2003.09.003. Cooke RG, Goossens LHJ (2004) Expert judgment elicitation for risk assessments of critical infrastructures. Journal of Risk Research 7, 643–656. doi:10.1080/1366987042000192237. Crowder L, Oshrenko G, Young O, Airame S, Norse E, Baron N, et al. (2006) Resolving mismatches in U.S. ocean governance. Science 313, 617–618. doi:10.1126/science.1129706. Dambacher JM, Gaughan DJ, Rochet M-J, Rossignol Pa, Trenkel VM (2009) Qualitative modelling and indicators of exploited ecosystems. Fish and Fisheries 10(3), 305–322. doi:10.1111/j.1467-2979.2008.00323.x. de Groot RS, Wilson MA, Boumans RMJ (2002) A typology for the classification, description and valuation of ecosystem functions, goods and services. Ecological Economics 41(3), 393–408. doi:10.1016/S0921-8009(02)00089-7. Elfes CT, Longo C, Halpern BS, Hardy D, Scarborough C, Best BD, et al. (2014) A regionalscale Ocean Health Index for Brazil. PLoS ONE 9(4), e92589. doi:10.1371/journal. pone.0092589. Greenstreet SPR, Rogers SI (2006) Indicators of the health of the fish community of the North Sea: identifying reference levels for an ecosystem approach to management. ICES Journal of Marine Science 63, 573–593. doi:10.1016/j.icesjms.2005.12.009. Guo S, Sanner S (2010) Real-time multiattribute Bayesian preference elicitation with pairwise comparison queries. Journal of Machine Learning Research 9, 289–296. Halpern BS, Longo C, Hardy D, McLeod KL, Samhouri JF, Katona SK, et al. (2012) An index to assess the health and benefits of the global ocean. Nature 488(7413), 615–620. doi:10.1038/nature11397. Halpern BS, Longo CS, McLeod KL, Cooke R, Frishhoff B, Samhouri JSF, et al. (2013) Elicited preferences for components of ocean health in the California Current. Marine Policy 42, 68–73. doi:10.1016/j.marpol.2013.01.019. Halpern BS, Longo C, Scarborough C, Hardy D, Best BD, Doney SC, et al. (2014) Assessing the health of the U.S. west coast with a regional-scale application of the Ocean Health Index. PLoS ONE 9(6), e98995. doi:10.1371/journal.pone.0098995. Halpern BS, Longo C, Stewart Lowndes JS, Best BD, et al. (2015) Patterns and emerging trends in global ocean health. PLoS ONE 10(3), e0117863. doi:10.1371/journal. pone.0117863. Hornborg S, Belgrano A, Bartolino V, Valentinsson D, Ziegler F (2013) Trophic indicators in fisheries: a call for re-evaluation. Biology Letters 9(1), 20121050. doi:10.1098/rsbl.2012.1050. Kang SM (2002) A sensitivity analysis of the Korean composite environmental index. Ecological Economics 43, 159–174. doi:10.1016/S0921-8009(02)00207-0. Kareiva P, Tallis H, Ricketts TH, Daily GC, Polasky S (2011) Natural Capital: Theory and Practice of Mapping Ecosystem Services. Oxford University Press, Oxford. Levin PS, Fogarty MJ, Murawski SA, Fluharty D (2009) Integrated ecosystem assessments: developing the scientific basis for ecosystem-based management of the ocean. PLoS Biology 7(1), e1000014. doi:10.1371/journal.pbio.1000014.

14: Building indicators for coupled marine ­socio-ecological systems

Levin PS, Kelble CR, Shuford RL, Ainsworth C, Dunsmore R, Fogarty MJ, et al. (2013) Guidance for implementation of integrated ecosystem assessments: a US perspective. ICES Journal of Marine Science doi:10.1093/icesjms/fst112. Link JS, Yemane D, Shannon LJ, Coll M, Shin Y, Hill L (2010) Relating marine ecosystem indicators to fishing and environmental drivers: an elucidation of contrasting responses. ICES Journal of Marine Science 67, 787–795. doi:10.1093/icesjms/fsp258. Mace GM (2014) Whose conservation? Science 345(6204), 1558–1560. doi:10.1126/ science.1254704. Nash CE (2004) Achieving policy objectives to increase the value of the seafood industry in the United States: the technical feasibility and associated constraints. Food Policy 29, 621–641. doi:10.1016/j.foodpol.2004.10.001. Niemeijer D, de Groot RS (2008) A conceptual framework for selecting environmental indicator sets. Ecological Indicators 8, 14–25. doi:10.1016/j.ecolind.2006.11.012. O’Connor B, Balasubramanyan R, Routledge BR, Smith NA (2010) From tweets to polls: linking text sentiment to public opinion time series. In Proceedings of the Fourth International AAAI Conference on Weblogs and Social Media. (Eds M Hearts, W Cohen and S Gosling) pp. 122–129. AAAI Press, Menlo Park, CA. OECD (2008) Handbook on Constructing Composite Indicators: Methodology and User Guide. OECD, European Commission, Joint Research Centre. OECD Publishing, Paris, France. Pauly D, Christensen V, Dalsgaard J, Froese R, Torres FJ (1998) Fishing down marine food webs. Science 279, 860–863. doi:10.1126/science.279.5352.860. Rice JC, Rochet M (2005) A framework for selecting a suite of indicators for fisheries management. ICES Journal of Marine Science 62, 516–527. doi:10.1016/j.icesjms.2005.01.003. Rochet M, Rice JC (2005) Do explicit criteria help in selecting indicators for ecosystem-based fisheries management? ICES Journal of Marine Science 62, 528–539. doi:10.1016/j. icesjms.2005.01.007. Rochet M, Trenkel V, Bellail R, Coppin F, Lepape O, Mahe J, et al. (2005) Combining indicator trends to assess ongoing changes in exploited fish communities: diagnostic of communities off the coasts of France. ICES Journal of Marine Science 62(8), 1647–1664. doi:10.1016/j.icesjms.2005.06.009. Samhouri JF, Lester SE, Selig ER, Halpern BS, Fogarty MJ, et al. (2012) Sea sick? Setting targets to assess ocean health and ecosystem services. Ecosphere 3(5), 1–18. doi:10.1890/ ES11-00366.1. Secretariat of the Convention on Biological Diversity (2014) Global Biodiversity Outlook 4. Secretariat of the Convention on Biological Diversity, Montréal. Selig ER, Longo CS, Halpern BS, Best BD, Hardy D, Elfes C, et al. (2013) Assessing global marine biodiversity status within a coupled socio-ecological perspective. PLoS ONE 8(4), e60284. doi:10.1371/journal.pone.0060284. Selig ER, Frazier M, O’Leary JK, Jupiter S, Halpern BS, Longo C, et al. (2015) Measuring indicators of ocean health for an island nation: the Ocean Health Index for Fiji. Ecosystem Services in press, doi:10.1016/j.ecoser.2014.11.007. Shannon LJ, Osman W, Jarre A (2014) Communicating changes in state of the southern Benguela ecosystem using trophic, model-derived indicators. Marine Ecology Progress Series 512, 217–237. doi:10.3354/meps10879. Shin Y, Rochet M, Jennings S, Field J, Gislason H (2005) Using size-based indicators to evaluate the ecosystem effects of fishing. ICES Journal of Marine Science 62(3), 384–396. doi:10.1016/j.icesjms.2005.01.004.

147

148

Indicators and Surrogates of Biodiversity and Environmental Change

UNESCO (2003) A Reference Guide on the Use of Indicators for Integrated Coastal Management – ICAM Dossier 1. IOC Manuals and Guides No. 45. UNESCO, Paris. White C, Halpern BS, Kappel CV (2012) Ecosystem service tradeoff analysis reveals the value of marine spatial planning for multiple ocean uses. Proceedings of the National Academy of Sciences of the United States of America 109(12), 4696–4701. doi:10.1073/pnas.1114215109.

15

The application of genetic indicators in wild populations: potential and pitfalls for genetic monitoring Jennifer Pierson, Gordon Luikart and Michael Schwartz

Things we know 1 Genetic diversity is the foundation of all biodiversity but is seldom considered in studies of biodiversity surrogates. 2 Genetic monitoring may soon be more tractable than many traditional ecological approaches. 3 Genetic indicators are influenced by multiple ecological and evolutionary processes. 4 Genetic indicators can be difficult to interpret. 5 A priori criteria and/or thresholds for interpreting indicators are important to define. Knowledge gaps 6 Determining the limits of non-genetic surrogates for assessing genetic diversity. 7 How will genomics change our understanding of patterns of diversity? 8 Ecologists and geneticists need to work together to define objectives and interpret indicators. 9 How do spatial processes influence metrics derived from idealised populations? 10 Identifying generalisations about how indicators perform.

Introduction ‘Nothing in biology makes sense except in the light of evolution’

– Theodosius Dobzhansky

‘Nothing in evolution makes sense except in the light of population genetics.’

– Michael Lynch

149

150

Indicators and Surrogates of Biodiversity and Environmental Change

The genetic aspects of biodiversity and conservation have been long recognised as important to the viability of populations and evolutionary potential of species (Lande 1988). Yet incorporating genetic considerations into conservation, management, and decision making has lagged behind this recognition (Mace et al. 2003; Laikre et al. 2010). Gene-level (genetic) diversity is required for maintaining fitness and for future evolution and consequently is fundamental to conservation of past and future biodiversity. Thus far, indicators of gene-level diversity have concentrated mostly on agricultural populations, such as food crops (Brown 2008) and forestry products (Boyle 2000). In wild populations, the primary application has been conservation of genetic diversity in the wild relative of crop plants (Laikre et al. 2010). The lack of genetic markers and data on a wide variety of species has contributed to the use of non-genetic surrogates (Mace et al. 2003). However, genetic data have become increasingly more feasible to attain through non-invasive sampling techniques, reduced costs for laboratory analyses, and improved data analysis approaches (Schwartz et al. 2007; Beja-Pereira et al. 2009; Luikart et al. 2010). The era of genomics will continue to increase the ease with which large quantities of genetic information are obtained (Allendorf et al. 2010). Therefore, applying genetic metrics as indicators is becoming more realistic in natural resource management. The increase in genetic monitoring of wild populations has spurred interest in the application and evaluation of genetic metrics as indicators used to describe patterns of genetic diversity (Schwartz et al. 2007; Robert 2011; Hansen et al. 2012; Graudal et al. 2014; Hoban et al. 2014). The same caveats generally apply to using genetic indicators as to any other indicator in that it is important to outline objective(s) and have a clear benchmark or criteria for identifying biologically significant change (Schwartz et al. 2007). In this chapter, we focus on the application of common genetic metrics as genetic indicators, or measures of patterns of genetic diversity, genetic erosion, or genetic vulnerability (Brown 2008). We outline five things we have learned regarding the current application of genetic indicators and their use in natural resource management, and five areas of research that will be most fruitful in moving the field forward.

What we know 1.  Genetic diversity is the foundation of all biodiversity, but is seldom considered in studies of biodiversity surrogates In 1992 at the Earth Summit in Rio de Janeiro, 190 of the world’s leaders signed the Convention on Biological Diversity (CBD) in which they recognised the need to conserve three levels of biological diversity – ecosystems, species within ecosystems, and genes within species – while providing economic opportunities to use these biological resources sustainably. According to article 7 of the CBD, each member nation is required to identify and monitor important components of biological diversity. Yet, only approximately one fifth of the National Biodiversity Strategic and Action Plans acknowledged the need for monitoring at the gene level. This is despite approximately two-thirds of the plans recognising the importance of genetic diversity in wild plant and animal species (Laikre et al. 2010). Laikre and colleagues (2010) suggest that this is partially a human perception problem, where degradation of ecosystems or loss of species is observable to the ‘human eye’, deterioration of the gene pool is largely invisible. At the national level, the critical environmental laws that protect biodiversity are also primarily focused on the protection and monitoring at the ecosystem and species level. For

15: The application of genetic indicators in wild populations

instance, The United States Endangered Species Act (1973) is a law with the goal of conserving two levels of biodiversity – species and ecosystems – while Canada’s Species at Risk Act 2002 (SARA) is more focused on species. In contrast, Australia has passed the Environment Protection and Biodiversity Conservation Act (1999), which focuses on all levels of biodiversity including ‘species, habitats, ecological communities, genes, ecosystem and ecological processes,’ although processes for protecting gene level biodiversity appear lacking. Thus, while many national acts are established to protect and monitor biodiversity, there is a deficiency of considering the most fundamental unit – the gene.

2.  Genetic monitoring may soon be more tractable than many traditional ecological approaches Genetic monitoring is becoming more tractable than traditional ecological and demographic approaches in several species and scenarios. This is primarily due to: (1) increased feasibility of genotyping poor-quality DNA; (2) developments in genomics that allow rapid, inexpensive genotyping; and (3) improved data analysis approaches that allow estimates of important population parameters. Large sample sizes or more representative samples of individuals can often be collected from non-invasive genetic samples, such as faeces, hair or feathers (Waits and Paetkau 2005; Beja-Pereira et al. 2009). Therefore, many parameters, including presence–absence, abundance, effective population size (Ne), number of breeders (spawners), population structure and connectivity (gene flow), can be more easily or reliably estimated using noninvasive genetic samples for species that are secretive or elusive. Certain single nucleotide polymorphism (SNP) chip technologies allow simultaneous genotyping of tens to hundreds of SNPs on tens to hundreds of individuals using low quantity DNA (Campbell and Narum 2008). Concurrent advances in data analysis techniques can provide improved estimates of several commonly used genetic metrics. As an example, the use of genetic indicators to assess and monitor spatial patterns of connectivity, in the context of landscape or climate related features, is increasingly feasible thanks to landscape genetic modelling approaches and software (Sork and Waits 2010). 3.  Genetic indicators are influenced by multiple interacting ecological and evolutionary processes Genetic indicators are often used to assess one aspect of a population (e.g. loss of genetic diversity) that is being influenced by multiple ecological and evolutionary processes (Fig. 15.1). Genetic drift, gene flow, selection and mutation are the primary evolutionary processes. Population size, dispersal behaviour, breeding behaviour, and selective pressures are the primary ecological processes. Many of these processes interact with each other, which can add greatly to the complexity of selecting and interpreting the appropriate indicator and associated criteria. For instance, the loss of genetic diversity over time could be affected by the evolutionary processes of genetic drift, gene flow, selection and the ecological processes of population size, dispersal behaviour and breeding behaviour as well as the interactions between processes (e.g. interaction between gene flow and drift). Careful thought regarding the particular processes that might be under pressure (e.g. habitat loss leads to reduced population size and increased drift) can lead to improved selection of an appropriate metric. Multiple genetic indicators (e.g. metrics) should be used. Simultaneous consideration of different indicators can often help understand causes of genetic changes and also help avoid misinterpretation, such as falsely concluding a population size decline when none occurred (Luikart et al. 1998).

151

152

Indicators and Surrogates of Biodiversity and Environmental Change

Objecves

Ecological factors

Evoluonary processes

Populaon a‰ributes

Genec indicator

Genec diversity

Genec erosion

Genec risk

Populaon size

Dispersal behaviour

Breeding behaviour

Genec dri‚

Migraon

Selecon

Genec variaon

Genec structure

Effecve populaon size

Gene flow

He A K F IS

PA FIS FST

N e Nb

m PA FST

Key to genec indicators He A K FIS PA FST Ne Nb m

expected heterozygosity allelic richness number of alleles inbreeding coefficient proporon admixture fixaon index effecve populaon size effecve number of breeders migraon rate

Fig. 15.1.  The overarching objectives of an assessment or monitoring program and the connections to ecological factors, evolutionary processes, the subsequent population attributes of interest, and the common population genetic metrics that can be used as genetic indicators for these objectives, as connected through the population attributes.

4.  Genetic indicators can be difficult to interpret The use of genetic indicators as an index of population change can be extremely powerful (e.g. Tallmon et al. 2010, 2012). Although genetic data have been widely used to monitor changes in populations of plant and animal species, there has been some reluctance by managers to employ genes as indicators of change. There are many reasons for this hesitation, but a prominent reason may be that changes in genetic signals can be difficult to interpret. For example, the classic population genetics metric FIS is a measure of non-random mating. This measure can become significantly positive if parents on average are more closely related than two mates chosen at random: a phenomenon that can occur in small populations. Similarly, not accounting for underlying population substructure (or a high number of immigrants) can lead to FIS being positive. Other population genetic metrics have the same types of problems, where the signal can be interpreted in multiple, often conflicting, ways. Although this argues for setting genetic questions in a hypothesis testing framework with clear expectations, this is seldom done in genetic assessments and monitoring programs. 5.  A priori criteria and/or thresholds for interpreting indicators are important to define Genetic indicators are used to assess either the current status of a population or to detect trends over time. Therefore, the choice of an appropriate indicator, and subsequent criteria for interpreting the indicator, are inherently tied to the objective of the program. Initial design of a program to assess or monitor patterns of genetic diversity should include clear objectives connected to a sampling design and appropriate analytical methods (Schwartz et al. 2007). Criteria are important to set a priori to ensure the sampling and experimental design are at the correct scale and have the power to detect trends when they exist. Criteria outline what specific conditions need to be met, for the program objectives to be met. This can represent a substantial, but essential, task in the application of genetic indicators in an assessment or monitoring program.

15: The application of genetic indicators in wild populations

For example, a common objective of genetic monitoring programs is to determine if genetic erosion is occurring. This requires selecting an appropriate indicator of genetic diversity and criteria for what constitutes genetic erosion. Criteria may include a consistent decline in allelic richness over a minimum of three sampling periods. General criteria have been proposed for several applications of genetic indicators (Boyle 2000; Hansen et al. 2012; Graudal et al. 2014). Hansen and colleagues (2012) outline criteria for detecting adaptive genetic responses to environmental change and suggested several approaches that can be used to demonstrate the criteria have been met. These criteria include six steps: (1) demonstrate adaptive genetic variation exists; (2) link this genetic variation with a specific environmental stress; (3) test for genetic change over time; (4) test that selection has occurred; (5) link observed genetic changes due to selection to environmental factors; and (6) rule out population replacement. Although each application of genetic indicators may require specific criteria, the examples here guide how simple criteria can be applied to give specific requirements of how patterns in the indicator will relate to the program objective.

Knowledge gaps 6.  Determining the limits of non-genetic surrogates for assessing genetic diversity Surrogates of genetic diversity that do not require genetic data from molecular markers (i.e. non-genetic surrogates) have been proposed as indicators of genetic diversity (Brown et al. 1997; Boyle 2000; Graudal et al. 2014). Boyle (2000) suggest obtaining genetic estimates of gene-level diversity as a last resort, and recommend demographic verifiers as a first step (e.g. population size, population isolation, mating system, species distribution, ecological amplitude). In practice, non-genetic indicators, such as a shift in species range or extent of occupation, are often used to assess genetic diversity and potential for genetic erosion (Forest Practices Authority 2012; Pauls et al. 2013; US Fish and Wildlife Service 2013). For example, the Australian Government’s most recent State of the Environment Report (State of the Environment 2011 Committee 2011) states that genetic diversity is ‘at the heart of biodiversity’. Yet, threats to genetic diversity are primarily assessed through reductions in species distributions (State of the Environment 2011 Committee 2011) and loss of populations (Forest Practices Authority 2012). The theoretical relationships between these non-genetic surrogates and genetic diversity are well understood (Frankham 1996). However, multiple ecological and evolutionary processes that shape patterns of genetic diversity (e.g. population supplementation, habitat fragmentation, dispersal patterns, population history) may be acting on wild populations that complicate these simple predicted relationships (Jackson and Fahrig 2014; Last et al. 2014). For example, the relationship between population size and genetic diversity can be complicated by population supplementation (e.g. the release of hatchery fish). Therefore population size may perform poorly as an indicator of genetic diversity in some cases (Osborne et al. 2012). An important issue with using shifts in species distributions as the sole indicator of genetic erosion is the likely oversight of cryptic diversity, or structured genetic diversity that may represent different lineages that may be on different evolutionary trajectories, within morphospecies (Pauls et al. 2013). This loss of cryptic diversity may result in large underestimates of predicted biodiversity loss (Bálint et al. 2011). As such, the use of coarse non-genetic indicators of genetic diversity in isolation can lead to incorrect management decisions.

153

154

Indicators and Surrogates of Biodiversity and Environmental Change

We recommend the use of genetic indicators in conjunction with non-genetic indicators when assessing genetic diversity status and risks (e.g. Alsos et al. 2012). A combined approach may require less molecular data, therefore provide the ease of non-genetic indicators, yet still include the minimum genetic information necessary to minimise incorrect management decisions. Pfenninger and colleagues (2012) outline a methodology that uses a combined approach. Their approach includes assessing genetic diversity and evaluating species distribution models to assess the risk of the loss of genetic diversity due to global climate change. Additionally, more empirical work in wild populations is needed to refine our understanding of the relationship between non-genetic metrics and patterns of genetic diversity, especially in situations that do not meet the assumptions of predicted theoretical relationships (e.g. spatially structured populations, moderate specialists) (Habel and Schmitt 2012; Neel et al. 2013).

7.  How will genomics change our understanding of patterns of diversity? Genomics will allow more precise estimates of most population genetic metrics, such as mean FST, Ne and Nm (gene flow), using neutral markers (Allendorf et al. 2010). Genomics also has the potential to dramatically improve our basic understanding of processes such as inbreeding depression and local adaptation through examining gene expression, epistasis and genomic architecture (Kardos et al. 2015; Narum and Campbell 2015). By looking at many genes across the genome, studies have been able to separate adaptive genetic diversity from neutral genetic diversity. Local adaptation can be evaluated by the identification of adaptive genes using high FST outlier loci as an indicator of genome regions that are putatively under directional selection, such as for local adaptation (Allendorf et al. 2010). From a conservation management perspective, an important application is defining management units based on both neutral and adaptive diversity (Funk et al. 2012; Moore et al. 2014). For example, Limborg and colleagues (2012) analysed 281 transcriptome derived SNPs in Atlantic Herring (Clupea harengus), a highly migratory small pelagic fish, for elucidating neutral and selected genetic variation among populations. They analysed 607 individuals from 18 spawning locations in the north-east Atlantic, including two temperature clines (5–12ºC) and two salinity clines. They found approximately nine loci that had excessively high FST (genetic differentiation) and also significant correlations with temperature and salinity differences among populations. In a cluster analysis used to identify population groups (conservation units), they identified only three genetically distinct groups of herring when using only the neutral loci. However, four distinct populations were identified when considering the putatively adaptive loci. This is just one example of how genomics can enhance our ability to apply and interpret genetic metrics for conservation management by providing deeper information about wild populations. 8.  Ecologists and geneticists need to work together to define objectives and interpret indicators There is a need for ecologists and population geneticists to work together to understand what are biologically meaningful indicators to monitor in the context of set objectives. There is a long history of debate regarding the importance of genetic considerations in short-term conservation efforts (Lande 1988) and subsequently the fields of ecology and population genetics have developed in parallel until fairly recently. A growing consensus contends that ecological and evolutionary processes, and the interactions between them – termed eco-evolutionary dynamics – are difficult, if not impossible, to separate (Schoener

15: The application of genetic indicators in wild populations

2011). Indeed, eco-evolutionary dynamics are fundamental to consider to design effective conservation efforts (Hendry et al. 2010). Eco-evolutionary dynamics is a swiftly emerging field, bringing ecologists and evolutionary biologists, including population geneticists, together to work on biodiversity conservation (Pierson et al. 2015). The effective use of genetic indicators relies on the combined knowledge of the ecological attributes of the system being evaluated and how these ecological attributes may affect evolutionary processes. This combined knowledge informs which population genetic indicators may inform objectives given the interaction of ecological and evolutionary processes likely acting on the particular system. Thus, collaborative efforts between ecologists and geneticists are essential.

9.  How do spatial processes influence metrics derived from idealised populations? Population genetics has much to offer the field of biodiversity assessments. However, much of population genetics theory has revolved around idealised (i.e. Wright-Fisher) populations, which are basically classical urn models, ideal for sampling and statistics. One of the biggest challenges in using population genetics to monitor natural or wild populations of plants and animals is to understand how spatial processes influence these models and their results. The notion that space can strongly influence our understanding of population genetics is not new. Phenomenon such as the Wahlund effect, which is the reduction of observed heterozygosity in a population caused by substructure, was first coined by Sten Wahlund in 1928. Despite the recognition that many of the classic population genetic metrics are influenced by spatial processes, this fact is often ignored, which can result in highly biased parameter estimates. For example, recent work by Neel and colleagues (2013) shows that Ne estimated from linkage disequilibrium is highly influenced by spatial dynamics. In fact, the interaction between the sampling frame, breeding dynamics and sample size can produce wildly different estimates of Ne on the same landscape. This is because the local spatial genetic structure, or neighbourhood dynamics, can create small-scale Wahlund effects. Currently, we treat these spatial processes as nuisances in our ability to use classic population genetics models. We hope future research will allow us to better describe the processes and use this autocorrelation structure to inform management decisions. 10.  Identifying generalisations about how indicators perform Population genetic metrics perform differently as indicators in different situations (Hoban et al. 2014), which can contribute to difficulty in their interpretation. Theoretical, experimental and empirical work can elucidate when general patterns emerge based on attributes or combinations of attributes such as life history or population history. A few generalisations have emerged in recent years that have been supported by a combination of theoretical and empirical work. For example, the number of alleles (K) and allelic richness (A) have consistently performed better than heterozygosity (He) as an indicator of genetic erosion often caused by population decline (Hoban et al. 2014; Pinsky and Palumbi 2014) because alleles are lost faster than heterozygosity declines (Luikart et al. 1998). An active area of research is on effective population size (Ne): a concept central to conservation management as it represents the population size that genetic drift acts upon. Genetic drift is a stochastic process that changes allele frequencies and reduces genetic diversity in small populations. As such, Ne can be a good indicator of the ‘genetic risk’ a population suffers. In theory, Ne represents the number of individuals that influence genetic change (or loss) in the population. In practice, Ne is notoriously difficult to estimate.

155

156

Indicators and Surrogates of Biodiversity and Environmental Change

The effective number of breeders (Nb), an annual estimate of the number of individuals contributing to the genepool, is more tractable to estimate than Ne in species with overlapping generations (Waples et al. 2013, 2014). However, the genetic indicator of interest is often Ne and the relationship between estimates of Nb and Ne can vary among species making interpretation of estimates of Nb difficult. Waples and colleagues (2013) examined Nb/Ne ratios across a wide variety of plants and animals (including invertebrates) and found that, although Nb/Ne varied widely (~0.3–1.6), two life history traits explained 67% of the variation in this ratio. Thus, these life history traits can be used to improve interpretation of Nb. Another area of research that will improve interpretation of genetic indicators is how the choice of molecular markers affects different metrics. For example, high throughput sequence data are an increasingly common choice that have some fundamentally different properties to microsatellites, which have been the most commonly used marker in recent years. These different properties, such as mutation rate, will likely influence estimates of popular genetic indicators such as heterozygosity and allelic richness (Lozier 2014). As such, research is needed to understand how to select and interpret genetic indicators in light of the type of molecular markers.

Conclusions As anthropomorphic pressures (fragmentation, habitat loss, introduced species and diseases, and climate change) on populations continue to increase, conserving genetic diversity will become more central to conserving the ability of species to rapidly adapt and persist (Barrett and Schluter 2008; Stuart et al. 2014). Increased capacity to readily obtain and analyse large amounts of genetic data from wild populations means this is an exciting time to carefully consider how best to apply genetic metrics as indicators of the genetic ‘health’ of populations. Many of the challenges in effectively applying genetic metrics in wild populations can be tackled with careful a priori objectives and criteria set in a hypothesis testing framework. Indeed, the next decades hold incredible promise for the application of genetic monitoring methods to tackle challenging aspects of biodiversity conservation (Beatty et al. 2014; Ficetola et al. 2015).

References Allendorf FW, Hohenlohe PA, Luikart G (2010) Genomics and the future of conservation genetics. Nature Reviews. Genetics 11, 697–709. doi:10.1038/nrg2844. Alsos IG, Ehrich D, Thuiller W, Eidesen PB, Tribsch A, Schonswetter P, et al. (2012) Genetic consequences of climate change for northern plants. Proceedings. Biological Sciences 279, 2042–2051. doi:10.1098/rspb.2011.2363. Bálint M, Domisch S, Engelhardt CHM, Haase P, Lehrian S, Sauer J, et al. (2011) Cryptic biodiversity loss linked to global climate change. Nature Climate Change 1, 313–318. doi:10.1038/nclimate1191. Barrett R, Schluter D (2008) Adaptation from standing genetic variation. Trends in Ecology & Evolution 23, 38–44. doi:10.1016/j.tree.2007.09.008. Beatty GE, Reid N, Provan J, Lowe A (2014) Retrospective genetic monitoring of the threatened Yellow marsh saxifrage (Saxifraga hirculus) reveals genetic erosion but provides valuable insights for conservation strategies. Diversity & Distributions 20, 529–537. doi:10.1111/ddi.12158.

15: The application of genetic indicators in wild populations

Beja-Pereira A, Oliveira R, Alves PC, Schwartz MK, Luikart G (2009) Advancing ecological understandings through technological transformations in noninvasive genetics. Molecular Ecology Resources 9, 1279–1301. doi:10.1111/j.1755-0998.2009.02699.x. Boyle TJ (2000) Criteria and indicators for the conservation of genetic diversity. In Forest Conservation Genetics: Principles and Practice. (Eds A Young, DH Boshier and TJ Boyle) pp. 239–252. CSIRO Publishing, Melbourne. Brown A (2008) Indicators of Genetic Diversity, Genetic Erosion and Genetic Vulnerability for Plant Genetic Resources for Food and Agriculture. FAO, United Nations, Rome. Brown A, Young A, Burdon J, Christidis L, Clarke G, Coates D, Sherwin W (1997) Genetic Indicators for State of the Environment Reporting. Department of Environment, Sport and Territories, Canberra. Campbell NR, Narum SR (2008) Identification of novel single-nucleotide polymorphisms in Chinook Salmon and variation among life history types. Transactions of the American Fisheries Society 137, 96–106. doi:10.1577/T07-011.1. Ficetola GF, Pansu J, Bonin A, Coissac E, Giguet-Covex C, De Barba M, Gielly L, Lopes CM, Boyer F, Rayé FPG, Taberlet P (2015) Replication levels, false presences and the estimation of the presence/absence from eDNA metabarcoding data. Molecular Ecology Resources 15, 543–556. doi:10.1111/1755-0998.12338. Forest Practices Authority (2012) State of the Forests Tasmania 2012. Forest Practices Authority, Hobart. Frankham R (1996) Relationship of genetic variation to population size in wildlife. Conservation Biology 10, 1500–1508. Funk WC, McKay JK, Hohenlohe PA, Allendorf FW (2012) Harnessing genomics for delineating conservation units. Trends in Ecology & Evolution 27, 489–496. doi:10.1016/j. tree.2012.05.012. Graudal L, Aravanopoulos F, Bennadji Z, Changtragoon S, Fady B, Kjær ED, et al. (2014) Global to local genetic diversity indicators of evolutionary potential in tree species within and outside forests. Forest Ecology and Management 333, 35–51. doi:10.1016/j. foreco.2014.05.002. Habel JC, Schmitt T (2012) The burden of genetic diversity. Biological Conservation 147, 270–274. doi:10.1016/j.biocon.2011.11.028. Hansen MM, Olivieri I, Waller DM, Nielsen EE, Ge MWG (2012) Monitoring adaptive genetic responses to environmental change. Molecular Ecology 21, 1311–1329. doi:10.1111/j.1365-294X.2011.05463.x. Hendry AP, Lohmann LG, Conti E, Cracraft J, Crandall KA, Faith DP, et al. (2010) Evolutionary biology in biodiversity science, conservation, and policy: a call to action. Evolution 64, 1517–1528. Hoban S, Arntzen JA, Bruford MW, Godoy JA, Rus Hoelzel A, Segelbacher G, et al. (2014) Comparative evaluation of potential indicators and temporal sampling protocols for monitoring genetic erosion. Evolutionary Applications 7, 984–998. doi:10.1111/eva.12197. Jackson ND, Fahrig L (2014) Landscape context affects genetic diversity at a much larger spatial extent than population abundance. Ecology 95, 871–881. doi:10.1890/13-0388.1. Kardos M, Luikart G, Allendorf FW (2015) Measuring individual inbreeding in the age of genomics: marker-based measures are better than pedigrees. Heredity online doi:10.1038/ hdy.2015.17. Laikre L, Allendorf FW, Aroner LC, Baker CS, Gregovich DP, Hansen MM, et al. (2010) Neglect of genetic diversity in implementation of the Convention on Biological Diversity. Conservation Biology 24, 86–88. doi:10.1111/j.1523-1739.2009.01425.x.

157

158

Indicators and Surrogates of Biodiversity and Environmental Change

Lande R (1988) Genetics and demography in biological conservation. Science 241, 1455–1460. doi:10.1126/science.3420403. Last L, Lüscher G, Widmer F, Boller B, Kölliker R (2014) Indicators for genetic and phenotypic diversity of Dactylis glomerata in Swiss permanent grassland. Ecological Indicators 38, 181–191. doi:10.1016/j.ecolind.2013.11.004. Limborg MT, Helyar SJ, Bruynm MDE, Taylor MI, Nielsen EE, Ogden R, et al. (2012) Environmental selection on transcriptome-derived SNPs in a high gene flow marine fish, the Atlantic herring (Clupea harengus). Molecular Ecology 21, 3686–3703. doi:10.1111/j.1365-294X.2012.05639.x. Lozier JD (2014) Revisiting comparisons of genetic diversity in stable and declining species: assessing genome-wide polymorphism in North American bumble bees using RAD sequencing. Molecular Ecology 23, 788–801. doi:10.1111/mec.12636. Luikart G, Sherwin WB, Steele BM, Allendorf FW (1998) Usefulness of molecular markers for detecting population bottlenecks via monitoring genetic change. Molecular Ecology 7, 963–974. doi:10.1046/j.1365-294x.1998.00414.x. Luikart G, Ryman N, Tallmon DA, Schwartz MK, Allendorf FW (2010) Estimation of census and effective population sizes: the increasing usefulness of DNA-based approaches. Conservation Genetics 11, 355–373. doi:10.1007/s10592-010-0050-7. Mace GM, Gittleman JL, Purvis A (2003) Preserving the tree of life. Science 300, 1707–1709. doi:10.1126/science.1085510. Moore J, Bourret V, Dionne M, Bradbury I, O’Reilly P, Kent M, Chaput G, Bernatchez L (2014) Conservation genomics of anadromous Atlantic salmon across its North American range: outlier loci identify the same patterns of population structure as neutral loci. Molecular Ecology 23, 5680–5697. doi:10.1111/mec.12972. Narum SR, Campbell NR (2015) Transcriptomic response to heat stress among ecologically divergent populations of redband trout. BMC Genomics 16, 103. doi:10.1186/s12864-0151246-5. Neel MC, McKelvey K, Ryman N, Lloyd MW, Short Bull R, Allendorf FW, et al. (2013) Estimation of effective population size in continuously distributed populations: there goes the neighborhood. Heredity 111, 189–199. doi:10.1038/hdy.2013.37. Osborne MJ, Carson EW, Turner TF (2012) Genetic monitoring and complex population dynamics: insights from a 12-year study of the Rio Grande silvery minnow. Evolutionary Applications 5, 553–574. doi:10.1111/j.1752-4571.2011.00235.x. Pauls SU, Nowak C, Balint M, Pfenninger M (2013) The impact of global climate change on genetic diversity within populations and species. Molecular Ecology 22, 925–946. doi:10.1111/mec.12152. Pfenninger M, Bálint M, Pauls SU (2012) Methodological framework for projecting the potential loss of intraspecific genetic diversity due to global climate change. BMC Evolutionary Biology 12, 224. doi:10.1186/1471-2148-12-224. Pierson JC, Beissinger SR, Bragg JG, Coates DJ, Oostermeijer JGB, Sunnucks P, et al. (2015) Incorporating evolutionary processes into population viability models. Conservation Biology 29, 755–764. doi:10.1111/cobi.12431. Pinsky ML, Palumbi SR (2014) Meta-analysis reveals lower genetic diversity in overfished populations. Molecular Ecology 23, 29–39. doi:10.1111/mec.12509. Robert A (2011) Is the F(ST) a good predictor of extinction? American Naturalist 177, 99–109. doi:10.1086/657434. Schoener TW (2011) The newest synthesis: understanding the interplay of evolutionary and ecological dynamics. Science 331, 426–429. doi:10.1126/science.1193954.

15: The application of genetic indicators in wild populations

Schwartz MK, Luikart G, Waples RS (2007) Genetic monitoring as a promising tool for conservation and management. Trends in Ecology & Evolution 22, 25–33. doi:10.1016/j. tree.2006.08.009. Sork VL, Waits LP (2010) Contributions of landscape genetics: approaches, insights, and future potential. Molecular Ecology 19, 3489–3495. doi:10.1111/j.1365-294X.2010.04786.x. State of the Environment 2011 Committee (2011) Australia State of the Environment 2011. State of the Environment 2011 Committee Canberra. Stuart YE, Campbell TS, Hohenlohe PA, Reynolds RG, Revell LJ, Losos JB (2014) Rapid evolution of a native species following invasion by a congener. Science 346, 463–466. doi:10.1126/science.1257008. Tallmon DA, Gregovich D, Waples RS, Scott Baker C, Jackson J, Taylor BL, et al. (2010) When are genetic methods useful for estimating contemporary abundance and detecting population trends? Molecular Ecology Resources 10, 684–692. doi:10.1111/j.1755-0998. 2010.02831.x. Tallmon DA, Waples RS, Gregovich D, Schwartz MK (2012) Detecting population recovery using gametic disequilibrium-based effective population size estimates. Conservation Genetics Resources 4, 987–989. doi:10.1007/s12686-012-9689-3. US Fish and Wildlife Service (2013) ‘Draft Species Status Assessment Report New Mexico meadow jumping mouse’ (Zapus hudsonius luteus). US Fish and Wildlife Service, Albuquerque, NM. Waits LP, Paetkau D (2005) Noninvasive genetic sampling tools for wildlife biologists: a review of applications and recommendations for accurate data collection. The Journal of Wildlife Management 69, 1419–1433. doi:10.2193/0022-541X(2005)69[1419:NGSTFW]2.0. CO;2. Waples RS, Luikart G, Faulkner JR, Tallmon DA (2013) Simple life-history traits explain key effective population size ratios across diverse taxa. Proceedings. Biological Sciences 280, 20131339. doi:10.1098/rspb.2013.1339. Waples RA, Antao T, Luikart G (2014) Effects of overlapping generations on linkage disequilibrium estimates of effective population size. Genetics 197, 769–780. doi:10.1534/ genetics.114.164822.

159

This page intentionally left blank

‡7HQ&RPPLWPHQWVSSLQGG

$0

16

Use of surrogates in medicine: ideas that may be useful for ecology Peter Lane and Philip Barton

Things we know 1 Terminology has been extensively developed in the context of surrogates in clinical medicine. 2 Causal frameworks are the basis of most approaches to establish surrogacy. 3 The evaluation of surrogates requires assessment of components of the causal framework. 4 Several alternative methods have been put forward to establish causality. 5 Consistency is a key property for surrogates to be of use in predicting the effect of interventions on an outcome. Knowledge gaps 6 Correlation can be misleading, leading to Simpson’s Paradox (aka the Surrogacy Paradox). 7 Covariates can complicate apparent surrogacy relationships. 8 The formalisation of surrogacy requires clear conceptualisation. 9 Ideas from medicine need to be adapted for ecology. 10 Surrogacy for one intervention does not imply surrogacy for another.

Introduction Surrogates are appealing in ecology because of the complexity and diversity of the natural world. Actions need to be planned and decisions made in the absence of information about many of the factors that contribute to the functioning of ecosystems. As a consequence, there is a widespread search for measurements that we can make relatively easily, and that can represent (i.e. be a surrogate for) the myriad interacting factors that characterise an ecosystem, at least well enough for practical decision making. This search has often been informed by correlation between an ecosystem characteristic that we would like to maintain or improve, and a surrogate measurement, or small set of measurements, that represent the ecosystem characteristic of interest. However, there has been limited use of targeted predictive assessment of these surrogates. How a surrogate and outcome behave 161

162

Indicators and Surrogates of Biodiversity and Environmental Change

in response to a specific intervention is central to understanding how the former can predict the latter. Surrogates are used extensively in clinical medicine. The evidence-based revolution in the 1990s grew out of earlier investigations and exposés of how expertise-based medicine, then current, had led to wide variation in practice and extensive weakness in decision making. The application of evidence-based ideas to the use of surrogate outcomes in medicine was well underway by the late 1990s (Biomarkers Definitions Working Group 2001), and individuals and working groups developed methods and processes to ensure that the adoption of surrogates was made on the basis of solid reasoning. There had been spectacular failures of potential surrogates in medicine. One example in cardiovascular medicine was that of arrhythmia as a surrogate for mortality, when it was discovered that the drugs encanaide and flecanaide reduced arrhythmia, which had been accepted as a surrogate, but actually led to a 3-fold increase in mortality (Buyse and Molenberghs 1998). It is important, therefore, that a thorough understanding be established of the relationship between intervention, surrogate and outcome. Below we outline some key ideas from clinical medicine and how they might be used to inform surrogates in ecology.

Things we know 1.  Terminology has been extensively developed in the context of surrogates in clinical medicine In clinical medicine, for each type of disease there is a focus on agreed clinical endpoints (Biomarkers Definitions Working Group 2001). These are measurements or statistics describing disease characteristics that reflect the effect of a therapeutic intervention. A biomarker is a characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes or pharmacologic responses to a therapeutic intervention. Biomarkers are used as diagnostic tools to identify disease, or to classify its stage, or to predict a response to an intervention. Some biomarkers are intended to substitute for clinical endpoints, and they are then called surrogate endpoints. Their use for drug registration is now carefully regulated: the US Food and Drug Administration requires evidence that the drug has an effect on the surrogate endpoint, and that the surrogate is reasonably likely to predict clinical benefit. A clinical surrogate endpoint is therefore conceptually analogous to a surrogate used in ecology. For example, a change in the species richness of a particular taxon (the surrogate) is commonly used as a predictor of a change in the diversity of a broader set of biodiversity (the true variable of interest) (Rodrigues and Brooks 2007). 2.  Causal frameworks are the basis of most approaches to establish surrogacy The main way in which medical statisticians have approached the problem of surrogate validation is with statistical causal frameworks, building on the model shown in Fig. 16.1. The assessment of a surrogate involves examination of the strength of associations between the treatment, surrogate and outcome variables. The widespread use of surrogates in clinical medicine, and the strong need for accuracy and rigour, has resulted in a considerable body of statistical research. Frameworks range in complexity, but build upon the basic model shown in Fig. 16.1, which links the effect of the treatment to the outcome, the treatment to the surrogate, and the surrogate to the outcome. Models can incorporate

16: Use of surrogates in medicine: ideas that may be useful for ecology

Effects not measured by surrogate

Treatment

Surrogate Effects

Outcome Measured to substitute for

Fig. 16.1.  A conceptual model relating a treatment or intervention to an outcome via a surrogate (solid lines) (Biomarkers Definitions Working Group 2001; Barton et al. 2015). This surrogate concept acknowledges the independent effect of the treatment on the outcome, which may not be detected or represented by the surrogate (broken line).

additional sources of variability, such as the effects of additional covariates on the surrogate and response variables.

3.  The evaluation of surrogates requires assessment of components of the causal framework The term validation has often been used in the assessment of surrogates, both in medicine (Buyse et al. 2010) and ecology (Noss 1999). This term can be unsuitable (Biomarkers Definitions Working Group 2001) because it can mean different things to different researchers. From a medical perspective, a surrogate is generally regarded as ‘valid’ when it allows correct inference to be drawn regarding the effect of a treatment on the outcome (Weir and Walley 2006). For this to occur, a strong association needs to be established between the surrogate and the outcome (Begg and Leung 2000). Thus, a valid surrogate is ‘a response variable for which a test of the null hypothesis of no relationship to the treatment is also a valid test of the corresponding null hypothesis of no relationship between the outcome and treatment’ (Prentice 1989, p. 432). Importantly, validation determines whether a surrogate can be generalised across multiple interventions, in different kinds of patients, or different ecosystems. The term evaluation is a more accurate description of this assessment process, and does not imply a simple dichotomy between ‘valid’ and ‘invalid’. Prentice (1989) put forward criteria for evaluation of a surrogate endpoint (S) for a given treatment or intervention (T) and a given clinical endpoint or outcome (O). 1 2 3 4

T must have a significant effect on O. T must have a significant effect on S. S must have a significant effect on O. After accounting for the effect of S, T must have no effect on O.

The last criterion is impractical, because proving the absence of an effect requires infinite power, so Freedman and colleagues (1992) proposed relaxing it. Specifically, estimate PE, the proportion of the effect of T on O that is explained by S, and require that the lower confidence limit of PE is larger than a given proportion (say 0.5, or 0.75). However, even that requires a large amount of data. Buyse and Molenberghs (1998) introduced alternative measures: RE, the relative effect of T on O compared with that of T on S, and γ T, the association between S and O after adjustment for T. These two quantities represent the population-averaged and the individual-specific associations between the treatment effects on S and O, and are represented by fixed and random effects, respectively.

163

164

Indicators and Surrogates of Biodiversity and Environmental Change

4.  Several alternative methods have been put forward to establish causality Surrogates have a better likelihood of being useful if there is a clear mechanistic link between the treatment, surrogate and outcome. This can be conceptualised by placing the surrogate in the causal pathway between treatment and the outcome (see Fig. 16.1). Several alternative statistical methods have been developed to determine causality in these models (Joffe and Greene 2009): 1 conditional independence of variables 2 analysis of direct and indirect effects 3 meta-analysis 4 principal stratification of variables. The first is based on Prentice’s original definition, and the second on work by Taylor and colleagues (2005). Both can be described as approaches in a causal-association paradigm, in which the effect of T on S is associated with its effect on O, across studies or subgroups, so allowing prediction of the effect on O from the effect on S. In contrast, the use of meta-analysis, and the method of principal stratification (Frangakis and Rubin 2002), can be described as approaches in a causal-effects paradigm, in which knowledge of the effects of T on S and of S on O is used to predict the effect of T on O. The meta-analysis approach (Burzykowski et al. 2005) combines data across studies to quantify the effects, but also can be applied across subgroups defined by baseline covariates within a single study.

5.  Consistency is a key property for surrogates to be of use in predicting the effect of interventions on an outcome The so-called surrogate paradox (another name for Simpson’s Paradox, see point 6) has led to interest in establishing consistent surrogates. S is said to be consistent for the effect of T on O if: 1 when S and O are positively associated, a non-positive (non-negative) average causal effect of T on S implies a non-positive (non-negative) average causal effect of T on O 2 when S and O are negatively associated, a non-positive (non-negative) average causal effect of T on S implies a non-negative (non-positive) average causal effect of T on O. A surrogate that is not a consistent surrogate is said to exhibit the surrogate paradox. VanderWeele (2013) showed that the surrogate paradox appears only if one of the following conditions holds: 1 There is a direct effect of treatment on the outcome not through the surrogate and in the opposite direction as that through the surrogate. 2 There is confounding for the effect of the surrogate on the outcome. 3 There is a lack of transitivity so that treatment does not positively affect the surrogate for all the same individuals for whom the surrogate positively affects the outcome. The use of consistent surrogates allows investigators to predict the direction of the effect of the treatment on the outcome simply from the direction of the effect of the treatment on the surrogate.

16: Use of surrogates in medicine: ideas that may be useful for ecology

Table 16.1.  Constructed example of correlated measurements on two taxa in 12 environments Env.

Taxon 1

Taxon 2

Env.

Taxon 1

1

31

24

9

25

Taxon 2 21

2

33

25

10

27

28

3

23

20

11

25

22

4

35

26

12

28

29

5

34

27

6

27

22

7

25

20

8

29

23

Knowledge gaps 6.  Correlation can be misleading, leading to Simpson’s Paradox (aka the Surrogacy Paradox) It is tempting to interpret correlation as a justification for substitution of one measure for another. However, it is easy to be misled by confounding with other variables, which can lead to the non-intuitive phenomenon known as Simpson’s Paradox. There are many instructive and real examples available, but consider here a simple constructed example to illustrate the idea. Say that we measure species richness of two taxonomic groups, with a view to demonstrating that we could use one of them (easier to measure, perhaps) as a surrogate for the other. Of course, we would probably take many measurements, but for simplicity we look at measurements of each taxon made in 12 different environments, presented in Table 16.1. The two counts of species richness from all 12 environments are positively correlated, with a coefficient of 0.61, which is significantly different from 0 (P = 0.036) even with so few observations. Suppose also that there is some variable of ecological interest, such as a management intervention aimed at increasing biodiversity, which takes one value for environments 1−8, and another value for environments 9−12. Again, the two sets of counts are positively correlated within both groups of environments, both with coefficients 0.98 (P < 0.001 and P = 0.02, respectively). The paradox is that the two taxa respond in opposite directions to the change in management: Taxon 1 has a mean of 30 species in Group 1 and 26 in Group 2, while Taxon 2 has 23 and 25, respectively. This arises here because there is a strong interaction between the intervention factor and the relationship between the two counts (Fig. 16.2). Thus, despite the correlation, Taxon 1 is not a useful surrogate for Taxon 2 for the purpose of assessing the choice between management interventions. 7.  Covariates can complicate apparent surrogacy relationships The diagram in Fig. 16.1 is incomplete, as it does not include common causes for the surrogate and the outcome. In randomised trials, randomisation guarantees that the treatment T has no common causes with any other variable: no such assumption is warranted for observational studies, which are the bread-and-butter of ecology. A more realistic diagram would include a further box labelled ‘Covariates’, representing unmeasured and measured covariates that can be considered to be common causes of S and O, with arrows leading to both the Surrogate and the Outcome boxes (Fig. 16.3). There is no reason to believe that all common causes are measured, and often reason to believe that some confounders of the effect of S are unmeasured. When both the surrogate and the clinical out-

165

Indicators and Surrogates of Biodiversity and Environmental Change

Fitted and observed relationship

32 30 28

Taxon 2

166

26 24 22 20 18 24

26

28

30

32

34

Taxon 1

Fig. 16.2.  Fitted model for the constructed data of Table 16.1. Triangles indicate the first level of the intervention factor, and circles the second level.

come are manifestations of a common disease process, such confounding is especially likely due to the presence of common physiological and environmental determinants of both outcomes.

8.  The formalisation of ecological surrogacy requires clear conceptualisation Compared with medical surrogates, ecological surrogates often lack a clear conceptualisation. To follow any of the statistical methodology developed in medicine, we need to be clear what the outcome is and what treatment or intervention we are concentrating on. In the absence of this, it is difficult to communicate a standardised way to assess whether a surrogate predicts an outcome, or to determine how a surrogate performs in relation to a treatment effect. Following a set of standard procedures, such as those proposed by Barton and colleagues (2015), could be used as a way to apply a medical approach to the examination of surrogates in ecology. This includes: (1) conceptualisation of the surrogate model; (2) validation of the surrogate; and (3) evaluation of the surrogate (Barton et al. 2015). 9.  Ideas from medicine need to be adapted for ecology Although established for use in clinical medical sciences, we suggest causal frameworks provide a well-developed foundation for the validation of surrogates in ecology. We use three examples of causal frameworks and give ecological analogues to illustrate how each causal framework might be used to validate a surrogate for three different situations. 1 A basic (naïve) model can apply to the problem of air pollution monitoring, such as the use of lichens as indicators of atmospheric contaminants (Conti and Cecchetti

Treatment

Surrogate

Outcome

Covariates

Fig. 16.3.  A ‘more realistic’ conceptual model relating a treatment to an outcome via a surrogate, with additional effects from unmeasured and measured covariates that might be considered to be common causes of both the surrogate and the outcome.

16: Use of surrogates in medicine: ideas that may be useful for ecology

2001, and see Chapter 8). In this scenario, the concentration of the contaminant is measured in the atmosphere (C) and in the tissue of lichens (CF), and is assessed relative to the desired background level (BL). 2 A general causal model might be applied to an ecosystem restoration problem, such as the application of an intervention aimed at increasing animal diversity in a range of vegetation types (Manning et al. 2013). Here, skink abundance can be measured as a surrogate for overall reptile diversity, and assessed in different types of vegetation. 3 A composite causal model might be applied to a biodiversity monitoring problem such as the use of woodpeckers as indicators of general avian diversity under a range of disturbance pressures (Drever et al. 2008). Notably, the composite model not only allows for incorporation of multiple covariates (e.g. areas of insect attack, different survey years), but also incorporates primary and secondary surrogate variables. In this case, the number of tree cavities is a secondary surrogate for the primary surrogate of woodpecker species richness.

10.  Surrogacy for one intervention does not imply surrogacy for another A drawback of inference about surrogacy based on a single study is that a variable that may be a good surrogate for one treatment–outcome combination may be a poor surrogate for the effect of a different treatment. With multiple studies, meta-analysis provides an approach for examining evidence that the effect of different treatments on the surrogate do a good job of predicting their effects on a given outcome (Burzykowski et al. 2005). The approach is somewhat ‘black-box’ in nature. This is because it is based not on understanding of causal mechanisms but on untestable assumptions that the new treatment is, in some sense, like previously considered treatments. Mechanistic understanding may help strengthen the case for the adequacy of a putative surrogate outcome for a new treatment. Incorporating ideas from the causal-intermediate paradigm into the meta-analytic approach may help with this task.

Conclusion Many aspects of the testing of surrogates in medicine are applicable to the way surrogates are used in ecology (Barton et al. 2015). This includes clear conceptualisation of a surrogate variable in relation to a treatment and outcome, and rigorous statistical evaluation of these relationships. Although challenging in complex ecological systems, a good understanding of causality and the mechanism underpinning potential surrogate relationships is likely to lead to better surrogates. There are also several considerations useful from a statistical perspective, including that correlations can be misleading and that covariates can complicate associations and the determination of causality. Nevertheless, we suggest there is much to learn from the examination of surrogates in other fields, particularly medicine where there is a history of rigorous research.

References Barton PS, Pierson JC, Westgate MJ, Lane PW, Lindenmayer DB (2015) Learning from clinical medicine to improve the use of surrogates in ecology. Oikos 124, 391–398. doi:10.1111/oik.02007.

167

168

Indicators and Surrogates of Biodiversity and Environmental Change

Begg CB, Leung DHY (2000) On the use of surrogate end points in randomized trials. Journal of the Royal Statistical Society: Series A (Statistics in Society) 163, 15–28. doi:10.1111/1467985X.00153. Biomarkers Definitions Working Group .(2001) Biomarkers and surrogate endpoints: preferred definitions and conceptual framework. Clinical Pharmacology and Therapeutics 69, 89–95. doi:10.1067/mcp.2001.113989. Burzykowski T, Molenberghs G, Buyse M (2005) The Evaluation of Surrogate Endpoints. Springer, New York. Buyse M, Molenberghs G (1998) Criteria for the validation of surrogate endpoints in randomized experiments. Biometrics 54, 1014–1029. doi:10.2307/2533853. Buyse M, Sargent DJ, Grothey A, Matheson A, De Gramont A (2010) Biomarkers and surrogate end points: the challenge of statistical validation. Nature Reviews Clinical Oncology 7, 309–317. doi:10.1038/nrclinonc.2010.43. Conti ME, Cecchetti G (2001) Biological monitoring: lichens as bioindicators of air pollution assessment - a review. Environmental Pollution 114, 471–492. doi:10.1016/S02697491(00)00224-4. Drever MC, Aitken KE, Norris AR, Martin K (2008) Woodpeckers as reliable indicators of bird richness, forest health and harvest. Biological Conservation 141, 624–634. doi:10.1016/j.biocon.2007.12.004. Frangakis CE, Rubin DB (2002) Principal stratification in causal inference. Biometrics 58, 21–29. doi:10.1111/j.0006-341X.2002.00021.x. Freedman LS, Graubard BI, Schatzkin A (1992) Statistical validation of intermediate endpoints for chronic diseases. Statistics in Medicine 11, 167–178. doi:10.1002/sim.4780110204. Joffe MM, Greene T (2009) Related causal frameworks for surrogate outcomes. Biometrics 65, 530–538. doi:10.1111/j.1541-0420.2008.01106.x. Manning AD, Cunningham RB, Lindenmayer DB (2013) Bringing forward the benefits of coarse woody debris in ecosystem recovery under different levels of grazing and vegetation density. Biological Conservation 157, 204–214. doi:10.1016/j.biocon.2012.06.028. Noss RF (1999) Assessing and monitoring forest biodiversity: a suggested framework and indicators. Forest Ecology and Management 115, 135–146. doi:10.1016/S03781127(98)00394-6. Prentice RL (1989) Surrogate endpoints in clinical trials: definition and operational criteria. Statistics in Medicine 8, 431–440. doi:10.1002/sim.4780080407. Rodrigues ASL, Brooks TM (2007) Shortcuts for biodiversity conservation planning: the effectiveness of surrogates. Annual Review of Ecology Evolution and Systematics 38, 713–737. doi:10.1146/annurev.ecolsys.38.091206.095737. Taylor JMG, Wang Y, Thiebaut R (2005) Counterfactual links to the proportion of treatment effect explained by a surrogate marker. Biometrics 61, 1102–1111. doi:10.1111/j.1541-0420. 2005.00380.x. VanderWeele TJ (2013) Surrogate measures and consistent surrogates. Biometrics 69, 561–565. doi:10.1111/biom.12071. Weir CJ, Walley RJ (2006) Statistical evaluation of biomarkers as surrogate endpoints: a literature review. Statistics in Medicine 25, 183–203. doi:10.1002/sim.2319.

17

Application of surrogates and indicators to monitoring natural resources John Gross and Barry Noon

Things we know 1 Surrogate species are of limited utility unless the monitoring objectives are clearly articulated and effectively applied by monitoring surrogate species. 2 Failure to recognise or accommodate differences in monitoring and research goals and constraints has impeded the application of surrogate species. 3 Because the complexity of an ecological system cannot be reduced to a single dimension, it is necessary to think of a suite of surrogates that covers the spectrum of key ecological attributes. 4 The set of surrogate measures should be comprehensive in response to the entire scope of the management/disturbance regime, and complementary in the sense that their overlap in information content is low. 5 A broad, coordinated research effort is needed to answer fundamental questions on the selection of surrogate species and their relationships across a range of temporal and spatial scales, methods and taxa. Knowledge gaps 6 The degree to which the status and trends of the surrogate species set reflects the state and temporal dynamics of all the unmeasured species, and the managed ecosystem, is generally unknown and must be validated before full implementation. 7 In a management context, species-specific behaviours, physiology and interspecific interactions complicate our ability to define a reference (or benchmark) condition and threshold for action. 8 A key challenge is to identify and implement surrogate measures that reflect not just the current state of the system but also the long-term dynamics of the system. 9 The scientific knowledge needed to promulgate clear, effective and efficient policy for widespread implementation of monitoring based on surrogate species is inadequate. 10 Wise use of surrogate species could be enhanced by application- or problemspecific ‘best practice’ guides. 169

170

Indicators and Surrogates of Biodiversity and Environmental Change

Introduction Monitoring the status and trends of all the species and ecological processes important to natural resource decision making is simply not possible. A key management challenge is thus to carefully select a set of ecological indicators that efficiently and accurately captures the most important and informative ecological features. Surrogate species have proven to be important indicators for some conservation applications, and they will continue to be an attractive option well into the future. The extensive literature on surrogate species confirms both their importance and the difficulties in their identification and interpretation. For the practitioner, there are significant challenges to designing and implementing monitoring with surrogate species, and these challenges may be amplified by policies or directives that fail to acknowledge important nuances and uncertainties. Challenges include articulating clear and concise monitoring goals, identifying suitable sets of indicators that include surrogates, designing credible and efficient field implementation, analysing observations and interpreting the broader meaning of changes in surrogate species. Our ability to appropriately use surrogate species for monitoring can have profound conservation consequences. The US Fish and Wildlife Service, for example, promulgated policy to use surrogate species as a core approach for monitoring natural resource conditions across the system of refuges, reserves and other lands under their management jurisdiction – an area more than 500  000 km2 (US Fish and Wildlife Service 2008, 2014). Furthermore, US Fish and Wildlife Service and the United States National Marine Fisheries Service proposed an amendment to regulations that implement the Endangered Species Act 1973, which would allow the use of surrogates to express incidental take of listed species (Murphy and Weiland 2014). This amendment would allow surrogate species to be used as proxies for the quantity or extent of incidental take. On the surface, surrogate species have a huge appeal to the unwary – they seem to promise an ability to inexpensively measure the exceedingly complex dynamics of species and ecosystem that underpin natural resource management. Our discussion of surrogates focuses on species or groups of species used to represent other species or aspects of the environment, and we mostly, but not exclusively, consider issues related to monitoring for conservation purposes. The term surrogate is also used generically and may include coarse-filter metrics such as landform, environmental gradients and vegetation communities. We use indicator to more broadly refer to any information-rich attribute of a system. In this context, surrogate species are a subset of indicators.

Things we know 1.  Surrogate species are of limited utility unless the monitoring objectives are clearly articulated and effectively applied by monitoring surrogate species Guidance for developing natural resource management plans or management-oriented research uniformly acknowledges the need for clear goals and objectives (e.g. ‘SMART’ – see www.wikipedia.org). Whether surrogates are being considered to identify the location of a biodiversity ‘hot spot’ or a new protected area, as a proxy for ecosystem health, to evaluate the status and trends of a biological community, or for another purpose, first ask ‘why?’. What do you want to know? Who is the audience for the results? How will they use the results and for what purposes? The need to answer these questions is not unique to the application of surrogate species, but for studies involving surrogates the answers are too frequently ambiguous, largely absent or misdirected. Caro (2010) describes a variety of

17: Application of surrogates and indicators to monitoring natural resources

situations where the fundamental conservation goals and objectives are inappropriate for evaluating the efficacy of a (sometimes pre-selected) surrogate species, or even whether a surrogate species strategy is warranted. The selection of monitoring strategies and methods must be driven by clearly articulated goals (Nichols and Williams 2006) and objectives that can be used to evaluate a full range of options. Surrogate species are used for three fundamental conservation approaches: identifying important conservation areas; evaluating status and trends (as an indicator); and to garner public support (e.g. ‘flagship’ species). Use of surrogate species will likely be unwarranted for purposes that fall outside these three broad areas.

2.  Failure to recognise or accommodate differences in monitoring and research goals and constraints has impeded the application of surrogate species Information in this book clearly shows that considerable expertise is needed to effectively integrate surrogate species into an operational natural resource monitoring program. At the scale at which monitoring is implemented, management organisations usually need outside expertise to help design monitoring. During the first years of development the United States National Park Service Inventory and Monitoring Program, our anecdotal evaluation of projects funded to design monitoring was that ~1/3 successfully delivered results consistent with agreements; ~1/3 eventually led to a useful product, and ~1/3 could be considered failures. Outside experts – mostly university or agency researchers – routinely treated monitoring objectives as research objectives (e.g. to investigate cause-andeffect, rather than status and trends), recommended lengthy and expensive (and frequently novel) programs to design and test sampling protocols, and generally failed to tackle key management needs. These issues are amplified when surrogate species are selected for monitoring because there are relatively few established ‘off the shelf’ protocols, site-specific validation studies are rare, and projections of future conditions – particularly climate – will lead to changes in species ranges and key interactions among species. Furthermore, monitoring objectives that involve surrogate species will need to be more carefully crafted than those involving direct measures. Compared with a direct measure, the additional level of separation between the surrogate and monitoring target introduces uncertainty and noise that must be accommodated. A review of recent scientific literature suggests the very practical constraints of planning and budget cycles, training, staff safety, and capacity to conduct in-house analyses and reporting are still not reflected in communications between researchers and managers. 3.  Because the complexity of an ecological system cannot be reduced to a single dimension, it is necessary to think of a suite of surrogates that covers the spectrum of key ecological attributes At management-relevant scales, ecological systems are characterised by complex sets of attributes that describe composition, structure and ecological functions (Franklin et al. 1981). Because species and other components of an ecological system interact in a multitude of ways, there is no way that the complexity of an ecological system can be reduced to a single dimension. It is therefore necessary to consider a set of surrogate measures. There may be broad indicators of general ecosystem function (e.g. net primary productivity), or measures of the physical environment strongly related to species diversity (Feld et al. 2009), but no single measure can capture the dynamics of multiple interacting species. This simple observation motivates a need to more clearly define criteria and characteris-

171

172

Indicators and Surrogates of Biodiversity and Environmental Change

tics of a parsimonious suite of surrogates. Various studies have examined the scope of coverage of surrogate species or surrogate groups, identifying constraints to application with regard to taxonomic groups (Lawler and White 2008; Sattler et al. 2014), scale (Hess et al. 2006; Cushman et al. 2010), ecological relationships (Hoare et al. 2012) and other factors (Caro 2010; Westgate et al. 2014). Although directed to reserve design, Lambeck (1997) provided an important conceptual foundation for using carefully selected suites of surrogate species, rather than one or a few species (but see Lindenmayer et al. 2002, 2014). We may not know the composition of the optimal set of surrogates, but we do know that a set is required.

4.  The set of surrogate measures should be comprehensive in response to the entire scope of the management/disturbance regime, and complementary in the sense that their overlap in information content is low In most protected areas, management goals encompass a broad range of species, habitats, ecosystems and other properties. Indicators to evaluate the full set of important ecological attributes must collectively be comprehensive of the scope of management/disturbance events and complementary in the sense that overlap in information of each indicator is low. To identify and implement a comprehensive and complementary set of surrogates, we advocate using a combined set of coarse-filter and fine-filter surrogates (Noon et al. 2009; Schwartz 1999). Species whose environmental requirements are not captured by coarsefilter attributes will require a fine-filter assessment. In addition, coarse-filter attributes based on aspects of the physical environment should be more stable than biotic elements in the context of climate change (Beier and Brost 2010; Anderson and Ferree 2010). Combining the two types of surrogates is a conceptually straightforward method that is likely to be the most efficient approach for many complex natural resource monitoring programs. Coarse-filters are characteristics of the environment measured at a landscape scale, often using existing inventory data from satellite images, vegetation classifications, digital elevation models and climate variables. Coarse-filter surrogate measures include environmental attributes that allow the prediction of individual species expected to occupy, for example, a particular vegetation community or to occur at specific locations along an elevation gradient. Coarse-filters can help identify the biophysical template that drives the overall pattern of biological diversity or concentrations of species across large landscapes. Importantly, coarse-filters are defined independently of the habitat requirements of any particular species, attempting instead to capture the habitat needs of an entire species assemblage. Thus, the fundamental assumption of coarse-filters is that measuring the quantities and spatial distribution of biophysical features allows one to predict the occurrence and distribution of most species in the ecosystem. To the extent that coarse-filter attributes capture a large component of the biological diversity within a management unit, they are useful for selecting areas for conservation or targeted management. For example, areas identified by coarse-filter assessments can be prioritised on the basis of their likelihood of loss (vulnerability), degree of irreplaceability (Margules and Pressey 2000) or to maintain landscape connectivity (Brost and Beier 2012). The coarse-filter approach has rarely been tested, and only with mixed results. Coarsefilters have often led to overestimating the presence of species in a region of interest (commission errors). Furthermore, they work better for small-bodied, abundant species than for

17: Application of surrogates and indicators to monitoring natural resources

large-bodied species that are comparatively rare. Species not covered by the coarse-filter are those that are often the focus of fine-filter (i.e. species-specific) approaches, either through direct assessment or by indirect inference from the monitoring of surrogate species.

5.  A broad, coordinated research effort is needed to answer fundamental questions on the selection of surrogate species and their relationships across a range of temporal and spatial scales, methods and taxa In addition to the need to synthesise past research, there is a need for coordinated research to more effectively investigate management-relevant hypotheses, particularly for application at the site or project level. Most published studies evaluating the utility of surrogates, such as the use of one species as a surrogate for other (similar) species, are limited by geographical scope, the ecological system(s) in which the research was conducted and the taxonomic groups examined. A coordinated research effort is required to implement broad-scale designs that are beyond the capabilities of any one, or even a few, research groups. While we agree that surrogates will continue to be used as a practical necessity, more coordinated research is needed to identify the generality of results, including studies of the effects of measurement scale (Cushman et al. 2010; Westgate et al. 2014), use of habitat surrogates versus surrogate species at finer scales (Grantham et al. 2010; Lindenmayer et al. 2014), and more basic considerations of the physiological and ecological similarly of responses among closely related species and other taxonomic groups (Banks et al. 2014). Each study is informative within its own domain, but those domains are too limited and gaps in our understanding of the generality of these relationships severely constrain the appropriate use of surrogate species by management agencies. We are sorely in need of critical examination of the conceptual and mechanistic foundations for the use of surrogate species, across a sufficiently broad range of geographies, biomes and taxonomic groups to support management application at landscape to regional scales.

Knowledge gaps 6.  The degree to which the status and trends of the surrogate species set reflects the state and temporal dynamics of all the unmeasured species, and the managed ecosystem, is generally unknown and must be validated before full implementation Use of surrogate species assumes that species that share biological traits, overlap in their geographic ranges, have comparable area requirements and show similar responses to environmental drivers will have similar abundances and respond in a parallel fashion to management or natural disturbances. These assumptions imply that sympatric species can be classified into distinct groups based on overlap in their ecological requirements, life history traits or trophic interactions (Che-Castaldo and Neel 2012). The ability to accurately predict the responses of the unmeasured target species from observed changes of one or more surrogate species depends critically on meeting these assumptions. Unfortunately, in most applications we are aware of, these assumptions have not been tested before designating surrogate species (Cushman et al. 2010). To provide reliable and defensible inference from the surrogate species set to the target species set, and to be consistent with the ‘best available science’ requirement of most land management agencies, it is imperative that these assumptions be validated (Murphy and Weiland 2014).

173

174

Indicators and Surrogates of Biodiversity and Environmental Change

7.  In a management context, species-specific behaviours, physiology and interspecific interactions complicate our ability to define a reference (or benchmark) condition and threshold for action Enduring management policies and resource plans are founded on quantitative objectives that are frequently anchored to benchmark or reference conditions. Identification of suitable benchmark conditions has posed significant challenges for non-surrogate indicators, and the relative paucity of statistically credible historical data on species distributions, abundances and interspecific relationships means this is even more of an issue for surrogate species. Concepts such as historical range of variation (Landres et al. 1999) are problematic due to the absence of species-specific data, and because both the trends and variances of future environmental conditions are expected to differ from those of the past (Milly et al. 2008). Furthermore, many ecosystems of conservation interest have been highly altered by human use before any species inventories, abundance estimates or detailed ecological studies. In these cases, there is no reference condition anchored to data. In other cases, only contemporary data are available, with a serious risk that interpretations will be coloured by the shifting baseline effect (Berger 2008). Further work is needed to better define methodology and criteria for defining benchmark or reference conditions for a variety of management applications (Stoddard et al. 2006). 8.  A key challenge is to identify and implement surrogate measures that reflect not just the current state of the system but also the longterm dynamics of the system Many species and communities are now exhibiting the manifold effects of rapid and directional climate change, and few, if any, will remain unaffected over the coming decades. Directional climate change provides both opportunities and challenges for the use of surrogates. A large and rapidly expanding body of evidence shows that many species are sensitive integrators of climatic changes, exhibiting changes in phenology, range shifts, sex ratios and a myriad of other traits (Peñuelas et al. 2013). The rapidity, scale and magnitude of climate-driven changes pose a challenge unique to the application of surrogate species because the identification and validation of surrogate relationships, especially those between species, are necessarily based on historical observations. Climate changes are already altering relationships among species in existing communities (Wolkovich and Cleland 2011), and these reorganisations are projected to result in entirely novel species assemblages (Williams and Jackson 2007). Most projections of changes in community composition are driven solely or primarily by climate (using climate-envelope or similar models), but a comprehensive meta-analysis found that biotic interactions were often more important than the direct physical climate effects as a cause of changes in species abundance and distribution (Ockendon et al. 2014). Thus, a key challenge is to identify surrogate species whose responses reflect not only the current state of the system but also provide insight to the long-term dynamics of the system in the face of unknown physical and biotic changes. 9.  The scientific knowledge needed to promulgate clear, effective and efficient policy for widespread implementation of monitoring based on surrogate species is inadequate To our knowledge, there is no clear policy – founded on the best available science – to guide the use of surrogate species for biodiversity monitoring by any large land management agency. A prominent USA management committee struggled for 2 years to draft

17: Application of surrogates and indicators to monitoring natural resources

such guidance and, based on science reviewers, they failed (AMEC 2014). Chapter 1 articulates technical issues with, for example, terminology, but the scientific community has not yet synthesised, distilled and effectively communicated what is known, and clearly not known, into guidance that informs on-the-ground implementation (e.g. at the site scale) and policy (e.g. at state to national scales) for any major biome or taxonomic group. Universal guidance is probably not an achievable goal, but there may now be sufficient knowledge to organise a coordinated effort to identify and document science-based guidance on specific applications of surrogate species. A working group with expertise in both the science and management contexts would be necessary to accomplish such an effort. Policy guidance will surely be restricted in, for example, purpose, scale and taxonomic relevance. An assessment of this sort would promote informed policy, and discourage policy that supports inappropriate use of surrogate species.

10.  Wise use of surrogate species could be enhanced by application- or problem-specific ‘best practice’ guides Initial decisions on the operational use of surrogate species for informing management decisions will normally be based on needs of a specific application (see point 1). Among managers, and apparently scientists, ‘best practices’ for using surrogate species are either unknown or not widely appreciated. When a surrogate-based methodology is well established and considers management goals, such as fish-eating birds as organo-chemical accumulators to detect pollution in the Great Lakes (Newman et al. 2007), decisions on adoption may be simple. But for most situations, there will be multiple approaches and potential indicators. Decisions will include the choice to use (or not use) a surrogate species approach (versus, for example, a direct measure; Lindenmayer and Likens 2011), selection of a sampling design, data analysis, interpretation of results, and so on. Most of these decisions – and the effort needed to implement the approach – could be significantly facilitated by concise and focused syntheses of ‘best practices’. These syntheses would necessarily be more technical and much narrower in scope than policy guidance. The ‘best practice guides’ could be organised around common applications and a published syntax, enunciating best practices with example applications and case studies. In addition to articulating best uses, the guides should include criteria to identify uses that are clearly unacceptable. Due to the scope of issues, scales and biomes, guidance would need to be developed for well-bounded and well-studied situations.

Conclusion Natural resource managers continuously seek indicators of ecological condition that are economical and that influence decision making. In some situations, surrogate species potentially meet these criteria, but the use of surrogate species is limited by both the state of knowledge and by the absence of syntheses that communicate emerging patterns and hypotheses. Integrative syntheses are needed to articulate what we know, what we don’t know, and to highlight emerging issues and promising new applications. Global change – especially climate change – poses huge challenges to natural resource managers and to the application of surrogate species. Rapid and directional change in physical and biotic interactions will surely modify the suitability of surrogate species for monitoring (Milly et al. 2008), and at the extreme some surrogate species will be extirpated from historical ranges. Novel species assemblages and previous unseen environmental conditions will challenge the interpretation of changes in status or trends of surrogate species (Williams and Jack-

175

176

Indicators and Surrogates of Biodiversity and Environmental Change

son 2007). Changing future conditions ensure there will be a need for continuing research and close collaborations between managers and scientists to improve the application of surrogate species to management of natural resources.

Acknowledgements We thank D. Lindenmayer for the opportunity to participate in the workshop that motivated this work. Comments from L. O’Brien and M. Kotzman significantly improved the manuscript.

References AMEC (AMEC Environment & Infrastructure Incorporated) (2014) ‘Final summary report: peer review of technical guidance on selecting species for landscape scale conservation’. AMEC Environmental & Infrastructure Inc., Golden, CO. Anderson MG, Ferree CE (2010) Conserving the stage: climate change and the geophysical underpinnings of species diversity. PLoS ONE 5, e11554. doi:10.1371/journal.pone.0011554. Banks JE, Stark JD, Vargas RI, Ackleh AS (2014) Deconstructing the surrogate species concept: a life history approach to the protection of ecosystem services. Ecological Applications 24, 770–778. doi:10.1890/13-0937.1. Beier P, Brost B (2010) Use of land facets to plan for climate change: conserving the arenas, not the actors. Conservation Biology 24, 701–710. doi:10.1111/j.1523-1739.2009.01422.x. Berger J (2008) Undetected species losses, food webs, and ecological baselines: a cautionary tale from the greater Yellowstone ecosystem, USA. Oryx 42, 139–142. doi:10.1017/ S0030605308001051. Brost BM, Beier P (2012) Use of land facets to design linkages for climate change. Ecological Applications 22, 87–103. doi:10.1890/11-0213.1. Caro TM (2010) Conservation by Proxy: Indicator, Umbrella, Keystone, Flagship, and Other Surrogate Species. Island Press, Washington DC. Che-Castaldo JP, Neel MC (2012) Testing surrogacy assumptions: can threatened and endangered plants be grouped by biological similarity and abundances? PLoS ONE 7, e51659. doi:10.1371/journal.pone.0051659. Cushman SA, McKelvey KS, Noon BR, McGarigal K (2010) Use of abundance of one species as a surrogate for abundance of others. Conservation Biology 24, 830–840. doi:10.1111/j.1523-1739.2009.01396.x. Feld CK, da Silva PM, Sousa JP, de Bello F, Bugter R, Grandin U, et al. (2009) Indicators of biodiversity and ecosystem services: a synthesis across ecosystems and spatial scales. Oikos 118, 1862–1871. doi:10.1111/j.1600-0706.2009.17860.x. Franklin JF, Cromack K Jr, Denison W, McKee A, Maser C, Sedell JR, Swanson FJ, et al. (1981) ‘Ecological characteristics of old-growth Douglas-fir forests’. USDA Forest Service General Technical Report PNW-118, Pacific Northwest Forest and Range Experiment Station, Portland, OR. Grantham HS, Pressey RL, Wells JA, Beattie AJ (2010) Effectiveness of biodiversity surrogates for conservation planning: different measures of effectiveness generate a kaleidoscope of variation. PLoS ONE 5, e11430. doi:10.1371/journal.pone.0011430. Hess GR, Bartel RA, Leidner AK, Rosenfeld KM, Rubino MJ, Snider SB, et al. (2006) Effectiveness of biodiversity indicators varies with extent, grain, and region. Biological Conservation 132, 448–457. doi:10.1016/j.biocon.2006.04.037.

17: Application of surrogates and indicators to monitoring natural resources

Hoare JM, Monks A, O'Donnell CFJ (2012) Can correlated population trends among forest bird species be predicted by similarity in traits? Wildlife Research 39, 469–477. doi:10.1071/ WR11087. Lambeck RJ (1997) Focal species: a multi-species umbrella for nature conservation. Conservation Biology 11, 849–856. doi:10.1046/j.1523-1739.1997.96319.x. Landres PB, Morgan P, Swanson FJ (1999) Overview of the use of natural variability concepts in managing ecological systems. Ecological Applications 9, 1179–1188. Lawler JJ, White D (2008) Assessing the mechanisms behind successful surrogates for biodiversity in conservation planning. Animal Conservation 11, 270–280. doi:10.1111/j.1469-1795.2008.00176.x. Lindenmayer DB, Likens GE (2011) Direct measurement versus surrogate indicator species for evaluating environmental change and biodiversity loss. Ecosystems 14, 47–59. doi:10.1007/ s10021-010-9394-6. Lindenmayer DB, Manning AD, Smith PL, Possingham HP, Fischer J, Oliver I, McCarthy MA (2002) The focal-species approach and landscape restoration: a critique. Conservation Biology 16, 338–345. doi:10.1046/j.1523-1739.2002.00450.x. Lindenmayer DB, Lane PW, Westgate MJ, Crane M, Michael D, Okada S, et al. (2014) An empirical assessment of the focal species hypothesis. Conservation Biology 28, 1594–1603. doi:10.1111/cobi.12330. Margules CR, Pressey RL (2000) Systematic conservation planning. Nature 405, 243–253. doi:10.1038/35012251. Milly PCD, Betancourt J, Falkenmark M, Hirsch RM, Kundzewicz ZW, Lettenmaier DP, et al. (2008) Climate change – stationarity is dead: whither water management? Science 319, 573–574. doi:10.1126/science.1151915. Murphy DD, Weiland PS (2014) The use of surrogates in implementation of the federal Endangered Species Act—proposed fixes to a proposed rule. Journal of Environmental Studies and Sciences 4, 156–162. doi:10.1007/s13412-014-0167-y. Newman SH, Chmura A, Converse K, Kilpatrick AM, Patel N, Lammers E, et al. (2007) Aquatic bird disease and mortality as an indicator of changing ecosystem health. Marine Ecology Progress Series 352, 299–309. doi:10.3354/meps07076. Nichols JD, Williams BK (2006) Monitoring for conservation. Trends in Ecology & Evolution 21, 668–673. doi:10.1016/j.tree.2006.08.007. Noon BR, McKelvey KS, Dickson BG (2009) Multispecies conservation planning on U.S. Federal Lands. In Models for Planning Wildlife Conservation in Large Landscapes. (Eds JJ Millspaugh and FR Thompson III) pp. 51–84. Elsevier Science, San Diego, CA. Ockendon N, Baker DJ, Carr JA, White EC, Almond REA, Amano T, et al. (2014) Mechanisms underpinning climatic impacts on natural populations: altered species interactions are more important than direct effects. Global Change Biology 20, 2221–2229. doi:10.1111/gcb.12559. Peñuelas J, Sardans J, Estiarte M, Ogaya R, Carnicer J, Coll M, et al. (2013) Evidence of current impact of climate change on life: a walk from genes to the biosphere. Global Change Biology 19, 2303–2338. doi:10.1111/gcb.12143. Sattler T, Pezzatti GB, Nobis MP, Obrist MK, Roth T, Moretti M (2014) Selection of multiple umbrella species for functional and taxonomic diversity to represent urban biodiversity. Conservation Biology 28, 414–426. doi:10.1111/cobi.12213. Schwartz MD (1999) Choosing the appropriate scale of reserves for consideration. Annual Review of Ecology Evolution and Systematics 30, 83–108. doi:10.1146/annurev. ecolsys.30.1.83.

177

178

Indicators and Surrogates of Biodiversity and Environmental Change

Stoddard JL, Larsen DP, Hawkins CP, Johnson RK, Norris RH (2006) Setting expectations for the ecological condition of streams: the concept of reference condition. Ecological Applications 16, 1267–1276. doi:10.1890/1051-0761(2006)016[1267:SEFTEC]2.0.CO;2. US Fish and Wildlife Service (2008) A Guide to Implementing the Technical Elements of Strategic Habitat Conservation (Version 1.0). United States Fish and Wildlife Service, Arlington VA, . US Fish and Wildlife Service (2014) Technical Guidance on Selecting Species for Landscape Conservation. United States Fish and Wildlife Service, Arlington VA, . Westgate MJ, Barton PS, Lane PW, Lindenmayer DB (2014) Global meta-analysis reveals low consistency of biodiversity congruence relationships. Nature Communications 5, 3899. doi:10.1038/ncomms4899. Williams JW, Jackson ST (2007) Novel climates, no-analog communities, and ecological surprises. Frontiers in Ecology and the Environment 5, 475–482. doi:10.1890/070037. Wolkovich EM, Cleland EE (2011) The phenology of plant invasions: a community ecology perspective. Frontiers in Ecology and the Environment 9, 287–294. doi:10.1890/100033.

18

Indicators and surrogates in environmental management William H. McDowell

Things we know 1 2 3 4 5 6

Indicator organisms can successfully protect public health. Legal mandates to use indicators can reduce environmental controversy. Simple indicators are effective in assessing the health of running waters. Charismatic indicators generate public support. Indicator organisms can be used to sample pollutants for later analysis. Indicator organisms can be used to directly measure pollutant levels.

Knowledge gaps 7 8 9 10

Are biogeochemical indicators valuable for environmental management? Will measuring environmental DNA reduce reliance on indicators? Can use of indicator organisms as sensors be refined and expanded? To what extent can real-time environmental sensors be used to replace some biological indicators?

Introduction Sound environmental management is of paramount importance in an ever-changing world. With increasing pressures on many key resources, it is incumbent upon the scientific community to anticipate future shifts in community composition, loss of keystone or iconic species or loss of key biogeochemical pathways. Because most legal statutes highlight aspirational or largely unattainable goals, the use of indicators or surrogates becomes extremely important from both a practical and philosophical standpoint. No survey can capture all the species in a habitat, or all the contaminants that might be affecting biotic integrity. Currently, indicators are used to provide rapid assessments, better capture environmental heterogeneity or serve to protect whole systems by protecting a key species. They can also be used to quantify success or failure of an environmental policy or approach, when directly measuring outcomes would not be practical. A wide variety of surrogates and indicators have been used for over 50 years to inform environmental management. The value of indicators lies in their ability to increase our 179

180

Indicators and Surrogates of Biodiversity and Environmental Change

capability to make environmental assessments, simplify the process of determining cause and effect, integrate a wide variety of potentially confounding individual variables, and streamline the monitoring and assessment of desired management outcomes. The earliest use of indicators for environmental management appears to be the use of various groups of microbes to serve as indicators of the hazard posed by pathogenic microbes present in drinking water supplies. This usage of indicators is driven by a need for a rapid, inexpensive way to gauge the likely health hazard of a water resource. Indicators for ecosystem health (as opposed to human health) have also been widely used for over 30 years in river management, based on the principle that composition of invertebrate or fish communities reflects biotic integrity of the study site. Although indicators are widely used, and even codified by statute in some countries, the literature supporting the use of specific indicators is neither comprehensive nor particularly forward looking.

What we know 1.  Indicator organisms can successfully protect public health One of the earliest and best developed uses of indicators for environmental protection is provided by water-quality management to reduce public exposure to water-borne diseases (e.g. Laws 1993; Mason 2002). The use of indicators such as faecal coliforms to alert managers to the possible presence of water-borne pathogens has its origins in the 1890s, when the ‘Wurz method’ of culturing and counting Bacillus coli (later re-named Escherichia coli) was developed, and was employed to protect the drinking water supply of London and other cities by the early 1900s (Ashbolt et al. 2001). In the USA, use of indicators was legally recognised as an important way to protect public health, when the Illinois Supreme Court ruled that use of a poor indicator (colon bacilli) could not be considered an adequate test for the presence of the dangerous pathogen of the genus Salmonella that results in typhoid fever (People v. Bowen (1941) 376Ill.317, 33 NE (2nd) 587). This spurred use of a whole series of other, better indicators – first faecal coliforms, then E. coli, and now more recently enterococci (Mason 2002). An entire body of literature highlights the characteristics of an appropriate indicator: it should be present in sufficient numbers to be readily measured when less common pathogens are present; it should be reliably correlated with these dangerous pathogens that are difficult to work with (and culture) directly; it should be easily measured at modest cost; and measurement must be rapid enough to provide near realtime threat detection in order to treat water supplies or close down bathing beaches in order to protect public health (Mason 2002). Although not all uses of indicators and surrogates need to meet each of these criteria, they do provide a time-tested starting point for assessing the appropriateness of an organism as an indicator or surrogate. 2.  Legal mandates to use indicators can reduce environmental controversy Experience in the USA shows that legal mandates to use indicator or surrogate species for environmental impact assessment can result in reduced controversy and enhanced environmental management. The use of a Representative Important Species for environmental management was codified in the Federal Water Pollution Control Act Amendments of 1972 (FWPCA), Public Law 92–500. The Act defines Representative Important Species as those ‘species which are representative, in terms of their biological needs, of a balanced, indigenous community of shellfish, fish and wildlife.’ The concept was further developed by the US Environmental Protection Agency and its contractors (EPA 1977) to provide a frame-

18: Indicators and surrogates in environmental management

work to assess whether focus on the protection and restoration of a single species would be sufficient to protect the entire fish community of the Hudson River in the state of New York (McDowell 1986). The Striped Bass, Morone saxatilis, was chosen as the Representative Important Species based on its role as a top predator in both coastal ecosystems and tidally influenced rivers such as the Hudson, its sensitivity to the major environmental threat faced by the entire pelagic community (entrainment and impingement of fish and fish larvae in the cooling systems of the power generating systems), its importance as both a sport and commercial fishery, and finally its public appeal as an iconic element of the both the river and the entire coastal fish community. The fact that the presence or absence of these large and energetic fish could be easily noted by anglers and other members of the public who spent time on the river also was important, as success in protecting this iconic species would increase the likelihood of continued support for their management by the general public. Although the use of a Representative Important Species was developed in the 1970s, and is still part of the legislation that governs the siting and operation of power plants, it has never been applied more broadly in the US. Passage of the Endangered Species Act 1973 tended to shift focus to preservation of individual species threatened with extinction, rather than use of indicator species.

3.  Simple indicators are effective in assessing the health of running waters Early work in stream ecology provides strong support for the use of indicator organisms to assess the overall biotic integrity of aquatic ecosystems. Some of the earliest work in this area was conducted by Patrick (1949), who, with considerable prescience, recognised the value of specific indicator taxa of diatoms in assessing the overall health of streams and rivers. Her early work spawned considerable interest in other indicator organisms in streams (Hynes 1960). The development of the current paradigm for rapid assessment of stream health – the proportion of Ephemeroptera, Plecoptera and Trichoptera found in the total community of benthic invertebrates – was developed in the 1980s and still remains an extraordinarily useful assessment that is widely used today in rapid assessments of stream health in the USA (Barbour et al. 1992). The fundamental basis for the metric is the fact that, on average, certain families of aquatic invertebrates are strongly associated with higher and lower levels of various contaminants such as organic matter (Hilsenhoff 1988). This principle has also been applied to fish communities, which show similar relationships between community structure and levels of disturbance or pollution inputs. These relationships been formalised in metrics such as the Index of Biotic Integrity (Karr 1981; Kerans and Karr 1994), which compares the presence and absence of fish that would be expected based on species composition in relatively undisturbed streams. The Ephemeroptera–Plecoptera–Trichoptera framework has been widely applied, because it is a relatively simple assessment of biodiversity at the family level that provides useful insights into the long-term status of stream conditions that could not be easily established with individual grab samples of water quality. The frequency with which these indicator groups is found integrates stream conditions over time, because these cleanwater taxa typically live almost a full year in the stream as aquatic larvae before emerging as winged adults. A second strength is that the organisms are accessible and charismatic. They are easily observed and effectively identified by citizen scientists as well as trained aquatic scientists, and do not require the specialised gear, collecting permits or larger collection volumes required for motile vertebrates such as fish. Furthermore, the study of

181

182

Indicators and Surrogates of Biodiversity and Environmental Change

benthic invertebrates forms the basis of an important leisure pastime, fly fishing, which links individuals to the world around them and rewards careful observation and study.

4.  Charismatic indicators generate public support The case of the Bald Eagle (Haliaeetus leucocephalus) shows the importance of charismatic species as indicators. The Bald Eagle was designated in the USA as Federally Endangered soon after passage of the Endangered Species Act 1973. Successful protection of the Bald Eagle required more aggressive enforcement of restrictions on hunting, harassing and poisoning of all raptors. The banning of DDT in 1972, and the subsequent decline in DDT concentrations and increase in eggshell thickness of many birds, was driven in part by early concerns for the Bald Eagle, with positive spillover for a wide variety of birds. As with the Striped Bass on the Hudson River, the effectiveness of protection efforts for the Bald Eagle was enhanced by its visibility to the public, as well as by its iconic status as an emblem of the USA. Attempts to prevent extinction of the Red Cockaded Woodpecker (Picoides borealis), an elusive bird found at low densities in the south-eastern US, have been less successful. The bird is a medium-sized (18–20 cm) woodpecker that is an obligate cavity nester in mature trees of Longleaf Pine (Pinus palustris). Efforts to reverse its population decline have been underway since before the inception of the US Endangered Species Act 1973. The woodpecker had suffered a catastrophic decline in the early part of the 20th century associated with loss of habitat and nesting cavities. Its biology is distinctive, as it is the only cavitynesting woodpecker that exclusively nests in live trees. Population size remains low compared with historic levels (estimated to be over 1 million breeding clusters), but has increased from 4694 breeding clusters in 1993 to 6105 in 2006 (US Fish and Wildlife Service 2014). The bird is elusive and seldom seen. In this example, the de facto indicator for success of the bird has been preservation of mature stands of Longleaf Pine, as that is relatively easily tracked and managed. Longleaf Pine has been decimated by extensive conversion of stands to high-yield plantation forestry for Yellow Pine (Pinus taeda), a very important commercial species with similar ecological requirements. This use of an indicator tree to serve as a proxy for success of the woodpecker has not been particularly successful. The woodpeckers themselves are not as apparent in the landscape as the Bald Eagle or Striped Bass, likely contributing to lower public support and lower likelihood of success in the preservation and restoration effort for this bird with highly specialised habitat requirements. 5.  Indicator organisms can be used to sample pollutants for later analysis Indicators and surrogates can also play an entirely different role in environmental management when they serve as time-integrated samplers of pollutants for which direct environmental measurements are technically challenging due to issues such as spatial and temporal heterogeneity in pollutant concentrations, or analytical issues such as low levels in complex matrices, such as soils, which require extraction and concentration of the pollutant before chemical analysis. Biomonitors have been developed that are known to accumulate a toxin over time, are relatively stationary, and thus can be used to provide an integrated measurement of contaminant loading in a given geographic location. Use of both dragonflies (Odonata) and mussels as biomonitors has recently been proposed (Buckland-Nicks et al. 2014; Coelho et al. 2014). Mercury (Hg) is still very difficult and expensive to measure at ambient levels in water, despite considerable advances in analytical technology over the past 30 years. Accumulation of Hg in aquatic invertebrates facilitates Hg anal-

18: Indicators and surrogates in environmental management

ysis, because the metal accumulates in the tissue of many aquatic organisms. Recent work by Buckland-Nicks et al. (2014) highlights the use of dragonfly larvae as biomonitors in the relatively pristine lakes of North America, which are subject to significant inputs of mercury via atmospheric deposition from remote sources (Horowitz et al. 2014). The use of dragonfly larvae has an additional benefit, because they are good vehicles for engaging citizen scientists due to their high visibility in the environment as adults and their charismatic ‘monster-like’ appearance as aquatic larvae, which captures the public imagination (Zoellick et al. 2012). Successful citizen science benefits both scientific research and the citizen participants. Use of biomonitor organisms that are visible in the environment, and easily studied, enhances such efforts.

6.  Indicator organisms can be used to directly measure pollutant levels Indicator organisms can be used to directly measure pollutant levels, with the response of the individual organism or the structure of the biotic community used as an indicator of pollutant levels. This is particularly appropriate where pollutants are difficult to measure chemically, vary dramatically over time or are of unknown identity. Lichens have been successfully used as novel indicators that directly respond to the level of pollutants in an urban environment, as described in detail in Chapter 9. The lichens are not analysed in the laboratory for their contaminant levels in this case, but instead the health and persistence of the lichen crust itself is taken as an indication of the level of an environmental contaminant. Monitoring of air pollution levels provides an instructive example. In the case of pollution with ammonia, the critical load to prevent damage to lichens is as low as 1 kg ha–1 year–1 (Johansson et al. 2012). The health and persistence of the lichens thus serves as surrogate for a biogeochemical flux (ammonia deposition) that is hard to measure directly (Pinho et al. 2011). The use of lichens as biomonitors has been proposed as a way to meet Italian regulatory requirements, and thus shows considerable promise as a general biomonitor (Augusto et al. 2010). Overall, indicator or biomonitor species are under-utilised, yet they have significant potential to quantify the nature and magnitude of a variety of environmental contaminants (Shibata et al. 2015).

Knowledge gaps 7.  Are biogeochemical indicators valuable for environmental management? Biogeochemical indicators can provide valuable insights for environmental management. In lakes, for example, the quantity of oxygen dissolved in water is often thought of as a ‘master variable’, and thus might be considered an important indicator of ecosystem health (Wetzel 2001). Dissolved oxygen levels reflect fundamental biological processes – oxygen is produced during photosynthesis and consumed during respiration, with the net balance providing an indicator of the overall trophic status of a lake or other body of water, as well as its suitability for fish and other aquatic life. In addition, dissolved oxygen concentration also drives many other processes, such as phosphorus release from sediments or the balance between production of CO2 and CH4 from lake sediments, which has important implications for greenhouse gas impacts of aquatic ecosystems (Cole et al. 2007). Because of the large amount of ancillary information that can be inferred from studies of the dissolved oxygen profile with depth in lakes, it is a strong candidate for use as a biogeochemical indicator. Another potential biogeochemical indicator is the concentration of nitrate in streams and rivers. Decades of work on forested ecosystems has shown that nitrate concentrations

183

184

Indicators and Surrogates of Biodiversity and Environmental Change

respond quickly to a wide variety of disturbances, such as ice storms (e.g. Houlton et al. 2003), insect outbreaks (Swank et al. 1981), frost damage (Tierney et al. 2001) or hurricanes (McDowell et al. 2013). Nitrate concentrations also may serve as a proxy for biogeochemical changes that are associated with climate change (Shibata et al. 2013). Because streams serve as effective integrators of individual watershed processes that have considerable spatial heterogeneity, such as nitrification, stream chemistry provides an indicator of overall ecosystem function.

8.  Will measuring environmental DNA reduce reliance on indicators? The availability of new sequencing techniques for measuring DNA in environmental samples may reduce reliance on indicator species. An example of the potential for use of environmental DNA to measure the presence of a specific organism is provided by the example of Asian carp that threaten to enter the Laurentian Great Lakes. Two species of carp, Bighead and Silver Carp, were originally imported into the USA for aquaculture in the 1970s. In the 1990s, severe flooding in the southern USA resulted in the escape of these nonnative fish into the Mississippi River, and they have travelled upriver towards the Great Lakes since their escape, earning fame along the way as the notorious flying carp that jump high above the water, periodically landing in boats or striking unwary boaters. Asian carp have enormous potential impacts on the functioning of the Great Lakes, as they are very efficient filter feeders that would compete effectively for the phytoplankton and zooplankton that form the basis of the food web for fisheries such as Walleye. Considerable expense has been incurred to develop ways to prevent these carp from entering the Great Lakes basin, such as development of electrified curtains to prevent their migration into the canals that link the Mississippi drainage to the Great Lakes. An innovative assessment of DNA (environmental or eDNA) has been employed to directly measure the presence of the invasive carp in the susceptible waterways. Fish surveys with nets or electroshocking would be enormously expensive, due to the labour involved in such intensive field work and the wide area that would need to be surveyed. In lieu of developing an indicator of carp presence, recent work by the US Army Corps of Engineers suggests that genetic markers can be used to directly assess actual presence of these fish in the Great Lakes. Use of indicators for public health microbiology also may decline with availability of new genetic assays. The use of faecal coliforms or E. coli as indicators is based on the link between direct entry of human wastes into a water supply and public health risks; the indicator does not directly measure the likelihood that a given disease organism is present in the water supply (Mason 2002). New genetic probes, however, may allow the direct measurement of pathogen presence and abundance. For the cholera-causing Vibrio bacteria, for example, one can test for the presence, type and even geographic origin of a particular strain of cholera, as was demonstrated during the cholera outbreak in Haiti following the 2010 earthquake (Hendriksen et al. 2011). With the decreased costs per analysis and increased speeds of modern genetic techniques, use of indicator organisms to reduce public exposure to contaminated waters may quickly become outmoded. In a similar way, indicators that describe population density of various organisms may become outmoded when remotely sensed data is of sufficient resolution to directly count the organism of interest (Turner 2014). More broadly, development of DNA barcoding or remote sensing technology for a wide variety of species would seem to limit the need for, or utility of, a wide range of indicator organisms. The argument can be made that indicators and surrogates are, in many cases, deflecting attention from a thorough analysis of the threats facing a specific endangered

18: Indicators and surrogates in environmental management

species, the public health risk posed by actual pathogenic microbes, or the danger to ecological integrity posed by an actual invasive species. As argued by Lindenmayer and Likens (2011), some indicators may be better foregone, with effort instead focused on actually measuring the organism at hand.

9.  Can use of indicator organisms as sensors be refined and expanded? Although use of indicator organisms may decline in favour of direct measurements using novel genetic probes, equally compelling arguments can be made for the increased use of individual species or taxa as biomonitors, or overall indicators of environmental health. The use of clam clearing rate to monitor ambient pollution levels provides such an example (Kramer et al. 1989). Bivalves clear their body cavities as part of their feeding and respiratory processes, and the rate of clearing is characteristic of the species as well as its environment. When exposed to toxins, or other irritants, the clearing rate increases, analogous to the coughing response in humans that can be triggered by foreign matter such as dust or pollen in the airways. In the case of clams, the clearing rate can be measured with relatively simple electrodes attached to their shells, with clearing rates responding to a wide variety of pollutants. The mussel clearing rate thus provides a real time indication of changes in pollutant load, which can be an especially valuable early warning system for water supply managers, even though it does not highlight the specific nature of the contaminants that are present. Monitoring animal behaviour thus becomes a proxy for chemical analysis, and modern microelectronics may provide additional opportunities to enhance the use of biomonitors in environmental management. A related effort is underway to use community structure of lichens as the measured variable in response to a chemical pollutant. This concept is developed in Chapter 9. 10.  To what extent can real-time environmental sensors be used to replace some biological indicators? Real-time environmental sensors provide an unparalleled opportunity to measure physical and chemical parameters in the environment in real time. To the extent that these parameters can be used as environmental indicators or surrogates, they can provide valuable real-time information on the status of aquatic and terrestrial ecosystems, and the likely impacts on a wide variety of individual species. Robust sensor technology is now available for analysis of temperature, specific conductance and oxygen in both terrestrial and aquatic systems. Each of these sensors is now in its second or third generation, providing good reliability, precision and accuracy. More recently developed sensors are now available for measuring nitrate and organic matter in water, and they have already proven valuable in understanding ecosystem dynamics of these important solutes (Pellerin et al. 2011). Many of these sensors will be deployed in a new national network in the USA, the National Ecological Observatory Network, which will include dozens of aquatic and terrestrial sites with sensors across North America that will provide unparalleled datasets that can be used to characterise the function of terrestrial and aquatic ecosystems at the continental scale (McDowell 2015; Utz et al. 2013). Because these sensor networks will be augmented with more standard biological measurements, the network will provide opportunities to assess the utility of continuously sensed data to replace or enhance the use of more traditional biological indicators. In the case of small streams and rivers, for example, collection of data on the well-established indicator groups Ephemeroptera, Plecoptera and Trichoptera can be used to determine if sensor output can serve as a real-time indicator of stream ecosystem integrity.

185

186

Indicators and Surrogates of Biodiversity and Environmental Change

Conclusion Indicators have been used for over 100 years in environmental management. The earliest uses involved protection of public health by measuring indicator microbes that were indicative of faecal contamination that carries with it a significant risk from of water-borne diseases. Other indicators have been developed to assess the ecological integrity of aquatic ecosystems, specifically small streams. Community composition of various groups, such as algae, aquatic insects and fish, has been used to indicate the health of the stream ecosystem. Legal requirements to use indicator organisms such as the mandate to use ‘Representative Important Species’ in determining the environmental impacts of power plant siting and operation in the USA have proven successful, and appear to have reduced environmental conflicts. Indicator species that are charismatic or highly visible tend to engender more public interest and support, and thus can lead to more effective environmental management. The availability of new approaches to environmental sensing and genetic analysis of the environment may reduce the use of indicators in favour of direct measurements. New approaches in which indicator organisms are used to sample the level of environmental contaminants have been used successfully for a few taxa, such as lichens. It is likely that the approach could be used more widely in the future. Use of living organisms to indicate the level of contaminants in air or water has been successfully implemented in a few cases, but it is unclear whether this approach could be used more broadly. Development of new realtime environmental sensors that can directly assess environmental conditions may reduce the use of indicator organisms in the future.

References Ashbolt NJ, Grabow WOK, Snozzi M (2001) Indicators of microbial water quality. In Water quality: Guidelines, Standards and Health. Assessment of Risk and Risk Management for Water-related Infectious Disease. (Eds L Fewtrell and J Bartram). IWA for World Health Organization, London. Augusto S, Maguas C, Matos J, Pereira MJ, Branquinho C (2010) Lichens as an integrating tool for monitoring PAH atmospheric deposition: a comparison with soil, air and pine needles. Environmental Pollution 158(2), 483–489. doi:10.1016/j.envpol.2009.08.016. Barbour MT, Plafkin JL, Bradley BP, Graves CG, Wisseman RW (1992) Evaluation of EPA’s rapid bioassessment benthic metrics: metric redundancy and variability among reference stream sites. Environmental Toxicology and Chemistry 11(4), 437–449. doi:10.1002/ etc.5620110401. Buckland-Nicks A, Hillier KN, Avery TS, O'Driscoll NJ (2014) Mercury bioaccumulation in dragonflies (Odonata: Anisoptera): examination of life stages and body regions. Environmental Toxicology and Chemistry 33(9), 2047–2054. doi:10.1002/etc.2653. Coelho JP, Duarte AC, Pardal MA, Pereira ME (2014) Scrobicularia plana (Mollusca, Bivalvia) as a biomonitor for mercury contamination in Portuguese estuaries. Ecological Indicators 46, 447–453. doi:10.1016/j.ecolind.2014.07.015. Cole JJ, Prairie YT, Caraco NF, McDowell WH, Tranvik LJ, Striegl RG, et al. (2007) Plumbing the global carbon cycle: integrating inland waters into the terrestrial carbon budget. Ecosystems 10(1), 172–185. doi:10.1007/s10021-006-9013-8. EPA (Environmental Protection Agency) (1977) Interagency 316(a) Technical Guidance Manual and Guide for Thermal Effects of Nuclear Facilities Environmental Impact Statements. Office of Water, Permits Division, Industrial Permits Branch, Washington DC.

18: Indicators and surrogates in environmental management

Hendriksen RS, Price LB, Schupp JM, Gillece JD, Kaas RS, Engelthaler DM, et al. (2011) Population genetics of Vibrio cholerae from Nepal in 2010: evidence on the origin of the Haitian outbreak. mBio. 2(4), 1–6. doi:10.1128/mBio.00157-11. Hilsenhoff WL (1988) Rapid field assessment of organic pollution with a family-level biotic index. Journal of the North American Benthological Society 7(1), 65–68. doi:10.2307/1467832. Horowitz HM, Jacob DJ, Amos HM, Streets DG, Sunderland EM (2014) Historical mercury releases from commercial products: global environmental implications. Environmental Science & Technology 48(17), 10242–10250. doi:10.1021/es501337j. Houlton BZ, Driscoll CT, Fahey TJ, Likens GE, Groffman PM, Bernhardt ES, et al. (2003) Nitrogen dynamics in ice storm-damaged forest ecosystems: implications for nitrogen limitation theory. Ecosystems 6(5), 431–443. doi:10.1007/s10021-002-0198-1. Hynes HBN (1960) The Biology of Polluted Waters. Liverpool University Press, Liverpool, UK. Johansson O, Palmqvist K, Olofsson J (2012) Nitrogen deposition drives lichen community changes through differential species responses. Global Change Biology 18, 2626–2635. Karr JR (1981) Assessment of biotic integrity using fish communities. Fisheries (Bethesda, Md.) 6(6), 21–27. doi:10.1577/1548-8446(1981)0062.0.CO;2. Kerans BL, Karr JR (1994) A benthic index of biotic integrity (B-IBI) for rivers of the Tennessee Valley. Ecological Applications 4(4), 768–785. doi:10.2307/1942007. Kramer KJM, Jenner HA, de Zwart D (1989) The valve movement response of mussels: a tool in biological monitoring. Hydrobiologia 188–189, 433–443. doi:10.1007/BF00027811. Laws EA (1993) Aquatic Pollution. 2nd edn. John Wiley and Sons, New York. Lindenmayer DB, Likens GE (2011) Direct measurement versus surrogate indicators for evaluating environmental change and biodiversity loss. Ecosystems 14, 47–59. Mason CF (2002) Biology of Freshwater Pollution. 4th edn. Prentice Hall, New York. McDowell WH (1986) Power plant operation on the Hudson River. In The Hudson River Ecosystem. (Eds KE Limburg, MA Moran and WH McDowell) pp. 40–82. Springer, New York. McDowell WH (2015) NEON and STREON: opportunities and challenges for the aquatic sciences. Freshwater Science 34, 386–391. McDowell WH, Brereton RL, Scatena FN, Shanley JB, Brokaw NV, Lugo AE (2013) Interactions between lithology and biology drive the long-term response of stream chemistry to major hurricanes in a tropical landscape. Biogeochemistry 116, 175–186. doi:10.1007/s10533-013-9916-3. Patrick R (1949) A proposed biological measure of stream conditions based on a survey of Conestoga Basin, Lancaster County, Pennsylvania. Proceedings of the Academy of Natural Sciences of Philadelphia 101, 277–341. Pellerin BA, Saraceno JF, Shanley JB, Sebestyen SD, Aiken GR, Wollheim WM, et al(2011) Taking the pulse of snowmelt: in situ sensors reveal seasonal, event and diurnal patterns of nitrate and dissolved organic matter variability in an upland forest stream. Biogeochemistry 108(1–3), 183–198. Pinho P, Dias T, Cruz C, Sim Tang Y, Sutton MA, Martins-Loução M-A, et al. (2011) Using lichen functional diversity to assess the effects of atmospheric ammonia in Mediterranean woodlands. Journal of Applied Ecology 48(5), 1107–1116. doi:10.1111/j.1365-2664.2011.02033.x. Shibata H, Hasegawa Y, Watanabe T, Fukuzawa K (2013) Impact of snowpack decrease on net nitrogen mineralization and nitrification in forest soil of northern Japan. Biogeochemistry 116, 69–82. doi:10.1007/s10533-013-9882-9.

187

188

Indicators and Surrogates of Biodiversity and Environmental Change

Shibata H, Branquinho C, McDowell WH, Mitchell MJ, Monteith DT, Tang J, et al. (2015) Consequence of altered nitrogen cycles in the coupled human and ecological system under changing climate: the need for long-term and site-based research. Ambio 44, 178–193. doi:10.1007/s13280–014-0545–4. Swank WT, Waide JB, Crossley DA Jr, Todd RL (1981) Insect defoliation enhances nitrate export from forest ecosystems. Oecologia 51, 297–299. doi:10.1007/BF00540897. Tierney GL, Fahey TJ, Groffman PM, Hardy JP, Fitzhugh RD, Driscoll CT (2001) Soil freezing alters fine root dynamics in a northern hardwood forest. Biogeochemistry 56, 175–190. doi:10.1023/A:1013072519889. Turner W (2014) Sensing biodiversity. Science 346, 301–302. doi:10.1126/science.1256014. US Fish and Wildlife Service (2014) Red Cockaded Woodpecker Recovery. US Fish and Wildlife Service, Arlington, VA. Utz RM, Fitzgerald MR, Goodman KJ, Parker SM, Powell H, Roehm CL (2013) The National Ecological Observatory Network: an observatory poised to expand spatiotemporal scales of inquiry in aquatic and fisheries science. Fisheries (Bethesda, Md.) 38(1), 26–35. doi:10.1080/03632415.2013.748551. Wetzel RG (2001) Limnology. 3rd edn. Academic Press, New York. Zoellick B, Nelson SJ, Schauffler M (2012) Participatory science and education: bringing both views into focus. Trends in Ecology & Evolution 10(6), 310–313.

19

A diversity of approaches to ecological surrogates and key knowledge gaps David Lindenmayer, Jennifer Pierson, Philip Barton, Peter Lane, Ayesha Tulloch and Martin Westgate

Introduction The body of published work on ecological surrogates and indicators is enormous (Caro 2010; Westgate et al. 2014) and it is almost impossible for any one researcher or manager to be familiar with even a relatively small subset of it. Rather than add unconstructively to an already crowded topic, our overarching aim was to contribute new perspectives to the literature. We have sought to do this in two ways. First, by asking chapter authors to define key learnings and knowledge gaps. Second, we used an inter-disciplinary approach that sought the collected wisdom and new insights of experts, each of whom is working on surrogates, but in very different fields and often in quite different ways. Indeed, the diversity of material in the preceding 18 chapters clearly illustrates the breadth and depth of approaches and perspectives in work on ecological surrogates, as well as research and application of surrogates in other disciplines such as medicine and statistical science. In this final chapter, we distil some of the key insights from both the written work of the chapter authors and the detailed discussions held at a workshop in Queensland (Australia) in October 2014. The workshop discussions entailed considerable extensions of ideas that appear, albeit in very brief and summarised form, in the respective chapters. Therefore, this chapter aims to: (1) outline some of the key areas of overlap and divergence between fields in the use and application of ecological surrogates; and (2) identify important knowledge and research gaps that need to be filled. Perhaps one of the most striking features of this book is the diversity of perspectives and approaches towards the way surrogates are developed and applied in different disciplines. Accompanying these different approaches is the large array of terms and concepts used to describe surrogates and their attributes. Indeed, many of these terms are used in different ways to mean the same or sometimes different things. Even the word indicator will sometimes be used to mean a proxy for something else, but at other times simply to mean the entity that is being measured directly (without any suggestion that it is a proxy for something else) (Chapter 3). Some organisations, such as the US National Park Service, have already developed well-defined meanings for surrogates and indicators, and these are 189

190

Indicators and Surrogates of Biodiversity and Environmental Change

unlikely to change terminology given the amount of effort they have invested in creating these agreed set of terms (Chapter 17). We therefore suggest that the most tractable solution is to simply recognise there is an array of different definitions and usage of different terms. In this book, we have therefore given the freedom to each chapter author to use the terms as they might be applied within their own disciplinary area. An ambitious goal for future attention is to define a set of terms that transcends distinct disciplines yet are agreed upon by all surrogate researchers, as well as managers and policy makers. This would greatly facilitate communication and collaboration among researchers and users of surrogates in different disciplines.

Similarities and differences in the use and application of ecological surrogates The chapters in this book suggest there are many similarities, but also some large differences, in the development and application of ecological surrogates among disciplines within the broader environmental sciences. This was further underscored by the discussions among chapter authors. The use of surrogates is more developed in some fields in terms of accuracy and rigour, including atmospheric pollution and aquatic ecosystems, as well as areas outside of the environmental sciences, such as clinical medicine and pharmacology (Chapter 16). As an example, the use of ecological surrogates in the field of atmospheric monitoring is underpinned by an agreed set of methods and protocols, as well as systematic ways to analyse and interpret data – a markedly different approach to that in many other fields (Chapters 8 and 9). To some extent, the differences between fields reflects the seriousness of direct impacts of environmental conditions on human health and wellbeing – poor air and water quality can lead to deaths. Therefore, the amount of knowledge (and hence scientific and management progress) about ecological surrogates is strongly biased towards fields that directly affect humans, in part because there is strong pressure to ensure that the ‘right’ ecological surrogates are identified and applied in an accurate and scientifically defensible way. Conversely, ecological surrogates in domains that have less direct effects on humans, such as the intactness of marine or terrestrial species assemblages, are less advanced than those used as proxies in assessing airshed and watershed quality. Despite the differences in how ecological surrogates are used between disciplines, there also are some important common themes among them. These include: • The need to identify well-developed objectives and goals for the use of ecological surrogates. It is not possible to determine the effectiveness of an ecological surrogate without a goal against which to judge its efficacy. This includes defining what constitutes management success. This common theme among disciplines may seem trite, but the literature on ecological surrogates and indicators in all fields is replete with innumerable examples where the goal or objective of surrogate application has not been articulated. • The need to develop a robust conceptual model of the ecosystem in question, which can then guide the identification of appropriate surrogates. This is required to identify causal relationships between an ecological surrogate and the entities for which it is a proposed proxy. That is, an effective ecological surrogate will have a clear mechanistic relationship with the target. Such conceptual models must be based on a good scientific knowledge, including some of the key ecological processes and the potential interactions between processes, as well as interactions among species.

19: A diversity of approaches to ecological surrogates and key knowledge gaps

• The need for rigorous statistical assessment and testing of ecological surrogates. Evaluation is a critical part in the development of surrogates but is rarely done. Rigorous evaluation is important because all disciplines have examples of surrogates that have failed (e.g. Millington and Walker 1983; Saraux et al. 2011). A key issue in all disciplines is the extent of transferability of a given ecological surrogate between ecosystems and over time (both within the same system and in different systems). Thus, a part of the evaluation process in all disciplines should include determining the spatial and temporal boundaries of surrogate–target relationships. For example, before a surrogate being applied in a new region beyond the one in which it was first developed, re-evaluation of the relationship in the new region needs to be completed. Similarly, the surrogate relationship may change over time and a new surrogate may become a better choice due to either environmental changes or technological advances (Chapter 8). • Assessment and testing of ecological surrogates also should include cost-effectiveness analyses, particularly as key areas like environmental monitoring and management in which surrogates are widely used can be both expensive and time-consuming. • The need to overcome widespread problems of translating the body of science on ecological surrogates into a form that promotes better use of surrogates in management and decision making. A way to solve this problem in all fields is to engage with key stakeholders at the outset of any project where ecological surrogates are planned to be used (Chapter 11). Failure to do this can result in a surrogate being rejected by groups of stakeholders, irrespective of its scientific validity.

Knowledge gaps and future research priorities Despite the enormous body of research on ecological surrogates, many key knowledge gaps remain. Below we outline a subset of what we consider to be the most important gaps, rather than a comprehensive list of existing deficiencies. 1 Undertake longer term work on ecological surrogates. Much of the scientific work on ecological surrogates is short-term or snap-shot studies of only a year or two (Chapter 2). Longer term, multi-year perspectives of surrogates are critically important for several reasons, especially as we want to determine the effectiveness of surrogates over time. Moreover, we should expect surrogacy relationships to exhibit temporal dynamics and hence be alert to the notion that ecological surrogates may need to be changed and replaced by other ecological surrogates over time. However, careful calibration may be needed during the transition from one surrogate to another, especially if there is an attempt to maintain the time series integrity in a long-term monitoring dataset. 2 Broaden the development of ecological surrogates to include social and economic perspectives. Many ecological surrogates are largely focused on ecological issues – with the social and economic dimensions of surrogate application to environmental problems largely ignored. A clear exception to this, in the context of this book, is the Ocean Health Index that was discussed in detail in Chapter 14. We suggest that the application of ecological surrogates in a decision theory framework may well help improve the integration of ecological social and economic dimensions of environmental problem solving. 3 Accommodating multiple interacting threats. A critical issue in the development and application of ecological surrogates is how to design them to be robust in the

191

192

Indicators and Surrogates of Biodiversity and Environmental Change

4

5

6

7

face of the multiple, and often strongly interacting, factors that influence species, assemblages and ecosystem dynamics (Chapter 6). This remains a critical and currently unresolved challenge for the future, but one that is increasingly important to consider given the extent of global change, the development of novel ecosystems, and the development of novel interacting threats in many environments. Increase the use of theory to underpin ecological surrogate use. Several authors have identified a lack of theory underpinning the development and application of ecological surrogates (e.g. Chapter 2). We suggest that this problem needs to be urgently tackled as it is particularly pertinent to other issues such as better quantifying the spatial and temporal boundaries of ecological surrogates. That is, better use of ecological theory may help better suggest where and under what general circumstances a surrogacy relationship might hold, and when they might not. Improve the predictive capacity of ecological surrogates. Few ecological surrogates can accurately predict changes in ecosystems. Some approaches are theoretically interesting but lack practical application (e.g. the fluctuating ecosystem state models of Carpenter et al. 2010). Similarly, the response of previously widely employed ecological surrogates to new threats remains poorly understood. As an example, it is not known whether the widely documented and widely employed use of lichens as indicators for monitoring airshed quality (and especially in response to pollutants such as sulphur dioxide) will remain robust as surrogates in a rapidly changing climate (Chapter 9). Embrace the potential of new surrogacy approaches. Important new frontiers are opening up in the development and application of surrogates, including those associated with the genetics metrics and those linked with the application of functional traits specifically as ecological surrogates. Therefore, taking ecological surrogates beyond the typical domains of species, species richness and community composition to examine additional forms of biodiversity offers some exciting prospects. For instance, Chapter 15 highlights the many potential values of using genetic metrics as surrogates and how they can provide information well beyond that derived from direct measures. We note that the field of functional diversity as a direct measure is not new (see, for example, Tilman 2001) and has been widely applied in ecology for many years. However, the application of functional traits as ecological surrogates is rare. It is nevertheless appealing because such kinds of proxies might better transcend spatial and temporal boundaries (and therefore prove to be more universal surrogates) than traditional measures such as species occurrence or species richness may change more rapidly in space and over time. However, the potential values of functional traits and genetic metrics as ecological surrogates are still working hypotheses that need rigorous testing. Indeed, the entire field of ecological surrogates may well be strengthened considerably if much of the underlying scientific work is restructured within a hypothesis-testing framework. Promote greater cross-disciplinary learning. The chapters in this book (and the insights from the associated workshop) clearly highlighted that some disciplines are far more advanced in their use of ecological surrogates than others. Yet, there is remarkably limited cross-learning between disciplines. For example, Chapter 16 suggests the fields of statistical science and medical science have much to offer the development of ecological surrogates and indicators. Similarly, approaches to the use of surrogates in monitoring and managing airsheds offers interesting perspectives in terms of standardised methods of data collection and analysis. New ways of

19: A diversity of approaches to ecological surrogates and key knowledge gaps

accelerated cross-learning between disciplines are needed. This has the potential to not only improve the use of ecological surrogates, but also to speed progress and ensure that mistakes made in a given field are not repeated in different fields.

Conclusion Although ecological surrogates have been heavily criticised by several authors (Landres et al. 1988; Niemi et al. 1997; Simberloff 1998; Carignan and Villard 2002; Seddon and Leech 2008; Collen and Nicholson 2014), they are here to stay. There are insufficient resources and time to measure all entities in all ecosystems at all times: the use of proxies in many (although certainly not all) cases is unavoidable. Given this, we need ways to ensure that existing surrogates can be improved, or replaced by better ones where they exist. Indeed, there is an increasing number of examples of the effective use of ecological surrogates. We have identified some of the similarities and differences among disciplines, and described important knowledge gaps. It is now critical to build a cross-disciplinary approach to surrogacy that integrates these learnings and guides some of the science that will be instrumental to developing and applying ecological surrogates in future. The reality is that there is now a large community of scientists, managers and policy makers working on ecological surrogates, but in very different fields. These workers have more in common than probably previously realised – in part because they are working in different ‘silos’ that rarely interact. Our sincere hope is that different perspectives on surrogates outlined in the different chapters in this book have led to new insights that can help improve the identification and application of ecological surrogates in the future.

References Carignan V, Villard MA (2002) Selecting indicator species to monitor ecological integrity: a review. Environmental Monitoring and Assessment 78, 45–61. doi:10.1023/A:1016136723584. Caro T (2010) Conservation by Proxy. Indicator, Umbrella, Keystone, Flagship, and Other Surrogate Species. Island Press, Washington DC. Carpenter SR, Brock WA (2011) Early warnings of unknown nonlinear shifts: a nonparametric approach. Ecology 92, 2196–2201. Collen B, Nicholson E (2014) Taking the measure of change. Science 346, 166–167. doi:10.1126/ science.1255772. Landres PB, Verner J, Thomas JW (1988) Ecological uses of vertebrate indicator species: a critique. Conservation Biology 2, 316–328. doi:10.1111/j.1523-1739.1988.tb00195.x. Millington PJ, Walker KF (1983) Australian freshwater mussel Velesunio ambiguus (Phillipi) as a biological indicator for zinc, iron and manganese. Australian Journal of Marine and Freshwater Research 34, 873–892. doi:10.1071/MF9830873. Niemi GJ, Hanowski JM, Lima AR, Nicholls T, Weiland N (1997) A critical analysis on the use of indicator species in management. The Journal of Wildlife Management 61, 1240–1252. doi:10.2307/3802123. Saraux C, Le Bohec C, Durant JM, Viblanc VA, Gauthier-Clerc M, Beaune D, et al. (2011) Reliability of flipper-banded penguins as indicators of climate change. Nature 469, 203–206. doi:10.1038/nature09630. Seddon PJ, Leech T (2008) Conservation short cut, or long and winding road? A critique of umbrella species criteria. Oryx 42, 240–245.

193

194

Indicators and Surrogates of Biodiversity and Environmental Change

Simberloff D (1998) Flagships, umbrellas, and keystones: is single-species management passe in the landscape era? Biological Conservation 83, 247–257. doi:10.1016/S00063207(97)00081-5. Tilman D (2001) Functional diversity. In Encyclopedia of Biodiversity. (Ed. S.A Levin) pp. 109–120. Academic Press, San Diego, CA. Westgate MJ, Barton PS, Lane PW, Lindenmayer DB (2014) Global meta-analysis reveals low consistency of biodiversity congruence relationships. Nature Communications 5, 3899. doi:10.1038/ncomms4899.

Index

abalone 114 abiotic surrogates  103, 125–35 absent species  97 acidic waterways  104 acoustic sensing  126, 129 active biomonitoring  10–11, 70, 75 adaptive management  53, 119, 154 African Elephant (Loxodonta africana) 28 Aichi target indicators  46, 52, 140 airborne sensors  129 air pollutants  190 monitoring of  166–7 mosses and  69–76 air quality  71, 79, 183 airsheds 192 air temperature increase  78 algal cover  85, 114, 116 allelic richness  153, 155, 156 alpine ecosystems  78 aluminium, soils and  86 ammonia  82, 87 lichens and  83–4, 183 soils and  86 analysis costs  131 animal traits  7–8, 185 anthropogenic stressors  91, 94, 95, 96, 98, 108, 109, 142 anti-surrogates  18, 39–40 ants  60, 61, 62, 63, 64, 65 apex predators  26 aquatic ecosystems  184, 190 monitoring of  181, 185 mosses 70 oxygen levels  183 Arabian Butterflyfish (Chaetodon melapterus) 115 arboreal habitats  61, 63 arctiid moths  61 arrhythmia 162 artificial habitats  39 Ashmore Reef  117 Asian carp  184 assemble first, predict later method  127–8

assessment  29, 105, 109 assumptions, surrogate species approach  173 Atlantic Herring (Clupea harengus) 154 atmospheric change, lichens and  77–90 Australia  49, 50, 52, 62, 107, 117, 151, 153 autocorrection analysis  10 avian surrogates  33–44 bacteria  95, 97, 106 Bald Eagle (Haliaeetus leucocephalus) 182 Bandicoot (Isoodon obesulus)  47, 48, 53 bark beetles  37 bathymetrics  127, 129 bauxite mines  63 Bayesian techniques  53, 132, 144 benthic cover  114, 121, 129, 131 see also seafloor habitats benthic diatoms  97, 102, 108 benthic invertebrates  97, 106, 118, 127, 181–2 best practice guides  175 bioaccumulators  70–1, 78 bioassessment  96, 97, 102 biodiversity  60, 143 decision theory and  45–57 eco-evolutionary dynamics and  155 genetics and  150–6 indicators  26, 52, 70, 91–100, 109, 162 loss of  34, 36, 39, 46, 140, 143, 153 marine systems  138 monitoring  7, 8, 11, 167 pollutants and  82–3 variables 86 biodiversity surrogates  5–13, 15–24, 36 bioeconomic metrics  141, 142 biogeochemical indicators  183–4 biogeographic processes  103 biological communities  170 biological diversity  60 biological traits  93, 95, 106, 173 biomass data  35, 115, 116, 117 biomes  26, 173, 175 biomonitoring  69, 70, 71, 162, 182, 183, 185 biotic integrity  94, 102, 180, 181 195

196

Indicators and Surrogates of Biodiversity and Environmental Change

biotic interactions  103, 174 birds  18, 27, 34, 39, 46, 50 birthweights, lichen monitoring and  79 bivalves 185 boosted regression trees  126 bottom up method, marine environments  127, 128 Brushtail Possum (Trichosurus vulpecula)  47 bryophytes  70, 78 bugs (Hemiptera)  60 build it and they will come paradigm  63 bulk precipitation  73 Bumphead Parrotfish (Bolbometopon muricatum)  115, 116, 121 butterflies  28, 62 Canada  38, 46, 151 Capercaillie Grouse (Tetrao urogallus) 17 carbon dioxide  1–2, 79, 107 carbon storage, oceans  138 cardiovascular medicine  162 causal frameworks  18–19, 162, 163, 164, 166–7, 180 cavity-nesting species  37, 40, 182 certainty level  49, 54 charismatic organisms  18, 181, 182, 183 Chile  35, 36 cholera 184 chrysomelid beetles  61 citizen science  38 clams clearing rate  185 classic umbrella species  26 classification, marine environments  127, 128 climate change  17, 115, 130, 152, 156, 172 coral reefs and  116 flagship species  50 genetic diversity and  154 lichen diversity and  79, 84, 85 marine protected areas (MPAs)  115 monitoring and  171, 172 nitrates and  184 nitrogen and  83 politics and  81 pollution and  86–7 surrogate species and  174, 175 climatic zones  71 clinical medicine  162, 163, 190 see also medical science clinical trials  19 cluster analysis  154 coarse-filter surrogates  172–3 coastal alpine habitat  38 coastal protection  106, 138, 144

commercial fishing  115 commercial logging  182 Common Starling (Sturnus vulgaris) 40 communication tools  16, 20 community-level assessments  115, 116, 119, 127, 128 threats-specific  120, 121 complementary indicators and surrogates  10, 26, 47–9, 53, 139 composite causal model  167 conceptual frameworks  140, 166, 190 congruent distribution patterns  6 coniferous forest  34, 39, 71 consensus priorities  144 conservation campaigns 28 data paucity and  131–2 forests 33–4 genetics and  154–5 goals 171 management 150–6 planning  17, 46, 127 tourism 51 values 50 Conservation by Proxy (Caro)  25 contaminants  167, 179, 181 controls 65 Convention on Biological Diversity (CBD)  125, 140, 150 copper mosses  71 coral reef ecosystems  113–23, 132 coral trout (Plectropomus spp.)  121 correlation analysis  21, 48, 165, 180 cost effectiveness  20, 49 biodiversity surveys  34, 126 indicator choices  46, 54, 65 monitoring and  72, 114, 191 multi-group assessments and  97 spatial scales and  131 costs management 53 moss bags  74 surrogacy studies  8, 9 trade-offs 49–51 see also financial constraints coupled systems  145 covariate distance sampling  37, 165 criteria  51, 152–3 cross-disciplinary research  192–3 cross-taxon congruence  6, 9, 28 cryptic diversity, loss of  153 cultural values  102 cuticle  71, 78 cyanobacterium  78, 85

Index

data analysis  144, 175 data gaps  105, 131, 141, 142 DDT 182 deciduous forest  34, 39 decision theory  45–57, 119, 144, 191 degradation  28, 102 demographic verifiers  153 deposition rates  73 desert ecosystems  78 detectability  36, 37 detoxification mechanisms, mosses  72 detritivores 63 diagnostic tools  139, 162 diatoms  94, 181 digital elevation models  172 Dingo (Canis lupus dingo) 27 direct gradients metrics  128, 129 direct measurement approach  8, 17–18, 21, 38–9 direct species sampling, costs  131 directional climate change  174 dispersal traits  92, 97, 103–4 distance to land mass  129 distribution, marine environments  78, 128 disturbance recovery  65, 121 diversity indices  78, 92, 93, 95, 96, 114, 120 see also biodiversity DNA  65, 151, 184–5 see also environmental DNA (eDNA) dragonflies (Odonata)  28, 182, 183 driver–pressure–state–impact–response (DPSIR) 140 dry depositions, mosses and  71–2 Dugong (Dugong dugon) 121 dung beetles (Coleoptera: Scarabaeidae)  60 dynamic relationships  36–7 Earth Summit (1992)  150 earthworms  28, 60, 62 echo sounders  129 eco-evolutionary dynamics  154–5 ecological aims  64 ecological disturbance  27 ecological factors  152 ecological indicators and surrogates  1–2, 16, 17, 92, 93–4, 103, 107, 109, 167, 170, 171–2, 189 abiotic surrogates  125–35 air pollutants and  69–76 avian species  33–44 bioaccumulators 70–1 biodiversity  5–13, 15–24, 36, 91–100, 109 complementary  10, 26, 47–9, 53, 139 coral reefs  113–23

decision theory and  45–57 environmental management  179–88 freshwater ecosystems  91–100 genetic diversity and  149–59 habitats  39, 128, 130, 173 indices of biological integrity (IBIs)  102–9 invertebrates 59–68 knowledge gaps  189–94 lichens as  77–90 marine systems  120, 125–35, 137–48, 190 medicine and  161–8 mosses 69–76 natural resources management and  169–78 Ocean Health Index (OHI) and  138–44 proxy conservation  25–32 theory 139 trade-offs 49 wetlands integrity  101–12 worldwide 84 ecological information, social goals and  145 ecological specialisation  34 ecological theory  35, 93, 192 ecosystem-based management (EBM)  145 ecosystems 5 change variations in  78 conservation of  150 coral reefs  113–23 decline risk of  107, 119 functions of  106–7, 109, 171 health of  114 invertebrate indicators  59–68 marine 137–8 restoration of  102, 167 stability of  131 surrogates and  1–2, 171–2 ectohydric mosses  70, 71 element uptake efficiency, mosses and  71 employment levels, marine environments and 141 endangered species  51, 184–5 Endangered Species Act 1973 (US)  151, 170, 181, 182 endohydric mosses  70–1 endpoints  162, 163 environmental change  36, 51, 65, 106, 153 environmental DNA (eDNA)  8, 108, 120, 184 environmental gradients  94, 95, 96, 170 environmental indicators  16, 46, 48, 93–4, 98, 128 environmental management  179–88

197

198

Indicators and Surrogates of Biodiversity and Environmental Change

Environmental Performance Index (EPI)  140 Environmental Protection Agency (EPA) (US) 180 Environmental Protection and Biodiversity Conservation Act 1999 151 environmental sensors  185 environmental stress  118 Ephemeroptera–Plecoptera–Trichoptera framework  181, 185 epiphytes  77–8, 81–2 Escherichia coli (E. coli)  1, 2, 180, 184 eukaryotic biodiversity  60 Europe  35, 50, 69, 70, 72, 73, 79, 81, 82, 85, 87, 97, 105 European standard method  81–2 European Starling (Sturnus vulgaris) see Common Starling eutrophication  85, 101 evaluation  46, 108, 163, 166, 191 evidence-based management  46 evidence-based medicine  162 evolutionary processes  9, 103, 151–5, 152 expertise, monitoring programs and  171 extinct species  47, 132, 143 faecal contamination  180, 184, 186 fairy shrimp  107 fauna restoration  63, 104 fecundity  118, 120 Federal Water Pollution Control Act Amendments 1972 (FWPCA) 180 fee-paying visitors  29 feral cat (Felis catus)  48, 53 ferns 78 field implementation  170 filters  104, 172–3 financial constraints  16, 47, 104, 105, 109, 114, 116 fire regime  48, 53, 63 fish  95, 97, 102, 105, 106, 114, 116, 118, 119, 120, 121, 130, 170 fish abundance  114, 115, 116, 117, 118 fish communities  181 fish-eating birds  175 fisheries  116, 138, 170, 181, 184 coral reefs  120 impact of  139 quantitative goals  141–2 sustainability  114, 119, 121, 142 fishing  115, 116, 120, 130, 131, 132, 144, 181 flagship species  27–8, 46, 50, 51, 62, 171 Florida  102, 108 flowering plants  62, 63 fly fishing  182

focal species approach  16, 18, 115, 128 food-web approach  48, 52–3, 65, 106, 138, 139 forests  17, 21, 33, 35, 46, 108, 126 management of  39, 79 freshwater ecosystems  91–100 fruticose lichens  81, 85 functional diversity  9, 35–6, 53, 85, 116 functional groups  106–7, 109, 115, 118 functional traits  7–8, 64–5, 65, 85, 192 fungi  78, 106 future conditions  176, 179 generalisation, validation and  163 generation times  60 genetic bar-coding  8 genetic diversity  149–59, 192 genetic drift  103 genomics  140, 151, 154 geodiversity 109 geographic limitations  173 geological data  128, 129 geostatistical analysis  87 giant squid  28 gliders 39 global change lichen and  80, 87 multiple threats  192 global food security  142 global warming, public opinion and  29, 30 Global Wild Bird Index  46, 50 goal setting  190 indices of biological integrity and  105, 107 local communities  142 Ocean Health Index and  141 Golden Toad (Incilius periglenes) 29 grazing  62, 120 Great Lakes  175, 184 green algae  78, 85 ground-dwelling invertebrates  61, 63 growth rates  118, 120, 121 guilds  34, 35, 37, 38 habitat links  63 habitat maps  126, 132 habitats  102, 106, 138, 172, 179 loss of  8, 29, 34, 35, 131, 151, 156, 182 recovery goals  142 habitat surrogates  39, 128, 130, 173 hatchery fish  153 heavy metals  70, 72, 84 herbivores  38, 62, 63, 65, 118, 119, 120 heterozygosity  155, 156 high-spatial-resolution mapping  82–3

Index

historical variation range  174 holistic approach, indices of biological integrity 109 hollow-bearing trees  37 honeybees 28 housing developments  101 Hudson River  181, 182 human activities, monitoring  77 human health research, lichens and  79, 83 human land use  174 human-made disturbance, marine environments 131 hurricanes 184 hydroperiods 107 Hylocomium splendens  71, 72 hypothesis discrimination  46 iconic species  138, 141, 181 impact study areas, marine environments  115, 131 inbreeding depression  154 Index of Atmospheric Purity (IAP)  80–1 Index of Biotic Integrity  181 Index of Plant Community Integrity (IPCI) 108 index values  93, 98 India  27, 28 indicator species  15, 16, 97, 114, 105, 181 see also ecological indicators and surrogates indices of biological integrity (IBIs)  102–9 indirect management, coral reefs  114 individual species modelling  128, 129, 163, 164 industrial regions  72 insectivores 38 insect outbreaks  184 integrated interactions, lichen  80 integrative syntheses, need for  52, 145, 175 interdependencies  47, 53 interdisciplinary method  189 international collaboration  107 international database  85, 86, 87 International Union for Conservation of Nature (IUCN) Red List  46–7, 48, 50, 143 interspecies variations, mosses  71 intervention factors  166 invasive species  39, 53, 62, 114, 130, 156, 184, 185 invertebrate indicators  59–68, 104, 107 investment optimum  49 irreplaceable species  35, 172 isotope studies  65 Italy  79, 83, 85

Keeling curve  2 keystone species  9, 26, 179 knowledge gaps  189–94 Koala (Phascolarctos cinereus) 62 Kyoto protocol  81 lakes  97, 183 Landsat imagery  39 landscapes  170, 172 biodiversity of  104, 107 changes in  38 land-use changes  27, 28, 46, 83, 107 larvae 181 lead pollution  69 leaf breakdown  107 least concern category  143 legacy effect  84 legal mandates  180–1 lichens  77–90, 186, 192 atmospheric change and  166, 167 pollutants and  183 light detection and ranging (LIDAR)  129 linear models  126, 143 litter bags  106 lobsters 114 local biotic interactions, dispersal and  104 local communities  104, 108, 142 Longleaf Pine (Pinus palustris) 182 long-term monitoring  52, 72, 116–17 long-term studies  8–9, 65, 191 long-term system dynamics  174 long-winged invertebrates  64 lung cancer  79, 83 macroinvertebrates  94, 97, 102, 106, 108 macro-lichens 81 macrophytes 97 macro scales  6, 7 Maine  106, 108 management 52–3 goals and objectives  8, 144, 172, 175, 190–1 guidelines 119 objectives 50 outcomes  49, 53 strategies  46–7, 131 management indicator species  26, 29, 50 management surrogates  107, 109 management zones, coral reefs  117 mangroves  115, 118, 143 mapping, pollutants  82–3 mariculture  138, 142, 144 marine protected areas (MPAs)  113–14, 115, 121, 125–6, 131, 132

199

200

Indicators and Surrogates of Biodiversity and Environmental Change

marine systems  26, 120, 125–35, 137–48, 190 marketing tools  29 Markov Decision Processes  53 marsupials 37 maximum entropy formalism  9–10 maximum sustainable yield (MSY)  142 mean trophic level of fisheries catch (MTLc) 139 measurement criteria  20, 161, 165, 173, 180 mechanisms, surrogacy  35–6 medical science, surrogacy studies and  19, 21, 161–8, 192 see also clinical medicine mercury monitoring  182–3 meta-analysis  52, 164, 167, 174 metal accumulation, mosses and  71 microbes, indicator  186 micro-lichen species  81 microscopy 60 mid-domain effect  129 millipedes 60 mine rehabilitation  62–3, 65, 108 mismanagement indices of biological integrity and  105 wetlands 108 mites 60 mobility, invertebrates  60 modelling  140, 144 molecular approaches  65, 156 monitoring  10, 51 air quality  79 coral reefs  114, 116–17, 119, 121 cost of  191 environmental change  65 genetics 152 lichen diversity  82 limits of  45–6, 118 mosses 72–3 objectives of  154–5, 170, 171 moorland environments  19 morphospecies approach  64, 66 mortality, cardiovascular medicine and  162 moss bag technique  70, 72, 73–4, 75 mosses, air pollutants and  69–76 motivation 104–5 multibeam acoustic data  127 multimetric indices  92, 93, 94, 95 multiple ecological threats  191–2 multiple indicators  46, 64, 96, 97, 98, 102, 139 marine environments  140 pollutants 86–7 wild populations  153 multispecies assemblages  28, 130 multivariate analysis  48, 98 mussels 182

mutual information analysis  48 myxohydric mosses  70 narrative indicators  140, 141 national indicators  46, 150, 185 national parks  29, 30, 171, 189–90 natural disturbance, marine environments 131 natural resources management  169–78 natural variability  108 negative surrogacy  18 nest boxes  39 New Zealand  46, 50, 108 niche theory  34–5, 60 nitrates  183–4, 185 nitrogen  70, 79, 83, 84, 85, 87 non-congruent taxa, coral reefs  118 non-genetic surrogates, limits of  153–4 non-invasive sampling  151 non-linear relationships  130, 143 non-random mating  152 non-surrogate indicators  174 non-transferable indicators  96 North America  35, 39, 70, 185 Northern Flicker (Colaptes auratus) 37 Northern Spotted Owl (Strix occidentalis) 35, 46 novel species  174, 175 novel surrogates  120, 121 objectives 20 monitoring  46–7, 152, 154–5, 170–1 need for  190 statement of  20 objectives-driven indicator choices, coral reefs 116–17 observational studies  165, 166 Ocean Health Index (OHI)  129, 138–9, 140, 191 goals  141, 144 scoring method  141, 143–4 off-the-shelf protocols  171 old-growth logging  46 oligotrophic species  83, 84 open habitats  17, 62 opossums 39 orchids 78 organic pollutants  70, 106, 108, 185 organism groups responses  95 orthogonal indicators, marine environments 139 outcomes  50, 52, 162–3, 164, 165, 166 output maps  132 owls  35, 36 oxygen, dissolved  183

Index

paradigm shifts, indices of biological integrity and 109 passive biomonitoring  70, 72, 75 pathogenic microbes  180 pattern-based surrogates, marine environments  126–7, 132 peat profiles  70 pest management  117 phosphorus, aquatic ecosystems and  183 photobiont component  78, 85 photosynthesis  85, 183 phylogenetic relations  92 physical environment, monitoring  171, 174 physiological tolerances  120 phytoplankton  97, 184 pitfall trap  61 places of special significance  138 plantation forestry  17, 39, 182 plant-dwelling herbivores  63 platform nests  39 poikilohydric species  70, 78, 85 Polar Bear (Ursus maritimus) 29 policy directions, lack of  174–5 pollination  62, 64 pollution  17, 73, 83, 85, 86–7, 101, 107, 190, 192 clams and  185 lichens and  80, 79 sampling 182–3 polycyclic aromatic hydrocarbons (PAH)  70 ponds 107 population density  184 population size, genetics and  151, 152, 153, 155–6, 163 Portuguese millipedes  66 power plants  181, 186 practical surrogacy, academic studies and  19 pre-colonial times benchmark  142 predators  6, 37, 47–8, 63, 103, 104, 120 predict first, assemble later method  127–8 predictive modelling  20, 21, 92, 93, 95, 96, 105, 128, 130, 161–2, 192 primary forest  33 prior information, lack of  52 prioritisation  20, 126 process-based surrogates, coral reefs  117–18 protected areas  29, 30, 46 proxy conservation  25–32 ptarmigans 38 public access bans  143 public health, indicator organisms and  180 public opinion, global warming and  29, 30 public participation  139, 181 qualitative indicators  50–1, 70

quantitative measurement  51, 70, 141–2, 144, 174 quarries 82 Queensland  3, 27, 131, 189 rabbitfish (Siganus spp.) 121 rainfall decrease  83 rainforest 27 Ramsar Convention  107 randomised trials  165 rapid bioassessment  106, 181 Rapoport’s rule  129 raptors  27, 35, 39, 182 rare species  50, 130 rarity-complementarity algorithms  126 reactive management, Marine Protected Areas 119 real-time environmental sensors  185 Red Fox (Vulpes vulpes)  27, 47, 49, 53 redundancies  47–9, 54, 85 reef protection status  131 reference points, Ocean Health Index  141 reforestation 62 refuges 170 regional diversity  95, 104, 129 region-specific context, bioindicators  96–7 regression equations  73 rehydration 70 reliability, detectability and  37 remote sensing  39, 83, 120, 126, 129, 132, 184 Representative Important Species  180–1, 186 research objectives  96, 171, 173 reserves  26, 29, 170 resources gradient metrics  128–9 respiration  106, 183 respiratory disease  83 restoration projects  17, 29, 62, 64, 65, 102, 121 results interpretation  175 Ringtail Possum (Pseudocheirus occidentalis)  47, 48 risk analysis  21, 53–4 river management  105, 180 road dust  73, 83 Rock Ptarmigan (Lagopus muta) 38 salinity 154 Salmonella 180 sampling methods  64, 66, 105, 155, 175 effectiveness variation  130, 131–2 lichen diversity  87 sand dunes  142 satellite images  120, 129, 172 scale-dependent patterns, marine environments 129 seafloor habitats  127, 129, 142

201

202

Indicators and Surrogates of Biodiversity and Environmental Change

see also benthic cover seafood sustainability  141 seascapes  127, 128 seasonal factors marine environments  78, 130 wetlands 103 Sea Urchin (Echinometra mathaei) 115 sediment covers  126, 127 seed-dispensing birds  27 selection-based indicators and surrogates  50, 126–7, 132 sense of place  138, 141 sensor technology  185 sentinel species  29 sequencing 19–21 shelf reefs  131 shifting baseline effect  174 Silver Carp  184 Simpson’s Paradox  164, 165 single-assessment index  143–4, 145 single-attribute indicators  139, 140, 141 single nucleotide polymorphism (SNP) chip technologies 151 single species approach  118, 181 coral reefs  114, 115, 120, 121 site-specific indicators  80, 97, 171, 175 situational analysis, population genetics and 155–6 skinks 167 snails 60 snapshot studies  8, 191 social acceptance  114 social marketing  27–8 social priorities  142–3, 144, 145 socio-ecological interactions, marine environments  132, 133, 138, 141 socio-economic factors  191 environment and  121 marine protected areas (MPAs)  115, 117 soil health  8, 60, 62, 77 soil pollutants  84, 86–7 Southern Brown Bandicoot (Isoodon obesulus) see Bandicoot spatial location, bioindicators and  98 spatial models genetic indicators and  151 lichen diversity  80, 82–3, 86–7 spatial processes  102–4, 105, 130, 142 conservation prioritisation  126 population genetics and  155 spatial scale  6, 7, 94, 95–6, 127, 129–30, 141 spatial variation, pollutants  72 species accumulation  37, 62, 130 Species at Risk Act 2002 (SARA) (Canada) 151 species distribution modelling  127, 128, 153 species indicators  46, 85

species interaction  9, 54 species replacement  38 species richness  7, 34, 37, 66, 80, 85, 115, 116, 117, 129, 131, 162, 165, 172, 173, 174, 192 species-specific approach  2, 25–30, 39, 95, 129, 130, 132, 137, 150, 151, 173 species traits  47, 93, 94, 95, 97, 171–2 Sphagnum spp.  72 spiders  60, 63 sponges 114 stakeholders 145 indices of biological integrity and  104, 108 marine protected areas  121 scientific validity and  191 starlings see Common Starling State of the Environment Report (2011)  50, 153 statistical assessment  19, 50, 162, 164, 191, 192 abiotic surrogates and  130 problems of  87 stick nest survey  39 stochastic processes  103 stream ecology  95, 181–4 stressors 120 coral reefs and  114 mosses and  72 Striped Bass (Morone saxatilis)  181, 182 structure, bioassessments and  104, 107 substrate data  71, 126 sulphur dioxide  80, 83, 87, 192 surrogacy variability, meta-analysis and  167 surrogate species  50 biodiversity and  5–13 definition  16, 166 endpoints 162 lack of need for  9 medicine and  161–8 natural resources monitoring and  169–78 see also ecological indicators and surrogates surrogate–target relationship  36–7 survival nodes  53 sustainability  142, 145, 150 Tammar Wallaby (Macropus eugenii)  47, 48, 53 target biodiversity  37 targeted management  172, 173 targeted predictive assessment  161–2 taxonomic groups  9, 35–6, 65–6, 172, 175 Simpson’s Paradox and  165 taxonomic limitations  105, 173 taxon-specific indices  93–4, 95, 96 technical expertise indices of biological integrity and  104

Index

marine monitoring and  132 temperature increase  2, 83 temporal processes  64, 97, 103, 115, 142 terminology development  16 terrestrial ecosystems avian surrogates and  33–44 monitoring 185 testing 192 absence of  18–19 criteria 21–2 therapeutic intervention  162 theridiid spiders  61 thermal stress, coral reefs  120 threatened species  17, 46, 47, 48, 50, 143 threat mitigation  51, 52, 53 threshold values, lichens  87 time dispersal and  104 environmental change over  153, 191 marine environments over  130 top-down method, marine environments  127, 128 tourism  143, 138 trade-offs  50–1, 143 transferability, biodiversity criteria  21, 35 transplantation  73, 75 treatments  162–3, 164 tree cavities  39–40 Tree Swallow (Tachycineta bicolor) 39 tree trunk grid  81, 82 Trembling Aspen (Populus tremuloides) 39 trigger values, coral reef monitoring  119 trophic levels  37, 38, 65, 106, 120 tropical forests  26, 71 turtles 121 typhoid fever  180 umbrella species  26, 27, 62 uncertainty reduction  51–2, 54, 144 United Kingdom (UK)  19, 52 United States of America (USA)  46, 50, 79, 81, 82, 83, 87, 102, 105, 107, 108, 142, 144, 151, 162, 170, 171, 174, 180–1, 182, 184, 185, 186

universal indicator, lack of  94 unmeasured covariates  165 unprotected species  10 unsupervised classification  127, 128 urbanisation  27, 72, 79, 108, 130 user–community consultation  105 validation  18–19, 21–2, 163 value of information (VOI) analysis  51–2 values, rescaling of  143 variables abundance of  179, 180 independence of  164 unmeasured 103 vascular plants  71, 77 vegetation recovery  60, 62, 63, 172 vernal pool ecosystems  106–7, 108 vulnerable species  35, 143, 172 water absorption  71 waterbirds 39 water quality  2, 85, 97, 102, 117, 138 coral reefs  121 water supplies, pollution of  91, 144, 180, 182–3, 184, 185, 186, 190 weighted average indices  140, 143 wellbeing indicators  140 Western Australia (WA)  47, 61, 63, 65 wet deposition data  73 wetlands integrity  101–12 White-tailed Ptarmigan (Lagopus leucura saxatalis) 38 wildlife corridors  28, 30 wild populations  149–59 wind farms  141 Woodland Caribou (Rangifer tarandus) 46 woodpeckers  34–5, 37, 39, 40, 167, 182 Woylie (Bettongia penicillata)  47, 48 Wurz method  180 Yellow Pine (Pinus taeda) 182 zooplankton  106, 184

203