Community

4,813 views

Published on

colombia guerra

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
4,813
On SlideShare
0
From Embeds
0
Number of Embeds
2
Actions
Shares
0
Downloads
55
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Community

  1. 1. Community Ecology
  2. 2. This page intentionally left blank
  3. 3. Community EcologyProcesses, Models,and ApplicationsEDITED BYHerman A. VerhoefVU University, Amsterdam, Department of Ecological Science,the NetherlandsPeter J. MorinRutgers University, Department of Ecology, Evolution & NaturalResources, USA1
  4. 4. 3Great Clarendon Street, Oxford OX2 6DPOxford University Press is a department of the University of Oxford.It furthers the University’s objective of excellence in research, scholarship,and education by publishing worldwide inOxford New YorkAuckland Cape Town Dar es Salaam Hong Kong KarachiKuala Lumpur Madrid Melbourne Mexico City NairobiNew Delhi Shanghai Taipei TorontoWith offices inArgentina Austria Brazil Chile Czech Republic France GreeceGuatemala Hungary Italy Japan Poland Portugal SingaporeSouth Korea Switzerland Thailand Turkey Ukraine VietnamOxford is a registered trade mark of Oxford University Pressin the UK and in certain other countriesPublished in the United Statesby Oxford University Press Inc., New York# Oxford University Press 2010The moral rights of the author have been assertedDatabase right Oxford University Press (maker)First published 2010All rights reserved. No part of this publication may be reproduced,stored in a retrieval system, or transmitted, in any form or by any means,without the prior permission in writing of Oxford University Press,or as expressly permitted by law, or under terms agreed with the appropriatereprographics rights organization. Enquiries concerning reproductionoutside the scope of the above should be sent to the Rights Department,Oxford University Press, at the address aboveYou must not circulate this book in any other binding or coverand you must impose the same condition on any acquirerBritish Library Cataloguing in Publication DataData availableLibrary of Congress Cataloging in Publication DataLibrary of Congress Control Number: 2009934138 ¨Cover illustration by Janine MarienTypeset by SPI Publisher Services, Pondicherry, IndiaPrinted in Great Britainon acid-free paper byCPI Antony Rowe, Chippenham, WiltshireISBN 978–0–19–922897–3 (Hbk.)ISBN 978–0–19–922898–0 (Pbk.)1 3 5 7 9 10 8 6 4 2
  5. 5. ContentsPreface xiList of Contributors xiiiIntroduction 1Part I Shape and Structure 51 The topology of ecological interaction networks: the state of the art 7 Owen L. Petchey, Peter J. Morin and Han Olff 1.1 Introduction 7 1.1.1 What do we mean by the ‘topology’ of ecological networks? 7 1.1.2 Different types of ecological networks 9 1.1.3 Three general questions 11 1.2 Competitive networks 11 1.2.1 Structural regularities 11 1.2.2 Mechanisms 13 1.2.3 Unresolved issues 15 1.3 Mutualistic networks 15 1.3.1 Structural regularities 15 1.3.2 Mechanisms 16 1.3.3 Unresolved issues 16 1.4 Food webs 17 1.4.1 Structural regularities 17 1.4.2 Mechanisms 18 1.4.3 Unresolved issues 21Part II Dynamics 232 Trophic dynamics of communities 25 Herman A. Verhoef and Han Olff 2.1 What types of dynamics can be distinguished? 25 2.1.1 Stable equilibria 25 2.1.2 Alternate equilibria 26 2.1.3 Stable limit cycles 26 2.1.4 Chaotic dynamics 26 2.2 Dynamics of food web modules 26 2.3 Internal dynamics in food web modules or simple webs 29 2.4 Dynamics enforced by external conditions 32
  6. 6. vi CONTENTS 2.5 Equilibrium biomass at different productivities 33 2.6 Dynamics of complex interactions 34 2.7 Conclusions 343 Modelling the dynamics of complex food webs 37 Ulrich Brose and Jennifer A. Dunne 3.1 Introduction 37 3.2 Simple trophic interaction modules and population dynamics 37 3.3 Scaling up keystone effects in complex food webs 39 3.4 Diversity/complexity–stability relationships 40 3.5 Stability of complex food webs: community matrices 40 3.6 Stability of complex food webs: bioenergetic dynamics 41 3.7 Stability of complex food webs: allometric bioenergetic dynamics 42 3.8 Future directions 444 Community assembly dynamics in space 45 Tadashi Fukami 4.1 Introduction 45 4.2 Determinism and historical contingency in community assembly 46 4.3 Community assembly and spatial scale 48 4.3.1 Patch size 48 4.3.2 Patch isolation 49 4.3.3 Scale of environmental heterogeneity 51 4.3.4 Synthesis 52 4.4 Community assembly and species traits 52 4.5 Conclusions and prospects 53Part III Space and Time 555 Increasing spatio-temporal scales: metacommunity ecology 57 Jonathan M. Chase and Janne Bengtsson 5.1 Introduction 57 5.2 The varied theoretical perspectives on metacommunities 58 5.2.1 Neutral 60 5.2.2 Patch dynamics 60 5.2.3 Species sorting 60 5.2.4 Mass effects 60 5.3 Metacommunity theory: resolving MacArthur’s paradox 61 5.4 As easy as a, b, g: the importance of scale 61 5.5 Species–area relationships and metacommunity structure 63 5.6 Effects of dispersal rates on local communities 64 5.7 Local–regional richness relationships 65 5.8 A synthesis of metacommunity models 65 5.9 Adding food web interactions into the equation 66 5.10 Cross-ecosystem boundaries 66 5.11 Conclusions 68
  7. 7. CONTENTS vii6 Spatio-temporal structure in soil communities and ecosystem processes 69 Matty P. Berg 6.1 Introduction 69 6.2 Soil communities, detrital food webs and soil processes 69 6.3 Soil organic matter 71 6.4 Variability in time in soil communities 72 6.5 Variability across horizontal space in soil communities 75 6.6 Variability across vertical space in soil communities is high 77 6.7 Spatio-temporal scales of community studies 79Part IV Applications 817 Applications of community ecology approaches in terrestrial ecosystems: local problems, remote causes 83 Wim H. van der Putten 7.1 Introduction 83 7.1.1 Issues in applied community ecology 83 7.1.2 Top-down and bottom-up go hand in hand 84 7.2 Community interactions across system boundaries 85 7.2.1 Linkages between adjacent or distant ecosystems 85 7.2.2 Linkages between subsystems: aboveground–belowground interactions 85 7.2.3 Consequences for application: find the remote cause of local effects 86 7.3 Community interactions and land use change 87 7.3.1 Land use change, predictability and major drivers of secondary succession 87 7.3.2 Secondary succession from an aboveground–belowground perspective 88 7.3.3 Consequences for restoration and conservation 89 7.4 Biological invasions 90 7.4.1 Community-related hypotheses that explain biological invasions 90 7.4.2 Mount Everest or tip of the iceberg? 91 7.4.3 Conclusions and consequences for management 92 7.5 Discussion, conclusions and perspectives 928 Sea changes: structure and functioning of emerging marine communities 95 J. Emmett Duffy 8.1 Introduction 95 8.1.1 Fishing as a global experiment in community manipulation 96 8.1.2 Physical forcing and the uniqueness of marine ecosystems 96 8.2 The changing shape of marine food webs 97 8.2.1 Conceptual background 97 8.2.2 Empirical evidence for trophic skew in the ocean 99 8.3 Trophic cascades in the sea 100 8.3.1 Conceptual background 100 8.3.2 Evidence for trophic cascades in open marine systems 101 8.3.2.1 Rocky bottoms 102 8.3.2.2 Continental shelves 103 vii
  8. 8. viii CONTENTS 8.3.2.3 Pelagic systems 103 8.4 Biodiversity and stability of marine ecosystems 104 8.4.1 Conceptual background 104 8.4.2 Evidence linking diversity and stability in marine systems 105 8.4.2.1 Comparisons through time 105 8.4.2.2 Comparisons across space 107 8.4.2.3 Mechanisms 108 8.5 Interaction strengths and dynamic stability in marine food webs 109 8.5.1 Conceptual background 109 8.5.2 Empirical evidence 110 8.6 Alternate stable states and regime shifts in marine ecosystems 110 8.6.1 Conceptual background 110 8.6.2 Empirical evidence for regime shifts in marine ecosystems 112 8.6.2.1 Mechanisms 112 8.7 Emerging questions in emerging marine ecosystems 1139 Applied (meta)community ecology: diversity and ecosystem services at the intersection of local and regional processes 115 Janne Bengtsson 9.1 Introduction 115 9.2 A theoretical background 116 9.2.1 A simplified historical narrative 116 9.2.2 Implications of metacommunity theory 118 9.2.3 Metacommunities in human-dominated landscapes: effects of habitat loss and fragmentation 121 9.3 A selection of empirical studies 123 9.3.1 Applied questions allow experimental studies on management scales 123 9.3.2 Biodiversity in human-dominated landscapes: local or landscape management? 124 9.3.3 Local and regional effects on ecosystem services 127 9.3.4 What have we learned in the context of metacommunity ecology? 12910 Community ecology and management of salt marshes 131 Jan P. Bakker, Dries P.J. Kuijper and Julia Stahl 10.1 Introduction 131 10.2 Natural salt marsh: the back-barrier model including a productivity gradient 132 10.3 Effects of plants on herbivores (bottom-up control) 133 10.4 Effects of intermediate-sized herbivores on plants (top-down control) 134 10.4.1 Experimental evidence 134 10.4.2 Effects of herbivores at high marsh 135 10.4.3 Low marsh 136 10.5 Large-scale effects of an intermediate herbivore on salt-marsh vegetation 137 10.6 Interaction of herbivory and competition 137 10.7 Competition and facilitation between herbivores 139 10.7.1 Short-term competition and facilitation between hares and geese 139 10.7.2 Long-term facilitation between herbivores 140 10.8 Exclusion of large herbivores: effects on plants 141 10.8.1 Natural marshes 141
  9. 9. CONTENTS ix 10.8.2 Artificial salt marshes 142 10.9 Exclusion of large herbivores: effects on invertebrates 143 10.10 Exclusion of large herbivores: effects on birds 144 10.10.1 Migrating birds 144 10.10.2 Breeding birds 145 10.11 Ageing of salt marshes and implications for management 146Part V Future Directions 14911 Evolutionary processes in community ecology 151 Jacintha Ellers 11.1 Introduction 151 11.1.1 Bridging the gap between evolutionary biology and community ecology 151 11.2 Evolutionary biology: mechanisms for genetic and phenotypic change 152 11.2.1 Benefits and maintenance of genetic diversity at the population level 152 11.2.2 The source and nature of genetic variation 154 11.2.3 The relationship between genetic and phenotypic diversity 154 11.3 Proof of principle: community properties result from genetic identity and selection at the level of individual organisms 155 11.4 Effects of genetic and phenotypic diversity on community composition and species diversity 156 11.4.1 Effects of genetic diversity on community functioning 156 11.4.2 Diversity begets diversity? 157 11.4.3 Phenotypic diversity is also important for community diversity and composition 158 11.4.4 Phenotypic plasticity and invasive success 159 11.5 Effect of community composition on the genetic and phenotypic diversity of single species 160 11.6 Future directions 16112 Emergence of complex food web structure in community evolution models 163 Nicolas Loeuille and Michel Loreau 12.1 A difficult choice between dynamics and complexity? 163 12.2 Community evolution models: mechanisms, predictions and possible tests 165 12.2.1 One or many traits? 165 12.2.1.1 Models in which species are defined by many traits 166 12.2.1.2 Models with a limited number of traits 166 12.2.2 Evolutionary emergence of body-size structured food webs 168 12.2.3 Advantages of simple community evolution models 170 12.2.3.1 Comparison with other community evolution models 170 12.2.3.2 Comparison with binary qualitative models 170 12.2.3.3 Testing predictions 172 12.3 Community evolution models and community ecology 173 12.3.1 Community evolution models and the diversity–stability debate 173 12.3.2 Effects of perturbations on natural communities 174 12.3.3 Models with identified traits: other possible applications 175 12.4 Conclusions, and possible extensions of community evolution models 176 12.4.1 Possible extensions of community evolution models 177 12.4.2 Empirical and experimental implications of community evolution models 177
  10. 10. x CONTENTS13 Mutualisms and community organization 179 David Kothamasi, E. Toby Kiers and Marcel G.A. van der Heijden 13.1 Introduction 179 13.2 Conflicts, cooperation and evolution of mutualisms 180 13.2.1 Mutualism can also develop without evolution 183 13.3 Mutualisms in community organization 183 13.3.1 Plant–pollinator interactions 183 13.3.2 Plant–protector mutualism 184 13.3.3 Plant nutrition symbiosis 185 13.3.3.1 Legume–rhizobia symbioses 185 13.3.3.2 Mycorrhizal symbioses 188 13.4 Conclusions 19114 Emerging frontiers of community ecology 193 Peter J. Morin 14.1 Introduction 193 14.1.1 Spatial ecology 193 14.1.2 Complex dynamics 193 14.1.3 Size-dependent interactions 193 14.1.4 Interactions between topology and dynamics 193 14.1.5 Evolutionary community dynamics 194 14.1.6 Applied community ecology 195 14.2 Future directions 196 14.2.1 Biotic invasions 196 14.2.2 Interaction networks beyond food webs 199References 203Index 245
  11. 11. PrefaceIn 2001 H.A.V. started a course for second-year turers were accompanied by some of their PhDundergraduate biology students at the Vrije Uni- students, creating an international group of com-versiteit Amsterdam entitled Community Biology. munity ecologists. The course was not intended toThis course has now been running successfully for be encyclopaedic, but rather it focused on the areas8 years. The course was obligatory for all biology of expertise of the invited speakers, many of whichstudents, and it differed from other courses in that share the theme of patterns and processes emergingit was multidisciplinary and provided the students from ecological networks. Participants addressedwith opportunities to perform their own research. the state of the art in theory and applications ofThe multidisciplinarity was emphasized by the dif- community ecology, with special attention to topol-ferent disciplines of the teachers on the course: soil ogy, dynamics, the importance of spatial and tem-ecology, plant ecology, systems ecology, microbial poral scale and the applications of communityphysiology and theoretical biology. The important ecology to emerging problems in human-domi-task of finding a textbook that could link all disci- nated ecosystems, including the restoration andplines and encourage participating lecturers to de- reconstruction of viable communities. The courseliver a unified course was solved by using finished with speculations about future researchCommunity Ecology by P.J.M. That book linked the directions. During the course, it became clear thatdifferent subjects of community ecology, and this international group of students appreciated theintegrated the more theoretical parts on modelling information presented by the various lecturers, de-with the empirical studies, including topics such as spite the fact that research topics exhibited greatbiodiversity and applied studies. Subsequently, diversity. It was during this very stimulating courseH.A.V. and P.J.M. met at an international meeting that the idea for this book took form. H.A.V. andon food webs, and discussed the possibility of par- P.J.M., the editors, convinced most lecturers toticipating in a similarly themed graduate-level transform their lectures into book chapters, andcourse. And, thus, our current collaboration began. asked other colleagues to fill in some gaps. TheIn The Netherlands PhD students from different result captures much of the excitement about com-universities are organized into interdisciplinary munity ecology expressed during the course, andthematic groups, called research schools, that pro- expands the coverage of topics beyond what wevide an intellectual support base for instruction and were able to discuss in an intensive week-longresearch. For example, students working in the field course. We recognize at the outset that certain sub-of socioeconomic and natural sciences of the envi- disciplines of community ecology are not coveredronment belong to the Research School SENSE. In here, and we do not claim otherwise. We know that ´2005, H.A.V., Andre de Roos, Claudius van de Vij- the topics addressed here will be of interest tover and Johan Feenstra organized a PhD course on advanced students and practitioners of communityCommunity Ecology for the SENSE PhD ecology. Ultimately, 19 colleagues participated inprogramme. During this 1 week course held in writing this book. We thank them all for their im-Zeist, leading researchers in the field of Community portant contributions. Writing book chapters,Ecology from Europe and the USA were asked to strangely enough, is less valued than writing arti-deliver lectures on recent and often unpublished cles for scientific journals in some academic circles.developments in their areas of expertise. The lec- Still, like the multidisciplinary course mentioned xi
  12. 12. xii PREFACEabove, we find that the interactive writing that hap- well as the students who have taken the communitypens when people from different subdisciplines ecology course that he has taught at Rutgers Uni-work together is a fascinating, synergistic and pro- versity since 1983. Their collective comments andductive process. feedback have helped to refine his perspectives We would like to thank friends and colleagues about the nature of community ecology over thewho were indispensable during the process of years. Thanks also go to participants in a recentwriting: H.A.V. thanks Nico van Straalen, who by seminar on Ecological Networks for critical feed-writing his book Ecological Genomics for Oxford back on some of the writing that appears here,University Press acted as an instigator for this including Mike Sukhdeo, Maria Stanko, Waynebook. H.A.V. also thanks his colleagues who made Rossiter, Tavis Anderson, Faye Benjamin, Denisethe Community Biology course a success for so Hewitt, Kris Schantz and Chris Zambel. ¨many years: Wilfred Roling, Bob Kooi, Matty Berg, H.A.V. and P.J.M. both thank Ian Sherman ofWilfried Ernst, Tanja Scheublin, Diane Heemsber- Oxford University Press, who was immediately en-gen, Stefan Kools, Marcel van der Heijden, Susanne thusiastic about this book project, and Helen Eaton,de Bruin, Lothar Kuijper, Rully Nugroho and Henk who as assistant commissioning editor played avan Verseveld. H.A.V. acknowledges colleagues crucial role in the development of the book.who were directly involved in the organization of H.A.V. thanks Emilie Verhoef, without whom ´the PhD course: Andre de Roos, Claudius van de this book probably would never have been pro-Vijver, Johan Feenstra and Ad van Dommelen. H.A. duced. P.J.M. thanks Marsha Morin for her under-V. is very grateful to his critical friend, John Ash- standing and support during another extendedcroft (Durham), for supportive focusing. writing project. P.J.M. thanks the many students who partici- Herman A. Verhoef, Amsterdampated in the Zeist Community Ecology Course, as Peter J. Morin, New Brunswick
  13. 13. List of ContributorsJan P. Bakker, Community and Conservation Tadashi Fukami, Department of Biology, Stanford Ecology Group, University of Groningen, PO Box University, Stanford, CA 94305, USA, 14, 9750 AA Haren, The Netherlands, Email: tfukami@hawaii.edu Email: j.p.bakker@rug.nl E. Toby Kiers, VU University, Amsterdam,Janne Bengtsson, Department of Ecology and Crop Department of Ecological Science, Production Science, PO Box 7043, Swedish De Boelelaan 1085, 1081 HV Amsterdam, University of Agricultural Sciences, SE-750 07 The Netherlands, Uppsala, Sweden, Email: toby.kiers@falw.vu.nl Email: Jan.Bengtsson@ekol.slu.se David Kothamasi, Centre for Environmental Man-Matty P. Berg, VU University, Amsterdam, Depart- agement of Degraded Ecosystems, University of ment of Ecological Science, De Boelelaan 1085, Delhi, Delhi 110007, India, 1081 HV Amsterdam, The Netherlands, Email: dmkothamasi@cemde.du.ac.in Email: matty.berg@falw.vu.nl Dries P.J. Kuijper, Mammal Research Institute,Ulrich Brose, Darmstadt University of Technology, Polish Academy of Sciences, ul. Waszkiewicza Department of Biology, Schnittspahnstr. 10, ˙ 1c, 17–230 Białowieza, Poland, 64287 Darmstadt, Germany; Pacific Ecoinfor- Email: dkuijper@zbs.bialowieza.pl matics and Computational Ecology Lab, 1604 Nicolas Loeuille, Laboratoire d’Ecologie, UMR7625, McGee Avenue, Berkeley, CA 94703, USA, ´ Universite Paris VI, 7 quai St Bernard, F75252 Email: brose@bio.tu-darmstadt.de Paris Cedex 05, France,Jonathan M. Chase, Department of Biology and Tyson Email: nicolas.loeuille@normalesup.org Research Center, Box 1229, Washington University Michel Loreau, McGill University, Department of in Saint Louis, Saint Louis, MO, USA, Biology, 1205 avenue du Docteur Penfield, Mon- Email: jchase@wustl.edu; chase@biology2.wustl.edu ´ ´ treal, Quebec, Canada, H3A 1B1,J. Emmett Duffy, School of Marine Science and Email: michel.loreau@mcgill.ca Virginia Institute of Marine Science, The College Peter J. Morin, Rutgers University, Department of William and Mary, Gloucester Point, VA of Ecology, Evolution, & Natural Resources, 23062-1346, USA, 14 College Farm Road, New Brunswick, Email: jeduffy@vims.edu NJ 08901, USA,Jennifer A. Dunne, Santa Fe Institute, 1399 Hyde Email: pjmorin@rci.rutgers.edu Park Road, Santa Fe, NM 87501; Pacific Ecoin- Han Olff, University of Groningen, Community and formatics and Computational Ecology Lab, 1604 Conservation Ecology Group, PO Box 14, 9750 AA McGee Avenue, Berkeley, CA 94703, USA, Haren, The Netherlands, Email: jdunne@santafe.edu Email: h.olff@rug.nlJacintha Ellers, VU University, Amsterdam, Owen L. Petchey, University of Sheffield, Depart- Department of Ecological Science, De Boelelaan ment of Animal and Plant Sciences, Alfred Denny 1085, 1081 HV Amsterdam, The Netherlands, Building, Western Bank, Sheffield S10 2TN, UK, Email: jacintha.ellers@falw.vu.nl Email: o.petchey@sheffield.ac.uk xiii
  14. 14. xiv LIST OF CONTRIBUTORSJulia Stahl, Landscape Ecology Group, University of Wim H. van der Putten, Netherlands Institute Oldenburg, PO Box 2593, D-26111 Oldenburg, of Ecology, Centre for Terrestrial Ecology Germany, (NIOO-KNAW) PO Box 40, 6666 ZG Heteren; Email: julia.stahl@uni-oldenburg.de Laboratory of Nematology, WageningenMarcel G.A. van der Heijden, VU University, University, PO Box 8123, 6700 ES Wageningen, Amsterdam, Department of Ecological Science, The Netherlands, De Boelelaan 1085, 1081 HV Amsterdam, The Email: w.vanderputten@nioo.knaw.nl ¨ Netherlands; Agroscope Reckenholz-Tanikon, Herman A. Verhoef, VU University, Amsterdam, Research Station ART, Reckenholzstrasse 191, Department of Ecological Science, De Boelelaan ¨ 8046 Zurich, Switzerland, 1085, 1081 HV Amsterdam, The Netherlands, Email: marcel.vanderheijden@art.admin.ch Email: herman.verhoef@falw.vu.nl
  15. 15. Introduction Herman A. Verhoef and Peter J. MorinFirst of all, we must clarify what we consider to be American botanist H.A. Gleason. His focus was onthe subject matter of community ecology. In his the traits of individual species that allow each toclassic textbook, Krebs (1972) described a commu- live within specific habitats or geographical ranges.nity as ‘an assemblage of populations of living A community can then be seen as an assemblage oforganisms in a prescribed area or habitat’. He de- populations of different species whose traits allowscribed this as ‘the most general definition one can them to persist in a prescribed area (Gleason 1926).give’. The ‘prescribed area’ suggests that we are This is a much more arbitrary unit than that envi-dealing with some sort of spatially bounded sys- sioned by Clements. In Gleason’s view, spatialtems, and it is the somewhat arbitrary nature of boundaries of communities are not sharp and thepossible boundaries that has caused continuing species assemblages may change considerably overdebate about the nature of communities (e.g. Rick- time and space.lefs 2008). It is debatable whether community ecol- We feel that the current broadly accepted viewogy is a proper discipline at all if communities do about the nature of communities is much closer tonot exist as natural definable units (Begon et al. Gleason’s opinion. A given species can occur in1996). At the beginning of the 20th century there rather different collections of species, or commu-was much debate about the nature of communities. nities, under different circumstances. CommunityThe driving question was whether the community ecology can, therefore, be described as the study ofwas a self-organized system of co-occurring species the community level of organization, rather than ofor simply a haphazard collection of populations a spatially and/or temporally definable unit (Begonwith minimal functional integration. At that time, et al. 1996).two extreme views prevailed: one view considered As suggested by Morin (1999), we include withina community as a superorganism whose member community ecology the study of two or more spe-species were tightly bound together by interactions cies at a location, including predator–prey inter-that contributed to repeatable patterns of species actions, interspecific competition, interactions ofabundance in space and time. This concept led to mutualism and parasitism. We recognize thatthe assumption that natural communities exist as some ecologists will argue that such pair-wise in-fundamental units, and this made it possible to teractions are simply aspects of population biology,classify communities in a manner comparable to but we also point out that they provide the basicthe Linnaean taxonomy of species. One of the lead- interactions that contribute to patterns involvinging proponents of this approach was the Nebraskan larger numbers of species. We also recognize that,botanist Frederick E. Clements. His view became even in well-studied communities, much of theknown as the organismic concept of communities. It complexity and daunting diversity involving bacte-assumes a common evolutionary history for the ria, fungi, protists and small invertebrates remainsintegrated species (Clements 1916). poorly known. Nonetheless, fascinating patterns The opposite view has been termed the individu- among well-studied groups of organisms demandalistic continuum concept and was advocated by the our attention and beg for explanation. 1
  16. 16. 2 INTRODUCTION In this book, we have grouped the various chap- Community ecology has many real-world appli-ters into several areas: (1) community shape, struc- cations deriving from the fact that the abundance ofture and dynamics, (2) communities over space and species strongly depends on competitive and pred-time, (3) applications of community ecology and (4) atory interactions. These interactions are major dri-future directions of community research. We start vers of ecosystem processes, and they are the key towith shape and structure, with a focus on the topolo- the delivery of ecosystem goods and services.gy of networks of interacting organisms in ecological Chapter 7 focuses on applications of communitysystems. Chapter 1 deals with subsets of the full ecology approaches in terrestrial ecosystems: localnetwork of interactions: competition networks, mu- problems, remote causes. In terrestrial ecosystems,tualistic networks and consumption networks (food there is a clear link between belowground andwebs). It involves recent studies and important ques- aboveground subsystems, which is to be comparedtions about the structure and the processes or me- with benthic–pelagic coupling in aquatic or marinechanisms that may result in their structure. Chapter systems. These above/belowground linkages influ-2 follows with dynamics, dealing with the trophic ence many major processes, such as ecosystem res-dynamics of communities. The focus is on types of toration following land abandonment or the fate ofdynamics that can be distinguished, the dynamics introduced exotic alien species. Some examples ofof simple and complex interactions and the internal above/belowground interactions in relation toand external determinants of food web dynamics. changing land usage and biological invasions areChapter 3 considers modelling of the dynamics of com- discussed and related to climate warming. Chapterplex food webs, with a focus on whether embedding 8 reviews the consequences of industrial-scalesimple modules in complex food webs affects the fishing on marine communities and is entitled seadynamics of those modules. The dynamic analysis changes: structure and functioning of emerging marineof complex food webs is organized around the rela- communities. Growing evidence from models andtionships between complexity and stability and empirical time series suggests that fishing pressurediversity and stability. Body-size-dependent scaling can cause relatively rapid regime shifts into distinctof biological rates of populations emerge as a possi- semi-stable states that resist change even after fish-ble solution to the instability predicted for complex ing has been relaxed. Although theory predicts thatfood webs. Chapter 4 forms a bridge between a preponderance of weak interactions stabilizesdynamics and the next section on space and time. food web dynamics, these weak interactions mayThe subject of this chapter is community assembly be unlikely to buffer marine ecosystems from thedynamics in space, with a focus on three elements of impacts of intense exploitation by humans. Thus,assembly dynamics: the rate of immigration to local the systematic reduction in average food chaincommunities, the degree to which the species pool is length documented in oceanic and coastal ecosys-external to local community dynamics and the tems can initiate regime shifts to alternate semi-amount of variation in immigration history among stable states, and is already affecting the provisionlocal communities. The hypothesis is that these ele- of marine ecosystem services to society. In Chapterments together determine the degree of historical 9, which focuses on applied (meta)community ecology:contingency. Chapter 5, on increasing spatio-temporal diversity and ecosystem services at the intersection ofscales: metacommunity ecology, discusses whether a local and regional processes, several expectationsmetacommunity perspective can be helpful in from metacommunity theory on the effects of landsynthesizing community ecology across spatial use intensification are suggested, based on the factscales, and whether it can provide insights into the that both local and regional processes are importantspatial mechanisms causing variation in species co- for diversity and ecosystem functioning. Examplesexistence, strengths of species interactions and pat- drawn from research on organic farming andterns of local and regional diversity. Chapter 6 deals biological control illustrate that metacommunitywith spatio-temporal structure in soil communities and theory combined with good knowledge of the sys-ecosystem processes, and highlights the complexity tem under management is useful to understandexhibited by soil communities at different scales. how human-dominated ecosystems can be managed
  17. 17. INTRODUCTION 3at the larger spatial scales that are relevant to ary community ecology is still in its infancy, butmanagers. holds great promise. Chapter 12 deals with the Community-level management is described in emergence of complex food web structure in communityChapter 10 for salt marshes in North-West Europe. evolution models and aims at synthesizing how evo-It deals with interactions among sedimentation, lutionary dynamics may help the understanding ofnutrient availability, plant growth and herbivores. food web structures and dynamics. CommunityNatural succession in these systems results in the evolution models incorporate both the dynamicaldominance of a single tall grass species to the detri- components of food webs and the complexity nec-ment of the natural herbivores such as geese and essary to understand empirical food web data. It ishares. Introductions of large herbivores can reverse shown how this model matches topological proper-succession to facilitate geese and hares. Without ties of empirical data, while giving information onthis management, these systems would develop the dynamics of the food web. It also discusses theinto low-diversity systems of reduced value. relevance of the allometric theory of ecology to the Finally, we consider future directions of community debate on the relationship between stability andresearch. These chapters include the study of the diversity, as well as the evolution of niche breadthevolution of community processes and patterns, and non-trophic interactions. Another promisingranging from local dynamics to large-scale diversity line of research involves mutualisms. Chapter 13gradients, and focus on the role of mutualisms in deals with the role of mutualisms and communitycommunity ecology. Chapter 11 focuses on evolu- organization. Key evolutionary events, such as thetionary processes in community ecology to try to bridge origin of the eukaryotic cell, the invasion of land bythe gap between community and evolutionary ecol- plants and the radiation of angiosperms, are linkedogy by considering the reciprocal effects of individ- to mutualisms. Mutualism probably brings order toual trait variation and community characteristics. the organization, structure and function of commu-Central principles from evolutionary biology are nities by regulating the acquisition of resources andused to extrapolate the consequences of genetic ameliorating stresses. Finally, Chapter 14 specu-diversity to community level and illustrate the ef- lates about emerging frontiers of community ecology.fects of genetic variation on community properties Among various emerging areas, a focus on the im-and vice versa. It is stated that more studies have portant topic of biotic invasions seems crucial, sincebeen performed on the effects of genotypic and they alter the world’s natural communities as wellphenotypic diversity on community properties as their ecological character. Further emergingthan on the effects of community diversity on ge- areas concern questions surrounding the linkingnetic and phenotypic diversity of single species. of the structure of ecological networks to empirical-This fascinating new field of integrative evolution- ly measured community dynamics.
  18. 18. This page intentionally left blank
  19. 19. PART ISHAPE AND STRUCTURE
  20. 20. This page intentionally left blank
  21. 21. CHAPTER 1 The topology of ecological interaction networks: the state of the art Owen L. Petchey, Peter J. Morin and Han Olff1.1 Introduction strength of their interactions (Fig. 1.1). The topology of an ecological network is a description of the1.1.1 What do we mean by the ‘topology’ patterns of interactions (‘who interacts withof ecological networks? whom’). Topology is thus the study of the arrange-Many diverse systems can be described as a network ment of links from an information/organizationof linked nodes (Barbarasi 2002); for example, trans- perspective. The distances and angles betweenportation networks of cities linked by roads, net- nodes have no meaning for the topology; they areworks of interacting populations on a landscape, often chosen so that the network can be convenient-networks of interacting individuals in a social net- ly graphically represented (Fig. 1.1a). If the geome-work, networks of neurones linked by synapses, try of the network is included, the Euclideannetworks of interacting genes in the genome or net- distance between nodes has a meaning (Fig. 1.1b);works of biochemical transformations within the for example, in the genetic or trait similarity ofcell. In the paradigm used in this chapter, species species or as the physical distance between speciesare the ‘nodes’ of ecological networks, and inter- within a landscape (e.g. Thompson et al. 2001) suchactions among species are the links. as in a metacommunity context (Leibold et al. 2004). Ecological communities are immensely complex; If the interaction strength is also included (Fig.variation exists at every level of organization (in- 1.1c), all interactions are not equivalent – some aredividuals, populations, species) and these entities stronger than others.interact with each other in any number of ways In this chapter, we focus on the origins and con-(consumption, competition, mutualism, facilitation, sequences of different network topologies. Chaptermodification). Ecological networks provide a tract- 2 will address the origin and consequences of vari-able simplification of this complexity – they can be ation in interaction strength within interaction net-constructed, modelled, experimentally manipu- works. Fundamental questions about the topologylated and analysed with available tools and re- of ecological networks include how many nodessources (Proulx et al. 2005). This chapter focuses (species) there are, how many links (interactions)on the networks of interactions among the species there are in total and how interactions arein ecological communities to highlight the advances distributed among species pairs (Fig. 1.1a).in ecological understanding that research about The possible interactions that can be representedthese networks can provide. in ecological networks theoretically include all of the The links in ecological networks can be charac- basic pair-wise interactions among species, includ-terized through three main aspects: (1) their topol- ing competition via various mechanisms, predator–ogy, (2) their geometry and (3) the direction and prey and host–parasite interactions, interactions 7
  22. 22. 8 SHAPE AND STRUCTURE (a) (b) Birds (Larus, Haematopus, Corvus) + + + + − Small starfish (Leptasterias) + + − − Snails (Nucella) + + − − − − Acorn Goose − 0 − − Mussels barnacles − barnacles (Semibalanus) (Pollicipes) (Mytilus) (c) Recycling Spiders Spartina 305 Insects 81 28 34580 6585 Photo. Resp. 23 224 53 5 Mech. Mech. 27995 Mud Crabs 6280 600000 59 4010 Bact. 462 Crabs 90 31 6 3890 Nemas 25 372 Algae 1620 1800 Photo. Resp. Mech. Mech. 3671 180 Net 28175 3890 644 Export 0.6% 4534 100 103 106 563620 32709 kc/m2/yr 93.9% 5.4%Figure 1.1 Examples of three kinds of ecological networks. (a) A food web for a portion of the grassland community nearSilwood Park, UK. Reprinted with permission from www.foodwebs.org using data from Dawah et al. (1995). (b) Aninteraction network showing the signs of interactions among a subset of species in a community from the rocky intertidalzone of the Pacific Coast of northwestern USA. Reprinted from Wootton (1994), with permission of the EcologicalSociety of America. (c) A network of energy flow from a salt marsh in southeastern USA. Reprinted from Teal (1962),with permission of the Ecological Society of America.
  23. 23. THE TOPOLOGY OF ECOLOGICAL INTERACTION NETWORKS 9through behavioural interference and positive inter- were recorded (Dawah et al. 1995). However, manyactions including mutualisms and commensalisms. food webs are more complex and less complete,In practice, few representations of ecological net- with less well-resolved taxa, especially in the smal-works include all of these kinds of interac- ler, relatively inconspicuous and taxonomically dif-tions, partly for historical reasons and partly ficult groups such as microbes and micro-because, for reasons of pragmatism and feasibility, invertebrates.the ecologists who describe these networks often One should note the very important assumptionfocus on only a subset of possible interactions, e.g. in the kind of the topological approach shown inwhen they study the importance of a particular type Fig. 1.1: every species is treated equally in the net-of interaction in structuring the ecological commu- work from an interaction perspective. For example,nity of interest. species that are very abundant in biomass and spe- cies that are very rare are treated exactly the same. Only if these differences in abundance result in a1.1.2 Different types of ecological difference in their number of interactions withnetworks other species will those differences show up in theWhat might constitute the links among species in topology of the web. In other words, the centralan ecological network? Very broadly, one can think assumption of the approach is that the key pointof two types of information represented by a link. for understanding the food web is that species in-The first is when a link represents one specific teract, while information on how abundant differ-biological mechanism or process, the second is ent species are and how strong these interactionswhen a link represents the net effect of a variety of are (e.g. expressed in per capita effects of predatorsmechanisms or processes. Links that represent on prey, and vice versa) can be disregarded. This isbiological mechanisms or processes can be recog- of course a strongly simplifying and yet poorlynized by the transfer of something tangible (such as tested assumption. Efforts to attach more informa-biomass, energy, nutrients, information or combi- tion to nodes have led to significant advances innations of these) between the linked entities, where understanding (Cohen et al. 2003). Nevertheless, athis usually requires close physical proximity be- huge body of research concerns the search for pat-tween the organisms involved in time and space. terns in food web topology (Cohen 1978). Food webs are an example of an ecological Plant–pollinator networks also use links to repre-network with this type of link, in which the sent transfers of something tangible among organ-consumer–resource interactions represent transfer isms that share close proximity in space and time.of energy and material and require a physical inter- Pollen moves from plant to pollinator, and again toaction between individuals. Food webs are perhaps the plant, which requires repeated physical contact.the most commonly encountered type of ecological These interactions involve different currencies.network (Elton 1927; Cohen 1978; Pimm 1982; Polis From the plant perspective, the exchange of geneticand Winemiller 1996; de Ruiter et al. 2005). An information through cross-fertilization is the mostexample of a food web, the Silwood Park, UK, important aspect. From the pollinator perspective,grassland food web (Dawah et al. 1995), is shown the resource rewards of visiting the flower are thein Fig. 1.1a and depicts species as ‘balls’ and feed- driving force. Interest in these so-called two-layering links as ‘sticks’. In this particular food web, networks is a rather recent development, at leastthere are plants, herbivores, primary parasitoids of compared with the long study of food webs, andthe herbivores, and hyperparasitoids. Trophic posi- the networks typically describe which plants aretion increases with height and is also coded by visited by which pollinators or seed disperserscolour (from www.foodwebs.org). This is a rela- (Bascompte et al. 2003; Jordano et al. 2003; Rezendetively simple food web (87 species) in which the et al. 2007).identity of each resource and consumer is resolved Networks of energy and material flow also docu-to the species level and sufficient sampling effort ment this first broad type of link, and they have aensured that virtually all species and feeding links venerable place in ecology, dating back at least to
  24. 24. 10 SHAPE AND STRUCTURELindeman (1942). Fig. 1.1c shows the quantification termed competition, and can result from consump-of the patterns of energy flow in a salt marsh in tion (a biological process) by two species of a sharedNorth America among the major functional compo- resource, as well as from a number of other non-nents (Teal 1962). Owing to measurement con- consumptive mechanisms, such as direct beha-straints and the assumed functional equivalence of vioural interference among species (Schoener 1983).different species involved with respect to nutrient When links represent effects of one species onand energy transformation, these networks typical- another, they may result from direct effects, indirectly have low taxonomic resolution, and have mostly effects or net effects. Competition via consumptionbeen developed as a way to understand and of a shared resource can be represented as adescribe energy fluxes and material in ecosystems, negative–negative link arising from an indirectand the associated ecosystem functions and ser- interaction between two competing speciesvices (such as primary and secondary productivity, mediated by their joint consumption of a thirdcarbon sink capacity, etc.). That is, the focus is the resource species; that is, there need not be a transferdynamics of energy and material, rather than of anything tangible from one competitor to thethe dynamics of species and their interaction struc- other, yet changes in the abundance of one willture. An exception is the work by Ulanowicz (1997), alter the abundance of the other. Other indirectwho is one of the rare authors who has extensively effects (or interactions) include apparent competi-considered origins and consequences of different tion (negative–negative effects caused by the pres-topological patterns of energy and nutrient flows ence of a shared consumer; e.g. Holt 1977) andin food webs. In his work, he stresses the impor- apparent mutualism (positive–positive effects, e.g.tance of positive feedback among species, resulting in a trophic cascade; Pace et al. 1999).in autocatalytic loops within subsystems of whole Ecological networks in which links represent thewebs, which may even lead to ecosystem-level net effect of one species on another (the outcome ofcompetition between such alternative loops. Also, all direct and indirect effects) are often, and per-the work on soil foods webs by de Ruiter, Moore haps confusingly, termed interaction webs. As thisand others (e.g. de Ruiter et al. 1998) has tried to also describes the general class of ecological net-map the flows of nutrients and energy from a much works in which species interact (irrespective of themore topological perspective than is done in classic nature of their interactions) they should be moreecosystem science. The outcome of this work will be correctly called interaction-sign networks. They de-addressed further in Chapter 2. pict the sign (positive, negative, zero) and some- The second broad type of information repre- times the magnitude of the net impact of changessented by a link is the effect of one entity on another. in the abundance of one species on another. TheseFor example, if individuals of two species each have net impacts can be the result of all kinds of directa positive effect on the growth and reproduction of and indirect interspecific interactions. Interactionindividuals of the other, they would be linked. This webs are less commonly depicted in the literature,positive–positive interaction is termed mutualism perhaps because of the practical difficulty ofand a range of biological mechanisms might under- determining the sign of net interactions, especiallylie this link. Pollination is an example of a mecha- within trophic levels, for which elaborate field ex-nism that can result in a positive–positive periments are often required. However, sometimesinteraction. Associational resistance, in which a the nature of interactions can be inferred fromspecies gains protection from its consumers by spa- large-scale natural disturbances to ecosystems. Atially associating with its potential competitors, is classic example is the effect of rinderpest on theanother example that can lead to a net positive food web structure and ecosystem functioning ofinteraction between species (Olff et al. 1999). Facili- the Serengeti (Sinclair 1979). In this work, abiotictation, such as often found among grazing herbi- components and processes, such as fire, are alsovores (Huisman and Olff 1998; Arsenault and taken into account as ‘nodes’ in the food web, be-Owen-Smith 2002), also belongs to this class of in- cause of the strong effect biotic feedbacks that theyteractions. A negative–negative interaction is receive. One good example of an interaction-sign
  25. 25. THE TOPOLOGY OF ECOLOGICAL INTERACTION NETWORKS 11web comes from Wootton’s (1994) work on inter- Hairston (1981) studied a group of six species ofactions among some of the species living in the rocky plethodontid salamanders found in moist forests ofintertidal zone of Washington State in the USA the mountains of North Carolina, USA. Based on(see Fig. 1.1c). The search for patterns in interaction their morphological similarity and similar life-networks has only begun rather recently. For a styles, all being carnivores feeding on small inver-recent review, see Ohgushi (2005), who focuses on tebrates, it seemed plausible that all six speciesindirect interactions in plant-based terrestrial food might compete for food (Fig. 1.2). Over a 5 yearwebs. period Hairston removed each of the two most common species at this site. Only one of the remain- ing five species responded positively to each re-1.1.3 Three general questions moval, and this was the most common species at the site. These results imply that only the two mostFor each type of ecological network, we suggest that common species compete. Hairston suggested thatthere are three general types of questions: those competition might be for nest sites, rather than forabout whether and what structural regularities shared prey, but subsequent research showed thatexist within and among observed networks; those direct territorial (interference) interactions betweenabout which mechanisms are responsible for the salamanders constituted the primary mechanism ofstructure of the networks and of any structural reg- competition (Nishikawa 1985). In this case, the net-ularities that exist; and lastly, those about the gaps in work of competitors is considerably simpler thanour knowledge. The aim here is to define a set of the completely connected system that would resultgeneral contemporary questions that we will use to if all species shared and competed for commonorganize each of the following sections. resources (Fig. 1.2). Another general and important question about the structure of competitive networks is whether1.2 Competitive networks the interactions form a hierarchy, or an intransitive loop. In a hierarchy, for example, species A out-1.2.1 Structural regularities competes species B and C, B outcompetes C, andReal-world competitive networks remain poorly C cannot outcompete either A or B. This will mostlystudied and documented. A fundamental question lead to predominance of a single competitively su-about their structure remains quite unresolved: perior species (A) and therefore will foster relative-given a set of co-occurring species that appear ly low biodiversity. In an intransitive loop, in whichto consume a similar class of resources (such as A outcompetes B, B outcompetes C, and C outcom-different-sized seeds, forming a potential guild, petes A, all three species can potentially coexist.sensu Root 1967), how many pairs of these species Consequently, mechanisms of interaction thatreally compete? That is, how many of the plausible allow intransitive competitive networks might ex-links are actually realized in competitive networks? plain high biodiversity in some ecosystems. Groups of taxonomically similar species are often At first hand, intransitive competitive networks,assumed to be potential competitors, but experi- especially within one trophic level, would be un-ments show that only a small fraction of the poten- likely to emerge, based on the theory of life historytial competitors actually compete. Sometimes this and competitive trade-offs. An organism’s life his-can be generalized to simple integrative traits, tory generally reflects a compromise in the use ofbuilding on the early work on limiting similarity available resources. Species generally specialize(for a review, see Brown 1981). For savanna large along environmental axes of resources axes andherbivores, Prins and Olff (1997), for example, stress factors in such a way that investing in somefound that large herbivores overlap more in life history/functional trait (e.g. the ability to crackresource use when they are more similar in body big seeds in birds) occurs at the expense of somesize. Thus, more similar-sized species are more other function (e.g. the efficiency with which smalllikely to interact. seeds can be collected). In plants, for example, the
  26. 26. 12 SHAPE AND STRUCTURE (what is the opponent sensitive to) than energy- Plethodon serratus or nutrient-based (which requires morphological adaptations, such as allocation shifts). Plethodon Plethodon One example of such an intransitive competitive glutinosus jordani network involves different strains of bacteria competing in spatially structured environments. Desmognathus Desmognathus ochrophaeus wrighti Kerr et al. (2002) explored how a mixture of me- chanisms can allow three strains of Escherichia coli Eurycea cirrigea to coexist via a set of intransitive interactions that are akin to a game of rock–paper–scissors. There are three key features of this system – one strain Plethodon produces a toxin, called colicin, whereas the two serratus other strains are either sensitive or resistant to the toxin (Kirkup and Riley 2004). The three strains Plethodon Plethodon interact as an intransitive network in the following glutinosus jordani way. All of the strains compete consumptively for a carbon source, such as glucose, but they differ in Desmognathus Desmognathus competitive ability such that colicin-sensitive ochrophaeus wrighti strains are the best consumptive competitors, followed by colicin-resistant strains followed Eurycea by colicin-producing strains. However colicin- cirrigea producing strains can displace colicin-sensitive strains by poisoning them, but colicin-producingFigure 1.2 Potential (top) and realized (bottom) networks strains can in turn be displaced by competitivelyof competing salamanders in the study of Hairston superior colicin-resistant strains. These patterns of(1981). Dashed lines in the lower graph indicate sequential displacements occur only in an un-unrealized competitive interactions; the heavy solid lineindicates that only two species, Plethodon glutinosus and mixed spatially complex environment, such asPlethodon jordani, actually compete. the surface of a culture plate. In a well-mixed spatially homogeneous environment, such as a well-stirred liquid culture, only one strain per-ability to compete well for nutrients has been sug- sists. It appears that limited dispersal can promotegested to happen at the expense of the ability to diversity by allowing an intransitive competitivecompete for light (Tilman 1990). Also, competing network to exist (Fig. 1.3). A similar role is playedneeds for growth and reproduction result in differ- by dispersal limitation in maintaining diversity inent optimal combinations in different species, de- neutral communities (i.e. where all species are as-pending on the conditions to which they are sumed to be functionally equivalent; e.g. Hubbelladapted. Such evolution-driven trade-offs make 2001).competitive hierarchies much more likely than Qualitatively similar but much more complextransitive networks in most cases. Said simply, it networks may explain the high diversity of bacteriais unlikely that the species adapted to the ‘left-tail’ that manage to coexist in soils, where dispersal isof some underlying niche axis will outcompete a limited. For example, Torsvik et al. (1990) madespecies from the ‘right-tail’ or vice versa. initial estimates using DNA hybridization techni- An important class of exceptions exists in organ- ques that suggested that thousands of bacterial taxaisms in which the outcome of interactions is not could occur per 30 g of soil. These conservativedriven by resource competition, but by chemical estimates were re-evaluated by Dykhuizen (1998),warfare, such as allelopathy. In this case, few who estimated that 30 g of forest soil could containtrade-offs are expected, as the defence and compe- over half a million species, depending on the as-tition is much more strongly information-based sumptions made in the analysis.
  27. 27. THE TOPOLOGY OF ECOLOGICAL INTERACTION NETWORKS 13a Static plate resources, might provide a mechanism to explain how so many bacterial taxa can coexist in soils 12 ´ ´ (Czaran et al. 2002). These authors present a theo- 10 retical analysis showing that, by including multiple Log (abundance) 8 toxins and multiple toxin resistance genes, a large number of taxa will persist in a two-dimensional 6 model grid of interacting bacteria. They find that a 4 system with a few toxin genes and many resistance 2 genes can support up to 1000 different strains in a 180 Â 180 spatial grid. This is obviously somewhat 0 0 10 20 30 40 50 60 70 less than the diversity estimates obtained for soils, but it is of the right order of magnitude.b Flask 12 1.2.2 Mechanisms 10 Log (abundance) 8 Intransitive networks of competitive interactions may explain the high diversity of some systems of 6 competitors. It appears that high diversity can be 4 maintained by intransitive competitive networks 2 only in spatially structured environments, such as soils or two-dimensional surfaces (Reichenbach 0 0 10 20 30 40 50 60 70 et al. 2007). In well-mixed environments without opportunities for much spatial structuring, suchc Mixed plate as aquatic systems, other mechanisms must be in- 12 voked to explain the coexistence of large networks 10 of competitors.Log (abundance) 8 What maintains competitor diversity in well- C mixed habitats with only few limiting resources? 6 S One possibility has been suggested by Huisman 4 R and Weissing (1999). Their models show that 2 large networks of many resource competitors, when competing for several resources (four to 0 0 10 20 30 40 50 60 70 six), can persist due to chaotic fluctuations in spe- Generations cies abundances (Fig. 1.4). They suggest that switches along alternative transitive networks alsoFigure 1.3 Dynamics of an intransitive competitor seem to play a role. The presence of multiple limit-network. C is a colicin (toxin producing) bacterial strain, ing nutrients in their model prevents the formationwhich excludes S, a colicin-sensitive strain, which excludesR, a colicin-resistant strain, which in turn competitively of clear competitive hierarchies. This might pro-displaces C. All three strains persist in a static two- vide an explanation for the coexistence of largedimensional habitat (a), but R eventually predominates in numbers of phytoplankton species in well-mixedmixed environments (b and c). Reprinted with permission environments such as lakes and oceans, a problemfrom Macmillan Publishers Ltd: Nature , 171–4, B. Kerr et al. that has interested ecologists for many years# 2002. (Hutchinson 1961). Recently, Huisman and collea- gues found experimental evidence for similar Intransitive networks involving multiple toxins chaotic dynamics in a closed experimental multi-and multiple resistance factors to those toxins, trophic plankton system that was observed for aalong with trade-offs between the ability to produce large number of generations (Beninca et al. 2008).or resist toxins and to compete consumptively for This seems to be the first well-documented example
  28. 28. 14 SHAPE AND STRUCTURE a 60 50 5 Species abundances 40 3 30 2 20 4 10 1 0 0 100 200 300 Time (days) b 55 c 150 50 120 Total biomass 45 Species 5 90 40 60 35 30 30 6 25 9 0 20 25 12 0 100 200 300 15 1 Spec 30 35 ci es Time (days) ies 3 40 18 SpeFigure 1.4 (a) Example of coexistence in a system of chaotically fluctuating consumptive competitors. Reprinted withpermission from Macmillan Publishers Ltd: Nature , 407–10, J. Huisman & F.J. Weissing, # 1999. (b) Populations of threespecies orbit around a strange attractor, typical of chaotic dynamics. (c) Although the populations fluctuate chaotically,total biomass summed over all species is invariant over time.of chaotic dynamics in a food web. Whether such limiting resources, leading to an expectation of onlychaotic dynamics also frequently occur in natural few species to coexist, each limited by a differentplankton communities still needs to be seen though, resource. However, different algae species seem toas the experimental set-up prevented many types of have specialized on different wavelengths of theinteractions between biota and abiotic factors (such light spectrum, which therefore is in fact a wholeas sediments). One can imagine that fluctuations in class of resources (Stomp et al. 2007a,b). These ‘col-external forcing factors, such as light and tempera- ourful niches’ seem to play a yet underestimatedture, may overrule the potential for chaotic dynam- role in the coexistence of plankton species.ics in such systems. Of course, other mechanisms that minimize the Recently, another way out of the ‘paradox of the importance of competition among species can alsoplankton’, or the ‘principle of competitive exclu- operate to maintain diversity in spatially structuredsion’ (no more species are expected to coexist than or well-mixed environments. Perhaps the best knownthe number of limiting resources), has been offered of these mechanisms is keystone predation (Paineby the same research group. Classic explanations 1966), in which consumers feed preferentially on com-for coexistence of plankton species have explored petitively superior species and prevent competitivelight, nitrogen, phosphorus and silicon as the main exclusions of weaker competitors (Leibold 1996).
  29. 29. THE TOPOLOGY OF ECOLOGICAL INTERACTION NETWORKS 151.2.3 Unresolved issues plant–pollinator systems (Feinsinger 1976; Petan- diou and Ellis 1993; Fonseca and John 1996; WaserExcept for a very few well-studied and relatively et al. 1996), and recently there has been a growth insimple examples, we know little about the structure the use of the network perspective to analyse theof real competitive networks. This is partly due to greater amounts of data collected on mutualisticthe difficulty of clearly establishing whether species networks (see following references in this section).actually compete, if so to what degree, and of identi- For the main part, the two types of mutualistic net-fying which mechanisms are in play. The necessary works examined are plant–pollinator networks andexperiments are often difficult to perform, except plant–seed disperser networks. The conventionunder simplified conditions in the laboratory. used to depict these networks is to show interac- How would inclusion of competitive networks tions in a two-layer network between one group ofinto food webs (networks in which links are con- species (plants) and their mutualists (pollinators orsumption) alter our predictions about web stabili- dispersers), as shown in Fig. 1.5. The species are notty? We can guess, in part, that inclusion of depicted on different trophic levels, as in foodcompetitive interactions into a community matrix webs, and only interactions between the twobased only on predator–prey interactions would groups of mutualists are considered (no direct com-tend to destabilize the network, to the extent that petitive interactions are included).May’s (1972, 1973) analysis holds true. Mutualistic networks appear to have several How common are subnetworks of intransitive types of structural regularity. First, there are nestedcompetitive interactions, and do they explain the sets of interactions (Bascompte et al. 2003; Lewin-persistence of high species richness? Although they sohn et al. 2006). That is, more specialized mutual-have long been recognized as a potential scenario for ists tend to interact with a proper subset of themaintaining diversity in systems (see Connell 1978), species that more generalist mutualists interactwe still have very few well-documented examples of with. One consequence is that there is a set of spe-non-transitive competitive networks of any sort. cies that form a highly connected ‘core’ in mutual-Minimally, such networks would require that com- istic networks. This makes the number of linkspetitors interact through more than one kind of com- across the network required to connect any twopetitive mechanism (Schoener 1983). The recurring species rather short (Olesen et al. 2006). Indeed,problem is that, based on traditions, expertise and mutualistic networks tend to be more nested thanpractical limitations, almost all studies still focus on food webs and to have shorter paths between anysubsets of entire food webs, such as beetle commu- two species than food webs (Bascompte et al. 2003).nities, plant communities, pollinator communities, This ‘core’ of highly connected interactors appearssoil food webs, soil microbial communities, but to have consequences for how mutualistic networkshardly ever study the entire system with the same respond to disturbance, and seems to make themlevel of aggregation. This will require large interdis- more robust to potential perturbations (Bascompteciplinary efforts and most likely new theoretical ap- et al. 2003).proaches to deal with the resulting data. Another structural feature of mutualistic net- works is of asymmetric patterns of interactions (Bascompte et al. 2006; Vazquez et al. 2007). In gen-1.3 Mutualistic networks eral, species with high numbers of connections tend to interact with those that are connected to relative-1.3.1 Structural regularities ly few species. This asymmetry in connectionsFor the most part, mutualisms do not figure promi- within mutualism webs is consistent with the fea-nently in traditional depictions of ecological net- tures of models that confer greater stability on theseworks. This reflects, in part, short shrift given to networks (Bascompte et al. 2006).reciprocal positive interactions in community ecol- Similar to competitive networks, mutualistic net-ogy (Boucher 1985), at least until recently (Brooker works can also show an intransitive structure, alsoet al. 2008). Some of the earliest works consider called hypercycles. Three species may be arranged

×