measuring biodiversity for conservation/media/royal_society... · measuring biodiversity for...

Measuring biodiversity for conservation

Policy document 11/03

August 2003

ISBN 0 85403 593 1

This report can be found atwww.royalsoc.ac.uk

Professor Peter Crane FRS

Chair of the Royal Society working group on measuring biodiversity forconservation

Nearly every week new evidence emerges of the steady erosion of life’s diversity, driven byhuman population growth, profligate use of natural resources and the loss and degradationof habitats. Since we depend on the Earth’s rich variety of life, for food, fibres, fuel,pharmaceuticals and many other goods and services – we should be concerned about whatimpacts these losses will have on our own species, and our prospects for the future.

This report was motivated by the recognition that science has an important role in supportingeffective conservation and sustainability practices. Considerable effort is already directed toward biodiversityassessment, conservation and sustainable development. However progress is impeded by major gaps in knowledge ofbiodiversity. Addressing these gaps requires the scientific community to place an urgent emphasis on the synthesis ofscattered data, which must also be supported by a more favourable attitude to such projects by the funding bodies. Inorder to better measure the loss of biodiversity, existing monitoring programmes should also be reviewed and newprogrammes, incorporating improved and consistent sampling methods, need to be implemented for the collection ofnew data. Applying the framework developed in this report would help ensure that measurements undertaken by suchprogrammes are both scientifically sound and appropriate to the purpose to which they are being applied.

In September 2002 the World Summit on Sustainable Development in Johannesburg set down the challenge ofsignificantly reducing the current rate of biodiversity loss by 2010. For governments charged with the responsibility todeliver against this demanding challenge, for the evolving work programme of the Millennium Ecosystem Assessment,and for the scientists, conservationists and businesses working to halt biodiversity loss, we hope that this report is avaluable contribution to the essential but difficult task of measuring the state of biodiversity.

My thanks go to all members of the working group who worked extremely hard to complete this report. We are alsoespecially grateful to all those who contributed to the report, by submitting their views or through the consultationworkshops.

Sir Patrick Bateson

Biological Secretary and Vice-President of the Royal Society

The World Summit on Sustainable Development made a commitment to a significantreduction in the rate of loss of biodiversity by 2010. How should this objective be monitored?Much work has been done in the attempt to assess the global status of biodiversity and awide range of measures is already in use. However, debate about losses has already revealedthat these measures are inadequate. Different sets of data collected over time are needed todemonstrate what is happening. Achieving this will require co-ordination and co-operation between conservation groups, academic scientists andgovernmental and intergovernmental agencies.

The Royal Society is keen to engage interested parties in its attempts to help monitor losses of biodiversity. A widerange of organisations and individuals are working on biodiversity measurement and we are particularly grateful tothose who responded to our call for views and participated in the two consultation workshops we hosted in November2002. I am especially indebted to Professor Peter Crane, the other members of the working group and the secretariatfor the considerable effort that has gone into this report. The report provides a valuable contribution for thedevelopment of a robust and scientific foundation for the global assessment of biodiversity. It will be of use to all thoseinvolved in commissioning, funding and undertaking biodiversity measurement. I warmly commend it to policy makers,national and international conservation bodies, professional scientists and their funding bodies, including commercial,non-governmental (NGO) and governmental organizations.

Foreword

Measuring biodiversity for conservation

Contents

pagePreparation of this report v

Summary vii

1 Introduction and context 11.1 The reality of ongoing biodiversity loss 11.2 Definition 11.3 The significance of biodiversity 11.4 Role of science in the conservation and sustainable use of biodiversity 21.5 International agreements and policy responses to biodiversity loss 21.6 Overview of the Report 4

2 Progress and challenges in measuring biodiversity 52.1 Progress towards measuring biodiversity 52.2 Challenges 62.3 Problems with existing measures and their application 92.4 Areas where rapid progress is possible 92.5 Refining assessments of biodiversity 11

3 A framework for measuring biodiversity 133.1 Introduction 133.2 Structure of the framework 133.3 The scoping stage 133.4 The design stage 173.5 The implementation and reporting stage 203.6 Conclusion 22

4 Case studies 234.1 Forest cover in the Eastern Arc Mountains of Tanzania 234.2 Global extent of open freshwater habitats 244.3 Trophic integrity of marine ecosystems 264.4 Meiofaunal indicators of pollution 274.5 Global state of plant biodiversity at the species level 284.6 Genetic diversity in wild lentils 304.7 Freshwater fish of conservation concern in the UK and Mexico 324.8 Barndoor skate 334.9 Trinidadian Guppy 354.10 Brown argus butterfly 364.11 The tiger in India 384.12 Case study conclusions 39

5 Conclusions and recommendations 41

Annex A List of submissions to call for views 43

Annex B List of participants in the consultation meetings 45

Annex C Contributors of additional case studies 46

The Royal Society Measuring biodiversity for conservation | August 2003 | iii

Annex D International agreements: policy responses to biodiversity loss 47D.1 International agreements and the Convention on Biological Diversity 47D.2 At the European Union level 48D.3 The United Kingdom 48D.4 NGOs and industry 48

Annex E Abbreviations and glossary 50

Annex F References and relevant websites 52Relevant websites 56

ISBN 0 85403 593 1

© The Royal Society 2003. Requests to reproduce all or part of this document should be submitted to:Science Advice SectionThe Royal Society6–9 Carlton House TerraceLondon SW1Y 5AG

Preparation of this report

This report has been endorsed by the Council of the Royal Society. It has been prepared by the Royal Society workinggroup on measuring biodiversity for conservation.

The members of the working group were:

Chair Prof Peter Crane FRS Royal Botanic Gardens Kew

Members Dr Andrew Balmford Dept of Zoology, University of Cambridge

Prof John Beddington FRS Dept of Environmental Science and TechnologyImperial College London

Prof Bland Finlay NERC Centre for Ecology and Hydrology

Prof Kevin Gaston Dept Animal & Plant Sciences, University of Sheffield

Dr Rhys Green Royal Society for the Protection of Birds and Department of Zoology,University of Cambridge

Prof Paul Harvey FRS Dept of Zoology, University of Oxford

Dr Sandy Knapp Natural History Museum

Prof John Lawton CBE FRS Chief Executive, NERC

Prof Georgina Mace OBE FRS Institute of Zoology

Prof Anne Magurran St Andrews University, Gatty Marine Laboratory

Lord May of Oxford OM AC PRS President, Royal Society

Dr Eimear Nic Lughadha Royal Botanic Gardens Kew

Prof Peter Raven ForMemRS Missouri Botanic Gardens

Prof John Reynolds School of Biological Sciences, University of East Anglia

Prof Callum Roberts Environment Dept, University of York

Prof Kerry Turner CBE Centre for Social and Economic Research on the Global Environment,University of East Anglia

Secretariat Mr Richard Heap Science Advice Section, Royal Society

Dr Rachel Quinn Science Advice Section, Royal Society

Mr Robert Banes Science Advice Section, Royal Society

The Royal Society Measuring biodiversity for conservation | August 2003 | v

The living world is disappearing before our eyes. Losses ofbiodiversity have accelerated over the last two centuriesas a direct and indirect consequence of humanpopulation growth, unsustainable patterns of resourceconsumption and associated environmental changes.Effective methods of measuring biodiversity are neededto monitor changes in the state of nature and to measureprogress towards the target, set by the World Summit onSustainable Development, of achieving ‘a significantreduction in the current rate of biodiversity loss by 2010’.No sound scientific basis currently exists for assessingglobal performance against this target.

Enough is already known of the distribution and drivers ofbiodiversity loss to indicate the scale of the problem andto provide a basis for the urgent conservation action thatis needed to prevent many species from beingirretrievably lost. This action must not be delayed.However without the right measures it will be impossibleto determine whether rates of loss of biodiversity aredeclining or accelerating, and therefore impossible todemonstrate the effectiveness of mitigating actions.Good measures of rates of biodiversity loss are lacking formany parts of the world as well as for many groups oforganisms. Better measures, that are cost effective andbased on sound science, are crucial to help assess successin managing biodiversity and preventing further losses.

A significant impediment to the development of measuresthat are improved in terms of efficacy and scope is theextraordinarily limited knowledge of many aspects ofbiodiversity. Knowledge about the total number of speciespresent on earth is poor, and for many of those that havebeen described little or nothing is known of theirdistribution, ecology, population size or evolutionaryhistory. The fate of organisms that have not yet beenrecognised by science cannot be measured. Likewise, howecosystems function cannot be understood until more isknown about the organisms of which they are comprised.Knowledge is most limited and patchy for the verygeographic areas and biomes where species diversity isgreatest – principally in the tropics; and next to nothing isknown of the deep sea. Understanding trends both in timeand distribution of biodiversity is further hampered by theabsence of reliable baseline data for most groups andhabitats, as well as by inconsistencies in methodology.

Measures of biodiversity vary in scale and purpose. Theyextend beyond the species level to encompass entirehabitats and ecosystems, and can also focus morenarrowly on the details of populations and genes. No onemeasure is best for all purposes. A broad suite ofmeasures is necessary to meet specific needs but thesheer multiplicity of current measures contributes to thedifficulty of building public awareness andunderstanding. Selecting the most appropriate measure,

especially at large scales, requires a careful considerationof the purpose of the assessment as well as the tradeoffsbetween usefulness, completeness and required effort interms of time and other resources.

We therefore recommend that the framework, developedin this report, be applied routinely in all situations, bythose commissioning, funding and undertakingbiodiversity measurement. Implementation of theframework would ensure that measures are appropriateto the purpose to which they are being applied. As aresult each biodiversity assessment would clearly identify:i) interested parties; ii) the attributes which those partiesvalue and are seeking to measure; iii) the extent ofexisting knowledge relevant to the assessment; iv) theassumptions used in the assessment and the limitations ofthe measure in addressing the valued attributes; v)precisely how each measure is defined; vi) the nature ofthe sampling strategy used; and vii) the data gatheringand analytical methods to be employed. Applying thisframework would also help to identify weaknesses incurrent approaches, as well as highlight major scienceand information gaps. A series of case studies in thereport demonstrates how this approach can be used in anumber of circumstances, for terrestrial, freshwater andmarine systems, and at the ecosystem, species andpopulation levels.

We recommend an urgent emphasis on synthesis ofexisting knowledge, by those working on biodiversity.Making otherwise scattered data more readily availableand more useful, for example by the use of web-basedtechnology, requires a more favourable attitude towardssuch projects by funding bodies. Key gaps in knowledge,revealed by such synthesis, should be addressed by theurgent development of new programmes with realisticgoals that can be completed in the next three to sevenyears.

Future data gathering and analytical techniques shouldaim to provide biodiversity information that is bothrelevant to interested parties and usable by other similarassessments. Enhancing the quality of baselineknowledge will facilitate the use of the framework as wellas the development of more effective measures withexpanded scope. Maximising the efficiency with whichbaseline information – including that generated bysystematists – is transferred, and made useful, toconservation biologists, is especially crucial. Enhancingthe level of taxonomic training, and linking such trainingmore directly to the ongoing measurement andmanagement of biodiversity, especially in countries wherebiodiversity is high, will also be essential. Combined withwell-designed sampling strategies and application of newtechnologies, these initiatives could transform the currentknowledge of changes in habitat types, patterns and

The Royal Society Measuring biodiversity for conservation | August 2003 | vii

Summary

rates of delivery of ecosystem services, distributions ofspecific taxa and changes in population abundance.

We also recommend that new and existing programmesof biodiversity assessment focus on establishing abaseline and rate of change. Effective initiatives, thatmeet their goals, should be maintained and, where

appropriate, expanded. Without a marked increase infunding, as well as coordination and cooperation amongnon-governmental organisations (NGOs), academics andgovernmental agencies, measuring progress towardsreducing rates of biodiversity loss by 2010, thecommitment of the 2002 World Summit on SustainableDevelopment, will be unachievable.

The Royal Society viii | August 2003 | Measuring biodiversity for conservation

1.1 The reality of ongoing biodiversity loss

The living world around us is disappearing before oureyes. Around a tenth of all the world’s bird species anda quarter of its mammals are listed by the WorldConservation Union (IUCN) as threatened withextinction (IUCN 2002). For less studied groups such asfish, mussels and crustacea, the proportion underthreat could be as high as one or two thirds (Master etal 2000; IUCN 2002). Between half and one percent ofthe world’s tropical forests are still being lost each year(FAO 2001; Achard et al 2002). Since the early 1980s,over one third of all mangroves have been cleared(Valiela et al 2001). Long-term studies indicate that wildvertebrate populations have declined in number by anaverage of over one-third since 1970 (Loh 2002), sharkpopulations in the Northwest Atlantic have fallen asmuch as 75% since 1986 (Baum et al 2003), andamphibian population sizes worldwide have decreasedby around 80% in 50 years (Houlahan et al 2000). Thisinformation is in the public domain but a systematicframework for assimilating data on the loss ofbiodiversity, and for assessing its impact on society,does not exist. The aim of this report is to clarify thescientific basis for measuring biodiversity in order tocontribute to an international consensus on howbiodiversity can be monitored.

Losses of biological diversity are being driven, primarily, byhuman population growth and by unsustainable patternsof resource consumption, reinforced by inappropriateeconomic structures and activities that maximise short-term gain, without considering long-term consequences(Raven 2002). There is broad scientific consensus thatwithout an adequate response to the resulting pressureson natural ecosystems – loss, fragmentation anddegradation of habitats, overexploitation of wild species,the introduction of non-native alien species, and climatechange – biological diversity will continue to be lost at arate that is unprecedented since the appearance ofmodern kinds of ecosystems more than 40 million yearsago (May et al 1995; Pimm & Askins 1995; Wilson 1999;Myers & Knoll 2001). Action is needed now, and ifconservation strategies and policies wait for perfectknowledge many species will be lost.

This section considers why these losses matter, whatinternational and national political tools already exist toaddress biodiversity loss and how science must play a rolein measuring change in biological diversity.

1.2 Definition

In this report we have adopted the Convention onBiological Diversity (CBD) definition of biological diversity

(biodiversity), which defines it as:

The variability among living organisms from allsources, including, inter alia, terrestrial, marine andother aquatic ecosystems and the ecologicalcomplexes of which they are part; this includesdiversity within species, between species and ofecosystems (CBD Art. 2).

Biodiversity exists as a complex web of life forms thatinteract with each other in an ecosystem, in a region orglobally. Biodiversity drives the functioning of ecosystemsthrough countless reciprocal interactions with thephysical and chemical components of the environment.

1.3 The significance of biodiversity

Human civilisation, indeed human life on earth, is ultimatelydependent upon myriads of other organisms with which weall share the planet. Our dependence on biodiversity isabsolute: without it humans would not be able to survive.All food is directly or indirectly obtained from plants andother photosynthetic organisms. Apart from direct benefitsof biodiversity from the harvest of domesticated or wildspecies for food, fibres, fuel, pharmaceuticals and manyother purposes, humans also derive benefit from itsinfluence on climate regulation, water purification, soilformation, flood prevention and nutrient cycling (i.e.ecological services); and the aesthetic and cultural impact isobvious (see Section 3, table 3.1) (Daily 1997; Balmford et al2002) . Biodiversity is thus fundamental for current andfuture social and economic livelihoods.

1.3.1 Sustainability and biodiversity conservationA series of ethical questions is at the core of thesustainability debate – sustainability of what, for whom,for how long and why? (O’Neill 2001). For example, oneapproach might be to seek sustainability in the levels ofwell-being necessary to meet individual preferences.Another might be to seek sustainability in the options oropportunities to meet broader societal needs. The issue ofequality also impinges directly on these questions. Formany people greater equity in contemporary society(intragenerational equity) is an urgent priority. For othersequality of opportunities between generations(intergenerational equity) lies at the heart of thesustainability debate.

In this report we take the view that, as a minimumrequirement, each generation should pass on a set ofopportunities no less than it itself inherited (the so-called‘justice as opportunity’ proposition). Nevertheless, werecognise the fundamental tension betweenintergenerational equity and the humanitarian imperativeof equality here and now.

The Royal Society Measuring biodiversity for conservation | August 2003 | 1

1 Introduction and context

Biodiversity provides substantial socio-economic,scientific, technical, and socio-cultural opportunities.These opportunities give rise to benefits that are based ondiversity within and among species and ecosystems. Theperspective on sustainability adopted here requires thatthese benefits continue to be available to futuregenerations. Thus biodiversity conservation is essential tosustainability.

1.3.2 Direct and indirect benefits from species andecosystemsThe delivery of sustainable development objectivesrequires the efficient and ongoing conversion of solarenergy into useful goods and services. Biodiversity fulfilsthis role, providing a vast range of significant use andnon-use benefits, as well as essential life-support services.Examples of these benefits can be found in Section 3,table 3.1. It is estimated that 40% of the global economyis based on biological products and processes (Packer2002). However, quantifying these benefits is notstraightforward, because they are mostly not captured byconventional, market-based economic activity andanalysis.

A synthesis of more than one hundred studies attemptingto value ecosystem goods and services estimated theiraggregated annual value to lie in the range of about $20trillion to $60 trillion (1012), around a rough average ofabout $40 trillion (Costanza et al 1997, updated to 2000US dollar value). These figures are of similar size to thetotal gross national product of the world (GNP). Althoughsuch estimates should be interpreted with caution, theynevertheless indicate the potential magnitude of theglobal ecological goods and services. At a more locallevel, estimates of the differences in benefit flowsbetween relatively intact and converted versions ofdifferent biomes suggest that, despite private, oftenshort-term, gains, the total economic value of naturalsystems, to society as a whole, is roughly halved followingconversion to farming, forestry or aquaculture (Balmfordet al 2002; Turner et al 2002).

1.3.3 Intrinsic value It can also be argued that biodiversity has intrinsic value inand of itself. One such perspective is that all living thingspossess inherent worth and their interests deserve respect.From this position it is only a short, but fundamental step,to the controversial claim that all living things, individuallyand collectively, possess rights (see Attfield 1981; Taylor1981; Baird Callicott 1995 for discussion).

1.4 Role of science in the conservation andsustainable use of biodiversity

In broad terms enough is already known of the distributionand drivers of biodiversity loss to indicate the scale of theproblem and to guide urgent conservation action.However, there are major gaps in knowledge. For example,

it is not known how much an ecosystem can be simplifiedbut still provide the ecological services on which humansdepend. Similarly, surprisingly little is known about thechanging state of populations, species and ecosystems (seeSection 2.1). Much of our existing knowledge has beendeveloped opportunistically, leaving us with informationthat is too patchy and selective for optimal long-termplanning. Improved knowledge, improved analysis andimproved synthesis at regional and global levels willenhance the effectiveness of attempts to measurebiodiversity for sustainability and conservation goals.

Working together with conservation practitioners,economists, lawyers and other social scientists, sciencehas a key role to play in developing ever more effectiveconservation and sustainability practices.

• Scientific principles must guide the systematic andobjective documentation, analysis and assessment ofbiodiversity as well as trends in the state of wildspecies, populations, habitats, and the ecologicalservices that these organisms and systems provide.

• Both large and small-scale scientific studies areessential in establishing the causes of biodiversitylosses, in identifying priority responses, and inevaluating their effectiveness.

• Scientific analyses (integrated with economic andsociological studies) are vital to developing a thoroughunderstanding of the significance and value ofbiological diversity, and to the development ofmanagement protocols that enable the benefits ofbiodiversity to be delivered sustainably.

• New and innovative techniques, methods anddiscoveries from across the natural and social scienceswill need to be integrated to address the complex andinter-disciplinary challenges inherent in biodiversityconservation.

1.5 International agreements and policyresponses to biodiversity loss

In the late 1980s the United Nations World Commissionon Environment and Development Report, Our CommonFuture (WCED 1987), introduced the concept ofsustainable development as a holistic way of approachingthe interrelated problems of environment anddevelopment. Subsequently, growing environmentalconcern culminated in the United Nations Conference onthe Environment and Development (UNCED), the EarthSummit, in Rio in 1992, from which emerged theConvention on Biological Diversity (CBD). This legallybinding treaty came into force on 29 December 1993 andto date has been ratified by over 186 countries. The CBDhas three main objectives: i) the conservation of biologicaldiversity; ii) the sustainable use of the components ofbiological diversity; and iii) the fair and equitable sharingof benefits arising from the utilisation of geneticresources (see Annex D for further detail).

The Royal Society 2 | August 2003 | Measuring biodiversity for conservation

Prior to the 2002 World Summit on SustainableDevelopment (WSSD) the CBD published the first of itsperiodic Global Biodiversity Outlook reports (CBD 2001).The report provides a summary of the status of globalbiodiversity with an analysis of progress towards the threeobjectives of the CBD. Much progress has been made, butthe persistent challenges of implementation anddevelopment of a comparable and consistent approach tonational reporting, are considerable. Uncoordinatednational reporting has led to some repetition of effort andcan obstruct the connection of disparate informationstrands into a regional or global overview.

The first Global Biodiversity Outlook report emphasised thatthe lack of biodiversity information was one of the majorimpediments, not only for reporting achievements, but alsofor the creation of meaningful targets against whichprogress could be measured. A number of initiatives havebeen launched in response to the lack of information, suchas the Global Taxonomy Initiative (GTI) (see Annex F) andOECD’s Global Biodiversity Information Facility (GBIF) (seeAnnex D for more detail). However inadequate focus andco-ordination of many of these global projects, combinedwith a lack of clearly developed and differentiatedobjectives, create the potential for wasteful overlap. As aresult no single mechanism exists for the easy collection orsynthesis of data, information and knowledge ofbiodiversity, or provisions for making it globally accessible.

Specific global targets for conservation (albeit for a singlegroup of organisms) were set for the first time in April2002, when the parties at the Conference of the Parties(COP 6) of the CBD, adopted the Global Strategy for PlantConservation. This initiative is important, along withother new approaches such as the ecosystem approachfor integrated implementation, but their goals will not berealised without adequate resources and improved co-ordination.

Taken as a whole, the situation with respect tobiodiversity loss has not improved markedly since 1992. Inrecognition of this continuing challenge, a key outcomeof the WSSD in Johannesburg in late August and earlySeptember 2002, was the commitment to:

Achieve by 2010 a significant reduction in the currentrate of loss of biological diversity.

The Summit implementation plan (United Nations 2002)stresses that this would need to be achieved through moreefficient and coherent implementation of the threeobjectives of the CBD alongside the importance ofsupporting initiatives, such as the GTI. However, whilestressing the importance of the CBD and relatedagreements, and despite establishing significantGovernment, NGO and business partnership initiatives, theWSSD did not develop explicit mechanisms to reduce theloss of biodiversity. Indeed no agreement exists on the bestways to measure the rate of biodiversity loss (Annex D).

Measurements and indicators are needed urgently to helpscientists and policy makers assess whether the WSSDgoal is being met. In other environmental arenas, forexample in reducing the use of chlorofluorocarbons(CFCs), agreed measures and indicators have made iteasier to assess the impact of global remediationstrategies. Similarly, measurements of carbon emissionsare the basis for international treaties and reductioncommitments, based on the work of theIntergovernmental Panel on Climate Change (IPCC), aninternationally recognised body comprised of the world’stop scientists and economists in this field.

The Millennium Ecosystem Assessment project (MA),launched in 2001, in which the Royal Society is activelyparticipating, will assess the ability of ecosystems to meetthe needs of people through the provision of goods andservices. The Assessment will seek to supportenvironmental decisions by responding directly to requestsfor information from intergovernmental conventions,governments, industry and society. The status ofecosystems will be assessed at many spatial scales (local toglobal) through a series of assessment reports to bedistributed to policy makers (Ayensu et al 2000). Beingbased on international scientific consensus, reports fromthe MA have the potential to provide an authoritativeassessment of the status of the world’s ecosystems in thesame way as IPCC reports do for climate change.

Without the awareness, participation and commitment ofbusiness and society, policies developed to combatbiodiversity loss and habitat decline will not be successful.Current political best practice is becoming more competentin engaging all sectors earlier in decision-making processes,and also in making better use of the breadth of resourcesand experience found outside of government. For example,many NGO-led initiatives in the area of biodiversitymanagement are now taken up and used by governments.

Industry and various businesses are starting to understandthe concepts of sustainable resource use and the potentialcommercial value of biodiversity, both as a resource and interms of generating consumer goodwill. In some cases,business has taken the lead in creating biodiversity actionplans and developing indicators to measure and limit theimpact of their activities. The global coverage of manybusinesses means that much could be achieved bypromoting conservation as an integral part of their activities,and in some countries the potential beneficial impact ofsuch an approach may outweigh that of nationalgovernments. It is therefore vital that responsibility is takenand that partnerships are forged in these areas.

For governments charged with the responsibility todeliver the biodiversity commitment made at the WSSD,for the evolving work programme of the MA, and for thescientists, conservationists and businesses working to haltbiodiversity loss, improvements in measuring the state ofbiodiversity are essential.


Coordination is also needed among the range of national,international, NGO and business led initiatives underwayto measure and monitor biodiversity, so that practitionershave a common approach and can develop theinformation base needed to achieve global conservationand sustainability objectives.

1.6 Overview of the Report

The primary aim of this report is to provide a scientificcontribution to the development of an internationalconsensus on how biodiversity is assessed. Such aconsensus will be vital if significant progress is to be made– and measured – by 2010. Because there are so manyexisting and potential ways of measuring biodiversity,commenting on them individually in this report isimpractical. Instead, we propose a conceptual frameworkfor biodiversity assessments that seeks to enhanceeffectiveness by clarifying goals, constraints andunderlying assumptions. The framework will also helpensure that the information gathered from individualassessments can be added to the overall wealth ofbiodiversity knowledge. It can also be used to highlightscientific and knowledge gaps, and to identify priorityareas where further research is required.

Many national and international conservation bodies,scientists, individuals, companies, governmental andnon-governmental organisations (NGO) already make

diverse and important contributions in the areas ofbiodiversity assessment, conservation and sustainabledevelopment. We are grateful to the representatives frommany of these organisations who provided valuable inputinto this study. Our initial call for written submission wasmet with 52 responses. This was followed by aconsultation workshop in November 2002 with UK andinternational academics, policy makers andrepresentatives from conservation and NGOorganisations. A further smaller meeting was held inDecember 2002 with representatives from the UKStatutory Conservation bodies. We are also grateful tothose individuals who independently tested theframework and submitted worked case studies. A full listof the contributors can be found in Annexes A, B and C.We hope that this report will be of interest and value to allthose who have contributed to its development, as well asthe many other parties concerned with the sustainableutilisation and conservation of biological diversity.

In the following section, some of the challenges inherentin measuring biodiversity are discussed. We also highlightseveral areas in which we believe progress is necessary toprovide a broader foundation of knowledge on whichbiodiversity assessments can be made. In Section 3 wepresent the framework for measuring biodiversity. This isfollowed in Section 4 by a series of case studies thatillustrates how the framework is used. Finally, in Section5, we offer our conclusions together with a short list ofrecommendations to be implemented now.


2.1 Progress towards measuring biodiversity

Many countries now have National Biodiversity Strategiesand Action Plans, which help monitor status and changesin biodiversity at national scales. For a few groups oforganisms in some places, very good knowledge ofcurrent status and certain recent trends at the specieslevel exist. For example, changes in distribution over timeof vascular plants in Britain in 10 km2 grids are known(Table 2.1; Preston et al 2002). Similarly, for many speciesof British birds, accurate time series of populationnumbers are available (Gregory et al 2002). Somecountries also have useful measures of abundance forother taxa, such as commercially exploited fish species.These kinds of information enable the identification ofthreatened species and provide a basis for policydevelopment and management action.

There has also been good progress with technicalapproaches that have increased the amount and utilityof data relevant to biodiversity measurement. Forexample, satellite data have helped increase awarenessamong the public and policy makers of the scale andmagnitude of habitat loss and transformation.Compilations of observational and specimen data andtheir manipulation and analysis in electronic databaseshave also permitted an improved understanding ofcurrent status and change in biodiversity in somecountries, such as Mexico and Australia. Some progresshas been made at the global level for certain well-known groups of organisms, such as mammals andbirds. These assessments have helped establish large-scale priorities for conservation.

Advances in molecular biology have facilitated geneticanalyses at the level of populations, which are now usedroutinely for a variety of conservation-related purposes.Examples include ex situ breeding, translocation andreintroduction programmes, assessment of diversity fordifficult groups such as protozoans, bacteria and fungi,and monitoring trade and international agreements.

Considerable progress has also been made towardssupplementing subjective assessments of species threatstatus with more quantitative assessments, where thedata are sufficient. An example is the quantitative criteriabeing used by the World Conservation Union (IUCN) toassess global threat status of species, which have beenadopted or modified by many other organisations forsmaller-scale assessments. Considerable advances havealso been made in developing explicit measures ofuncertainty in estimates of species status and trends. Afurther positive development, of recent years, has beenthat the context of biodiversity measurement has

broadened to include more consideration of ecosystem-level approaches to complement the long-standingspecies-level emphasis.

In addition, the ability to capture and utilise largequantities of useful information collected by non-specialists has been enhanced. Examples range fromcensuses of birds, butterflies and dragonflies in the UK tothe use of parataxonomists to collect data in Costa Rica.In the UK and elsewhere, direct involvement of amateurnaturalists in the Biodiversity Action Plan (BAP) process isbecoming more widespread (for example, the BritishBryological Society’s Survey of the Bryophytes of ArableLands; www.rbg-web2.rbge.org.uk/bbs/).

Good data sets have the potential to informenvironmental policy. Two examples illustrate this. Thefirst is the UK Wild Bird Index produced for Departmentfor Environment, Food and Rural Affairs (DEFRA 2002) bythe Royal Society for the Protection of Birds (RSPB) andthe British Trust for Ornithology (BTO). The UKgovernment uses this index as one of 15 headlineindicators for sustainable development. Together theseindicators are intended to form a barometer for ‘quality oflife’ and measure everyday concerns about health,wealth, services and jobs, but also include wildlife.Published since 1998, the UK Wild Bird Index includesdata on 139 common bird species from which regionaland national data sets are derived with separate indicesreported for farmland and woodland species. Farmlandspecies show a deteriorating index, with the majority ofspecies declining (Figure 2.1). Some woodland species arealso declining, for reasons that are very poorlyunderstood. The farmland bird index has undoubtedlybeen influential in the debate concerning the relationshipbetween agriculture and environment.

A second example is the Living Planet Index, a periodicupdate on the state of the world’s ecosystems producedby the World Wide Fund for Nature (WWF) (Loh 2002).The index is derived from trends over the last 30 years inhundreds of vertebrate species, and is calculatedseparately for forest, marine and freshwater species.During the period 1970–2000 the index fell by 37%. Ofthe three ecosystems, freshwater seems to be the mostheavily impacted (Figure 2.2). The power of the LivingPlanet Index is potentially great since it is based onundisputed population level data. Its potential toinfluence decisions is limited by the availability ofpopulation data (which are especially limited in thetropics) and the speed with which the data needed for theindex can be gathered and compiled. Nevertheless, this isan excellent example of how indices could and should bedeveloped and presented (Loh 2002).


2 Progress and challenges in measuring biodiversity

2.2 Challenges

From the above, it may seem that the currently availabledata and assessments of biodiversity are adequate tounderpin future management and policy development forsustainable development. Yet this is far from true.Alarming gaps remain in the data, where it is eitherabsent, very limited or biased. Comparison betweenlevels of detailed knowledge for vascular plants in Britainand the severe lack of information for the globally muchmore significant flora of Madagascar illustrates this point(Table 2.1). However, even in the UK, knowledge aboutthe biodiversity of some groups of organisms, such asfungi, is critically inadequate. In contrast, other taxa suchas birds have been well surveyed in many countries, and

are relatively well known on a global basis, including intropical regions where they are at their most diverse.

To demonstrate further the extent and nature of ourignorance, we list here five categories of informationdeficiency that we consider to be especially crucial.

2.2.1 Poor knowledge of biodiversity at the specieslevelToday the total number of living species named andrecorded has been estimated at around 1.7 to 1.8 million.This number is uncertain to within 5 to 10%, because nocentralised catalogue yet exists.

For some better-known groups, most notably birds and


0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

1970 1975 1980 1985 1990 1995 2000

Year

Freshwater species

Marine species

Forest species

Popu

latio

n in

dex

Figure 2.1 Mean population sizes of Britain’s commoner breeding bird species, 1970–2000; species counts arestandardized to a value of 1 for 1970, and then averaged across all species in a category (data from DEFRA 2002, re-drawn from Balmford et al 2003).

0.5

0.6

0.7

0.8

0.9

1

1.1

1.2

1970 1975 1980 1985 1990 1995 2000

Year

Farmland species (19)

Woodland species (33)

All species (105)

Popu

latio

n in

dex

Figure 2.2 An index of vertebrate populations for forest, freshwater and marine biomes, 1970–2000; the index is basedon 282, 195 and 217 populations, respectively; counts are standardized to 1.0 for 1970 and then averaged across allspecies in a category (data from Loh 2002, re-drawn from Balmford et al 2003).

mammals, global level catalogues exist, but more thanhalf (roughly 56%) of all known species are insects, forwhich few comprehensive catalogues are available. Byone estimate, around 40% of all named beetle speciesare known only from one site, and many from only onespecimen. The amount of current taxonomic effort variesvery widely from group to group, with about one third ofall taxonomists working on vertebrates, another thirdworking on the roughly 10 times more numerous plantspecies, and the remaining third working on invertebrateanimals, which outnumber vertebrate species by at least afactor of 100 (Gaston & May 1992).

Estimates of the total number of species listed are furtherhampered by changes in the concepts of species limitsand by problems of synonyms – the same speciesinadvertently given different names by different people.Estimates of the proportion of synonyms have been put atabout 20% (Hammond 1995; Solow et al 1995), but may

actually be about 40% (Solow et al 1995). For seed plantssynonyms may be as high as 56–78% (Govaerts 2001;Scotland & Wortley 2003). Deliberate accounting for theeffect of synonyms would put the current global total ofdifferent eukaryotic species (including plants, animals andfungi) that have been named and recorded at around 1.5million (May 1999).

The true total number of extant species, as distinct fromthose that have been named and recorded, is hugelyuncertain. One recent survey of surveys – emphasising theevidence and uncertainties – gives a plausible range of 5 to15 million extant species, with the best guess toward thelower end of the range (May 1999). Other estimates rangefrom as low as 3 million to as high as 100 million, theuncertainty in this number being dominated by insect totals.

Knowledge of the totality of species on Earth is thereforevery poor. Even using a ‘low’ estimate of about 7 million


Table 2.1. Comparison of biodiversity information available for vascular plants in the UK and Madagascar – two islandnations with very different levels of biodiversity and associated knowledge.

VASCULAR PLANTS: UK VASCULAR PLANTS: MADAGASCAR

Area: 234, 410 km2 Area: 587,040 km2

Total number of species: 1,403 (+1,600 non-natives) Total number of species: 9,345–12,000

Endemic species: < 4% Endemic species: > 80%

Biodiversity Action Plans in place for critical species No Biodiversity Action Plans in place

Species of conservation concern identified Most species of conservation concern still to be identified

Population genetic data available for species of particular Little population genetic data available for any species conservation concern

Abundance at population level known for species of Very few species for which abundance at population level conservation concern is known

Extinction rate – at national level – known Extinction rate – at national level – unknown

Immigration rate of exotics – known in broad terms Immigration rate of exotics – unknown

Phylogenetic relationships among species being addressed Phylogenetic relationships among species not addressed

Ecology and physiology of some species known in some Ecology and physiology known for very few species. detail through Biological Flora accounts1

Distributions of all species mapped to 10 km squares in Distributions of few species mapped in detail. Herbarium1960 and again in 2002 specimens provide most useful data.

DNA of almost all species in DNA Banks – some sequencing DNA of very few species in DNA Banks – little sequence data of some genes for many species – DNA IDs feasible but not available – DNA IDs not feasible.necessary

Almost all species protected ex situ in seed banks Few species protected ex situ in seed banks

Species inventory complete – new species added only rarely – Species inventory incomplete – new species discoveredgenerally aliens. frequently – generally natives.

Many specimens of most species in preserved collections. Few specimens of most species in preserved collections: many species presumed not represented

1 Biological Flora of the British Isles. Currently published in The Journal of Ecology by the British Ecological Society & Blackwell Science

living species, means there are between three and fourundescribed species for every one currently known. Forthose species that have been described and recorded,knowledge is often very inadequate. For many, only abrief morphological description based on a singlespecimen exists, with no information on distribution,habitat, ecology or past or current population levels.

2.2.2 Poor knowledge of geographical areas andbiomesThe deep sea covers two thirds of the planet yet all of ourquantitative knowledge of the communities of the deepsea floor comes from samples with a combined area of afew football fields (Paterson 1993). Less than onemillionth of the deep sea floor has been touched upon bybiological science and every mission brings forth newdiscoveries, such as entirely new habitats aroundmethane seeps, and the previously unappreciated extentof cold-water coral reefs in offshore waters of countrieslike Britain (Freiwald et al 1999).

Of approximately 500 species of carabid beetles knownfrom southern South America, 30% are known only fromone locality. Inventories in Amazonian forests still yield atleast one new plant species for every 100 specimensprepared, and equivalent figures for west-central Africamay be as high as 5%.

From what is known about certain groups, the tropicscontain many more species than temperate areas, but thismay not be true for certain groups of invertebrates andmicroorganisms. Our knowledge of ecological principlesand trends are very biased towards studies of relativelysimple, temperate communities. Nearly 90% of forestspecies population trend data in WWF’s Living Planet Index(see Section 2.1) derive from studies in temperate regions.

2.2.3 Poor knowledge of current status ofbiodiversityAssessments of conservation status have been made forup to 10% of described species, and an uncertainpercentage of all species (IUCN 2002). The extent of theseconservation assessments is highly skewed across groups,ranging from close to 100% of birds and mammals to lessthan 0.1% of insects. The available species-levelconservation assessments are also skewed geographically,with complete or near-complete coverage for somegroups in temperate regions contrasting sharply with verylow percentage coverage for tropical regions wherebiological diversity is high but scientific capacity is oftennot well developed. This geographic bias produces clearlyanomalous results, such as the US appearing to havenearly three times as many endangered species as Peru.Conservation assessments for tropical plants are availablefor only a fraction of the very large number of species intropical countries.

Even for our closest relatives, the two species ofchimpanzees, current estimates of global population sizes

vary by a factor of two (Butynski 2001). Global estimatesof threatened plant species, extrapolated from endemismdata, range from 22 to 74% (Pitman & Jørgensen 2002),two to four times as high as the figures reported by theWorld Conservation Union (IUCN 1997). The leastthoroughly assessed group of vertebrates is thefreshwater fishes, yet these appear to be among the mostthreatened (IUCN 2002).

Since conservation assessments depend on distributionaland population information, any attempt to increase theproportion of species for which formal assessments areavailable is likely to be impeded by the limited knowledgeof biodiversity (see Section 2.2.1). A particular concern isthat in any group of organisms, common and widespreadspecies are more likely to be discovered and describedthan scarcer ones. Many of the species as yetundiscovered and undescribed have very restricteddistributions and are thus more likely to be vulnerable toextinction through habitat loss. Thus it is expected thatthe proportion of species considered as threatened willrise as our knowledge of the earth’s biota increases.

2.2.4 Poor knowledge of trends in the state ofbiodiversityIn general, knowledge of trends in biodiversity loss ishampered by the absence of reliable baseline data formost groups of organisms as well as habitats. Forexample, global assessments of changes in habitat extentsince the 1992 UNCED Summit have been conducted forjust four out of 14 major biomes (temperate and tropicalforests; mangroves; and seagrass beds). No reliable globalestimates have been published of recent rates of changein the quantity of freshwater swamps, lakes and rivers,estuaries, continental shelf habitat, rock and ice habitats,grasslands, deserts or tundra (Jenkins et al 2003;Balmford et al 2003).

Despite over 20 years of data compilation, recentestimates of rates of loss of one of the biomes beststudied with respect to changes in extent – tropical forests– still vary by a factor of two (FAO 2001; Matthews 2001;Stokstad 2001; Achard et al 2002).

Coral reefs have long been acknowledged as one of theearth’s most important habitats, in terms of both speciesrichness and the delivery of ecosystem services to people.Even so, no reliable global estimates exist of recent ratesof change in their area, though assessments have beendone of the levels of risk from various adverse impacts(Spalding et al 2001).

Of those species whose conservation status has beenassessed, less than 10% are the subject of any sort ofongoing monitoring, even for birds, the best-knowngroup of organisms globally. Good population estimatesexist for less than 30% of threatened bird species (BirdLife2000) and it proves surprisingly difficult to trace the statusof individual species back through time. Routine


reassessments are still dominated by changes due to newknowledge (or past errors) rather than genuine changesin status. For example, between 2000 and 2002, three ofthe 14 changes in conservation status of bird species,were due to genuine alteration in status of populations(IUCN 2002).

2.2.5 Poor knowledge of ecosystem services and thelink to biodiversityServices such as nutrient cycling, carbon sequestration, orstream flow can be some of the key products ofbiodiversity. Available data on the economic value ofecosystems and their individual components generallyrelate to contributions at given locations and points intime. Very little is known about the transformation ofrelatively pristine habitats into degraded forms and theconsequent gains and losses in human welfare. Ofparticular importance to sustainability is the distributionof welfare changes within communities and acrosssocieties as a result of habitat change, but againinformation in these areas is currently deficient. Suchinformation and analysis of the impacts are needed toguide choices about ecosystem management.

2.3 Problems with existing measures and theirapplication

There is a vast range of biodiversity measures already inuse. Such variety is inevitable, given the great breadthinherent in the concept of biodiversity (see Section 1.2),and also the range of spatial, temporal and taxonomicscales across which measurement is useful. Manyexisting measures are well designed and informative.But, very often, existing measures are inadequate forpurposes beyond those for which they were specificallydesigned.

Many measures are only available for a small and oftenunrepresentative subset of species, habitats or areas, orspan too short a time period to provide meaningfulinformation about temporal trends in status. Groups oforganisms that are especially poorly known include mostspecies from the tropical regions and in the deep sea,invertebrates, protists (microscopic unicellulareukaryotes) and fungi.

It is often necessary to make do with data that werecollected for very different purposes than biodiversityassessment. For example, in Europe most data oncommercial landings of skates and rays do not distinguishbetween species. While this may be adequate for providingan overall picture of the health of skate and ray fisheries, ithampers attempts to monitor individual species that maybe endangered (Dulvy et al 2000). Similarly, in using datafrom museum collections to assess geographic distribution,it is important to separate patterns that reflect collectingactivity from those that reveal the underlying diversitydistribution (Nelson et al 1990).

Shortcomings have in turn been compounded by a lack ofclarity about exactly what particular data sets reveal. Forexample, attempts to assess changes in global forestcover have sometimes ignored the differences betweenprimary (undisturbed) forest, commercial forestry andsecondary (regenerating) vegetation. Similarly, catch-per-unit-effort data in fisheries sometimes provide moreinformation about the behaviour and efficiency of fishersthan about the populations of fish that they are catching.

Many aspects of biodiversity, including some of those ofgreatest interest to local stakeholders, are scarcely beingassessed at all. Opinion about the most important aspectsof biodiversity to measure also varies greatly. For example,while many stakeholders are interested in the delivery ofecosystem services, only five out of 300 recent studiesinto their value provide substantial data on the centralquestion of how both direct and indirect benefits changewhen intact habitats are converted to other forms of landuse (Balmford et al 2002).

Attempts to improve the quality, extent, coverage andcoordination of biodiversity assessment and analysis willcost money, yet no large-scale mechanism is in place tosupport such initiatives. For example, many organisationshold biodiversity data relevant to assessing change, butthe incentive to combine these data sets is frequentlyover-ridden by local concerns. Similarly, regional andglobal projects could be delivered through theConvention on Biological Diversity, but this depends onthe political will of the Parties involved. Action throughthe Convention often focuses within national boundaries,making it difficult to organise global databases andoverviews. The value systems in place in much ofacademia can also inhibit data sharing and free access,and incentives to focus on pure rather than appliedresearch may limit the contributions from a potentiallysignificant number of well-qualified experts in universitiesand research institutes.

2.4 Areas where rapid progress is possible

Gaps in the current knowledge of biodiversity status andchange are enormous. In many areas large-scalebiodiversity assessment exercises are hampered by lack ofan overview of existing knowledge, which is often patchyand widely scattered across diverse published andunpublished sources. Synthesis is necessary, not only tomake better use of the existing data, but also to pinpointmore accurately those areas where new data are mosturgently required. Equally, for some areas and groups,existing data are so few that almost any new observationsare likely to yield useful and worthwhile additions to thebody of knowledge.

Addressing the Johannesburg commitment of significantlyreducing the current rate of loss of biodiversity by 2010 willrequire striking a balance between synthesis of information


that is already available and the collection of new data usingsampling methods that are improved in efficacy and scope.This balance is important if future global biodiversityassessments are to rest on a firmer and broader foundationthan current knowledge permits.

2.4.1 Potential progress at the species levelThe lack of authoritative taxonomic treatments for manygroups of organisms limits our ability to identifyorganisms, make inventories and assess change. Existinglevels of knowledge differ widely, but groups for whichtaxonomic needs are exceptionally acute include insects,nematodes and unicellular eukaryotes. For other relativelywell-known groups (e.g. Lepidoptera, aculeateHymenoptera (ants, bees and true wasps), ciliatedProtozoa), including some considered to be goodindicators of overall levels of biological diversity at thespecies level (e.g. plants), a synthesis of existingknowledge is urgently needed to make it more accessiblefor purposes of conservation, measurement, monitoringand management. The results of such syntheses couldinitially take the form of synonymised checklists of specieswith information on distribution, but should evolve to bemore inclusive archives of information containingtaxonomic, identification, and ecological data, as well as,where feasible, preliminary assessments of conservationstatus. Such syntheses would provide nearcomprehensive baselines of current knowledge andwould greatly facilitate future conservation andmonitoring activities at a global scale.

Widespread dissemination of this information and itstransfer to facilitate conservation action in the field iscrucial (Royal Society 2002). Collections and expertise innational museums, on which determinations are based,are often away from the area of need and not readilydeliverable to conservation practitioners in the field.Provision of this information in a low cost format thatcould be used by non-experts, such as keyed-out guidesor compact discs and tapes of birdcalls, could greatlyenhance conservation efforts and the collection of newdata. Up-to-date lists of taxonomic experts, who are ableto respond either remotely or directly, could also behelpful to support situations where advice is required.Assembling data, which at present may be scattered andvery difficult to access, and supplementing it withadditional data (e.g. illustrations, biological data, links toother sites), would provide a universally accessible portalfor systematic information on particular groups (withsafeguards to allow alternative taxonomic hypotheses tobe put forward). For example through a web based‘unitary taxonomy’ as proposed by Godfray (2002).

2.4.2 Potential progress at the geographic levelKnowledge of biodiversity is particularly poor in those veryareas that are most biodiverse – the tropics and especiallythe wet tropics. Central Africa, South East Asia and northernSouth America are among the regions with the greatestbiodiversity where even basic inventories are incomplete for

most groups of organisms and completely lacking for many.Basic improvements in rudimentary baseline knowledge areneeded. In addition, a complete biological inventory of atropical forest area would strongly facilitate ecologicalstudies aimed at understanding how these systems work. Inturn this would highlight potential differences from thetemperate systems on which most ecological knowledge isbased. In Costa Rica, collaboration between scientists andparataxonomists within a comprehensive samplingprogramme makes a near-complete inventory a realisticgoal. Building human capacity in the areas of biodiversity,conservation and sustainable use is also crucial in mostmegadiverse countries that are seeking to measure theirbiodiversity and meet their obligations under theConvention on Biological Diversity.

2.4.3 Potential progress at the habitat levelReliable estimates of recent rates of change in the extent ofparticular habitats are lacking for several major biomesincluding coral reefs and grasslands. Changes incontinental shelf habitats are also still relatively unknown,despite their importance for world fisheries. The stratifiedsampling approach recently applied to humid forests(Achard et al 2002) could be usefully extended to coralreefs, tropical dry forests, temperate forests and wetlands(including freshwater habitats, see worked example inSection 4.2) in order to provide reliable baseline data onextent for these important habitats. Key elements in furtherbroadening the range of habitats being monitored are therefinement of techniques using remote-sensing data, andthe establishment of networks of field sites where suchapplications can be calibrated and supplemented.

2.4.4 Potential progress at the ecosystem levelRational planning for conservation actions might askquestions such as: Which has more effect on thefunctioning of global ecosystems – losing say 25% of allmammal species or 25% of the vastly more numerousinsect species? To address questions like this, moreknowledge is needed of the organisms that comprisebiodiversity and a clearer understanding of howbiodiversity contributes to the functioning of ecologicalsystems. Gaps in our knowledge of ecosystem services areeven more extensive. Our limited understanding of nutrientcycling, climate regulation and the delivery of freshwaterare of particular importance in this respect. Data on howthese services alter as ecosystems are converted for otheruses, how they vary over time and how they interact arealso lacking. Collation and synthesis of recent and historicaldata in these areas could result in substantial improvementsin understanding and provide the baseline against whichfuture changes could be monitored. The production ofglobal and regional maps of major ecosystem serviceswould be an important achievement.

2.4.5 Potential progress on types of dataAcross all groups of organisms, geographical areas andscales of study, certain types of data that are of greatpotential utility, are generally scarce or completely


lacking. Chief among these are time series data(systematic observations repeated at documentedintervals), data on abundance and how it varies, anddistribution data, especially where apparent absence andeffort invested in survey are also recorded (see workedexamples in Section 4). Databasing and geo-referencingthe data currently held in museum collections, forestry andwildlife department archives, expedition logs, and even thenotebooks of amateur naturalists, would mobilise muchhistoric information at relatively low cost per observationand provide an important baseline and context for theinterpretation of current and future observations. Synthesisof this data is essential and will quickly reveal key gaps inknowledge. These gaps should then be addressed by thedevelopment of realistic new programmes capable ofdelivering substantial improvements in knowledge ofotherwise poorly understood geographic areas, habitatsand groups of organisms. Cost-effective ways ofaddressing these gaps could be driven by technicaladvances in other fields (for example the internet,

molecular biology, remote sensing), they could also includemore effective deployment of amateur naturalists andother non-specialists to help document distribution,abundance and variation over time.

2.5 Refining assessments of biodiversity

Expanding and improving biodiversity assessment isessential if a clearer understanding of the changing stateof ecosystems, species and populations is to be provided,and protocols put in place with which to assess whetherthe Johannesburg 2010 commitment is met. Assessmentsmust be scientifically based to improve the accountabilityand rigour in the reporting of biodiversity trends andstatus. As the conservation and sustainable use ofbiodiversity become more prominent, nationally andinternationally and in political and legal debates, data setsand methods are needed that are well founded and thatwill withstand intense scrutiny.


3.1 Introduction

One problem in the measurement of biodiversity is thatexisting measures are often not well suited to the purposesof those making policy decisions, or measuring the effect ofpolicy. To address this issue we have developed a framework(Figure 3.1) for the assessment of some aspect ofbiodiversity. The framework does no more than makeexplicit best science practice, but is intended to be used byall scientists who measure biodiversity – whether academic,industry based, governmental or NGO – as well as thosewho commission and use the information generated.

3.2 Structure of the framework

The framework consists of a series of linked activities thatcomprises the assessment of some aspect of biodiversityor ecosystem function. The framework process can bedivided into three main stages: a scoping stage, a designstage and an implementation and reporting stage. Arational approach to any of the activities in the frameworkdepends upon the outcome of at least one other activity.Sections may have to be repeated if feedback fromactivities downstream indicates that changes are needed.

The framework can be regarded as a conceptual processthat can be applied to all levels of biodiversity. A series ofcase studies in the following section (Section 4)demonstrates its use for terrestrial, freshwater andmarine systems, and at the ecosystem, species andpopulation levels. It is also relevant to long-termmonitoring programmes and in emergency situations,such as an oil tanker disaster. Equally, in situations wherea potential disaster can be anticipated, the frameworkcan be used to develop a damage limitation programme,where potential risk areas are assessed in advance. Forexample, programmes of this kind are already carried outby oil companies along important shipping lanes.

Clearly, the quality of an assessment will be enhanced asmore time and effort is spent on each stage. At certainpoints an extra investment of time could have significantbearing on later activities and may increase the amountof information the study yields. For example, wideningthe range of interested parties at the scoping stage mayalter what needs to be measured. Additional time spentidentifying the assumptions and limitations used in thedesign stages, and making them explicit in the reporting,will increase the robustness of the assessment. Where arapid response is needed some of these activities couldbe quite quick, such as through a phone call to a few keystakeholders, rather than a full consultation process.

Routine use of the framework, in all situations, will helpensure stakeholder involvement and that measures are

fit for the purpose to which they are being applied. Itwould also help to identify weaknesses in currentapproaches, as well as highlight major science andinformation gaps.

3.3 The scoping stage

3.3.1 Context, stakeholders and interested partiesBiodiversity is embedded within human ecological andsocial systems. Those people involved with, and affectedby, the biodiversity of such systems are referred to here asstakeholders. The potential effects of decisions aboutmanagement of biodiversity within a system may varyamong stakeholder groups. For example the health,welfare, intellectual, recreational, spiritual and financialinterests of different groups of stakeholders may beaffected in different kinds of ways. Scientific assessmentsof biodiversity occur because some subset of stakeholders(interested parties) wish to have certain information.These wishes should influence the way in whichassessments are carried out. Other stakeholders, whoseinterests may be greatly affected by the outcome of theassessment, may not be among the initial set of interestedparties. This may be because they are not interested inencouraging a scientific study of the system, they do notknow about it, or because they do not have the resourcesor political influence to become involved.

A major objective of biodiversity assessments is to informpolicies that seek to reduce the rate of biodiversity loss aspart of a sustainable development strategy. However, thisbroad policy-driven objective has to be translated intopractical activities via a decision-support mechanism. Thiswill place the biodiversity component of a system withinthe broader context of relevant social, political, economicand scientific knowledge.

The Driver-Pressure-State-Impact-Response (DPSIR)approach (Turner et al 1998) is a tried and tested scopingprocedure. It recognises that at the root of environmentalchange are socio-economic drivers, such as exploitation ofwild natural resources, intensification of agriculture,urbanisation, and tourism development. The cumulativeeffect of these, together with climate change and otherfactors, is to exert pressure on environmental systems andcause changes in the state of the environment (biological,geochemical and physical) including changes and losses inbiodiversity. In turn those changes have impacts (positiveand negative) on human welfare. However, these changesin welfare affect different, often competing, stakeholdersin different ways and therefore frequently stimulatedebates about equity, values and ethics. These in turn canspur political systems to provide legal and managementregimes that seek to control driving forces and pressures.The result is a dynamic cycle with feedback loops.


3 A framework for measuring biodiversity


3.3.2 Identifying valued attributes, aims andobjectivesAt the outset, the object of the assessment, and theattributes of that object that are of interest, need to bedefined, both by the interested parties and those carryingout the assessment. Differences in timescales that areinherent in the stakeholders’ values will also be identifiedat this stage. Some examples of valued attributes andobjects are provided in Table 3.1. Interested parties willoften differ in the desired state of the attribute and thelevels of precision and accuracy they require from theassessment. The examples in Table 3.1 illustrate someways in which different interested parties might value thesame attributes. Care is then needed in defining themeasurements to be used (Section 3.4.1) and designingthe sampling strategy (Section 3.4.3). It is essential that

interested parties are clear about what the assessmentcan and cannot provide. A scientific assessment is onlylikely to satisfy the requirements of interested parties if itfirst establishes what their interests and values are.

For every attribute of interest, decisions must be madeabout how best to capture information about its state in aquantitative form. This involves the identification ofmeasurements that are feasible to obtain at the spatialscale needed and within the time available. Before themeasurements are defined precisely (Section 3.4) thereshould be consultation with interested parties, beginningwith the following straightforward questions.

• What do you care about?• What questions do you want the assessment to answer?

Table 3.1 Examples to illustrate the meaning of the terms interested party and valued attribute as applied to the value ofbiodiversity. Direct use benefits are valued through the market, whereas indirect use is valued by observation.

Valued attributes

Global species richness and thelocation, abundance and rangeof species as resources fordocumenting and understandingthe evolutionary process

Global species richness and theabundance, range extent andviability of species

Volume and reliability ofstreamflow (in part controlledby the moisture collection andretention properties of forestvegetation) as a determinant ofwater availability to people.

Volume of timber that can beextracted

Reliable ongoing source ofprotein and income

‘Beautiful’ and abundant coraland fish species

Range and population size(desired state – at leastmaintained well above theminimum viable level)

Range and population size(desired state – below the levelthat results in significantdamage to crops and grass bygrazing)

Type of value

Option value i.e. conservationallows time for new scienceand information to bediscovered

Existence value or non-consumptive use value

Forest vegetation – direct usevalue

Flood protection benefits interms of costs avoided –indirect use value

Direct use value

Direct use (consumptive) value

Direct use (non-consumptive,amenity) value

Existence value

Negative value above a givenpopulation size

Interested parties

Evolutionary biologists

People who like ‘wild’ nature

Local people whose health andlivelihoods depend upon areliable fresh water supply

Commercial foresters

Local fisheries

Marine ecotourists

Conservationists

Farmers in the winter range

Object

Diversity of life onEarth

A forested rivercatchment

Coral reef

The world populationof an arctic-temperatemigratory goosespecies


Reporting to interested parties (3.5.3)

Identifyingvalued attributes, aims & objectives

(3.3.2)

Discussion among assessorsand interested parties

Devising the sampling strategy(what / how / where / when)

and assessment of efficiency, accuracy,precision and utility of output

(3.4.3)

Biodiversity assessmentState of valued attributes

Identification and state of pressures and drivers(3.5.2)

Analysis(3.5.2)

Data gathering,checking and storage

(3.5.1)

Pilot project totest/develop

methods(3.4.3)

Choosingmeasures

(3.4.1)

Modelling the system

(3.3.4)

Assessing existingknowledge (3.3.3)

Scopingstage(3.3)

Check measureswith interestedparties

Check that the expectedoutput will satisfyinterested parties

Project/study/design stage(3.4)

Implementation & reporting stage(3.5)

Adaptivemanagement:

regular analysisof data to check

adequacy(3.5.1)

Publication in peer-reviewed literatureadds to knowledge

Data made availableon open access

database, if possible

Policy and researchrelevant information

Figure 3.1 Framework for biodiversity assessment, showing the various conceptual stages necessary for assessing anelement of biodiversity

• What are you likely to do with the results?• When is the result of the assessment needed?• Is it envisaged that the assessment will be repeated in

order to measure change through time?

Discussion of these questions is likely to lead to moredetailed queries such as the following.

• Is there a need to quantify as fully as possible the totalabundance and distribution of an attribute of interest,or will a reliable estimate based upon sampling beacceptable? For example, a full census of anendangered vertebrate population may be needed inorder to define and justify the designation of aprotected area, but otherwise a reliable estimate basedupon extrapolation from samples within the rangemight be adequate.

• Where the attribute of interest is difficult to measure isit acceptable to substitute measurements of anindicator? For example, changes in the species richnessof a single taxon may be shown to be a good proxy forchanges in species richness more widely.

• Is it necessary to have absolute measurements of thestate of a particular attribute or is an index that iscorrelated with population size and that will allow theassessment of change over time acceptable? Forexample, is an estimate of the total number ofbreeding age adults in a fish population required, or isan index such as catch per unit effort acceptable?

• What level of precision is needed and is thereasymmetry in the level of uncertainty that can betolerated? For example, for an endangered speciesclose to its minimum viable population size, theamount of uncertainty about current population sizebelow the best estimate is more important toconservation managers than uncertainty above it.

• What is the desired state or set of states of theattribute of interest? Asking this may help to define therequired accuracy of the estimates (see above). Forexample, species richness estimates for protectedareas can be based on species lists derived fromopportunistic sightings by tourists, but moresystematic surveys with records of sampling effort willbe needed to quantify changes over time.

• Is assessment of the valued attribute the only thingthat the interested parties require? This is unlikely if theDriver-Pressure-State-Impact-Response (DPSIR)scoping process has been fully implemented. Howeverthe value of the assessment could be increased atmodest extra cost by also measuring aspects ofenvironmental pressures and their socio-economic,biotic and abiotic drivers. For example, in remotesensing surveys of deforestation rates it may beefficient to expand the study to include a survey of landuse in the places converted from forest as part of thesame research programme.

3.3.3 Assessing existing knowledgeExisting knowledge may come from a variety of sources,

such as previous studies, the interested parties and fromthose carrying out the assessment. Where the knowledgecomes from previous scientific studies of the system or fromstudies of other systems, if the similarities are sufficientlystrong, judgements have to be made about the extent towhich it can be relied upon as a guide to the presentassessment. Such judgements should be identified asassumptions. Even if the information is derived from thesame system, it may have become an unreliable guidebecause of changes over time. For example, inferencesabout the sustainability of a bushmeat harvest from atropical forest may be made from current measurements ofthe volume of the harvest and densities of hunted species,but they may also depend critically upon pre-existingknowledge of the geographical distribution of consumersand markets and the catchment area used by hunters. Ifbushmeat from the study area begins to be exploited bynew hunters and traded in previously unknown marketsthen estimates of the harvest, and conclusions aboutsustainability, would become unreliable.

Existing knowledge is likely to be used to make samplesurveys efficient. For example, to measure changes in thethreat status of species, prior knowledge of geographicalrange and previous assessments of specific threats can beused to design efficient surveys for multiple speciessimultaneously.

Interested parties frequently have some knowledge of theobject to be assessed as well as being able to providesome insights about factors that affect the state of theattributes. The initial stages of an assessment ofbiodiversity may (and often should) involve formal reviewand critical testing of existing knowledge alternating withdiscussion with interested parties of the conclusionsdrawn from it. During this process of rigorousexamination, interested parties may change theirpriorities about which attributes of a system they value.

3.3.4 Modelling the systemConsideration of existing knowledge, together withvalues and requirements of interested parties, provides abackground to prepare a conceptual model of the system.Even a simple model will be of great value as it forcesassumptions to be recognised and made explicit and canhighlight defects in the reasoning that links the attributesof interest with the measurements that can be made.Without having collected any data developing a modelmay seem premature, but there is usually someinformation available from which it is possible to providesome level of quantitative or qualitative analysis.

The nature of an appropriate model will depend upon thesystem and the attributes to be measured. For example,suppose the object of the assessment is a measure ofspecies diversity and the attributes to be measuredconcern the effects of habitat fragmentation at a globalor continental scale. The underlying model might includeknown latitudinal trends in species numbers, range sizes


and levels of endemism. This knowledge, in turn, wouldinfluence the choice of study sites and the scale ofanalysis required in different regions. Similarly, in the caseof an animal or plant species being monitored forpopulation size and trends over time, there may only becertain life history stages for which individuals can becounted and certain areas that can be covered by surveys.An appropriate model might be a matrix of the number ofindividuals in various life history stages over a period ofseveral years. For a bird species, it may only be possible toobtain reliable counts for territorial males during thebreeding season. In this case having a model of the birdpopulation would lead to an enhanced understanding ofwhat a feasible assessment could achieve, and what it isnecessary to assume or find out in order to interpret theresults. Counting only the singing territorial adult malebirds either assumes that this provides an acceptableindex of the total population size, or that studies must beinitiated to work out the relationship between numbersof males and the size of the whole population.

If the assessment aims to measure aspects of thepressures and drivers that are thought to affect thebiodiversity attributes of interest, it is important that themodel of the system includes the relationships amongthem. These relationships can be expressed using a Driver-Pressure-State-Impact-Response (DPSIR) framework, evenif this is only possible in conceptual form.

Making a model based upon whatever is known aboutthe system forces the assessor to make explicit the linksbetween the attributes they are interested in and themeasurements that it is feasible to make. This will helpto identify the limitations of the scope of theassessment.

The process of identifying valued attributes, assessingexisting knowledge and developing a conceptual modelof the system, comprises the Scoping Stage. This shouldlead to the identification of a clear set of aims andobjectives for the assessment. In turn, these are likely tobe modified and refined during later steps in theframework as practical constraints become apparent.However, all but the most trivial of such changes shouldbe referred back to interested parties for discussion.

3.4 The design stage

3.4.1 Choosing measures

3.4.1.1 Advantages and disadvantages of differenttypes of measuresWith a model in place, the specifics of what to measurecan be considered. The model can also be extended toinclude the relationship between the measurement andthe true state of the valued attribute. Table 3.2 provides asimplified scheme of four broad categories thatencompass many common measures.

Extent of habitatThe first category of biodiversity measurementsestablishes the extent of large-scale ecosystems orhabitats. Knowing something about habitat area may bevaluable in itself, and when linked to knowledge aboutrate of change in habitat area can provide an indirectmeasure of the loss of populations and species associatedwith that habitat. It may also provide a basis for estimatesof certain kinds of ecosystem services.

Large-scale habitat measurements have been aided greatlyby advances in remote sensing and GIS software. Forexample, the extent of forest fires in Southeast Asia andlosses of primary forest globally have been monitored withsatellite images. However, the degree of resolution of thistechnique is still not adequate for many purposes, such asmonitoring more subtle habitat degradation ordistinguishing between regenerating forest and plantations.

Ecosystem processes and changes in functionThis broad category of biodiversity measures seeks toassess ecosystem processes and changes in ecosystemfunctioning. Often this is approached by the use ofproxies. For example, key aquatic plants can be valuableindicators of eutrophication of freshwater bodies.Surveying these aquatic macrophytes may be more rapidand cheaper than full-scale monitoring of water qualityand biological composition. However, measurements ofindicator species may make comparisons across habitattypes difficult.

Attempts to assess ecosystem processes also potentiallyinform improved understanding of the goods and servicesprovided by ecosystems. For example, understanding tidaland other hydrological effects is important inunderstanding how saltmarsh ecosystems work, andmeasurements of wave heights at saltmarshes are used toprovide a direct indication of the services provided bymarshes in reducing the risk of sea defences beingbreached. Measurements of services provided byecosystems can also be used to provide rapid andinexpensive indications of the state of habitats, thoughthey often have low sensitivity to changes in abundanceof specific taxa.

Lists and distribution mapping of taxaCounts of species (species richness) are probably the mostcommonly used surrogate for overall biodiversity at bothlocal and broader scales. The species level is an acceptedstandard, because species are the most familiartaxonomic unit for scientists, the public, and policymakers. Information on the presence of well-knowngroups of species such as birds and flowering plants isavailable for many areas and time periods from therecords of visiting naturalists, as well as more formalsurveys undertaken by governments and NGOs. The wideavailability of these data has led to their common use.However, such data also have drawbacks that are notalways fully appreciated.


In certain cases, levels in the hierarchy of biologicaldiversity other than species, can be used to indicatebiological diversity across sites. For example, geneticvariation within species is an important component ofbiodiversity in its own right. Recent advances in moleculartechniques are also allowing genetic assessments as aproxy for species diversity when taxa are difficult toidentify individually, as is the case for manymicroorganisms. Other assessments consider highertaxonomic levels, such as the number of families, whichare sometimes easier to identify and count than species.This may be a valuable approach if these higher taxaadequately capture phylogenetic distinctiveness.

Measurements based on mapping and listing particulartaxa often categorise species according to their rarity orthe extent to which they are threatened with extinction.

Measurements of rarity and extinction risk are extremelyuseful, but measurements of actual species extinctionrates are a poor way to monitor biodiversity loss. Mostextinctions have a long ‘tail’, whereby the species maypersist temporarily at low numbers with a negligiblechance of recovery and a severely diminished role in theecosystem. Measurements of extinction rates thereforeprovide a very crude indication of biodiversity loss,which reveal little about short-term changes in pressuresand cannot provide the kind of ‘early warning’ thatcould lead to successful conservation interventions.Measurements of extinction rates at the species levelalso suffer from other problems, including: i) most of theworld’s species have yet to be described, ii) estimates ofextinction rates, such as those derived from habitat loss,have significant associated uncertainties, and iii) formost groups of organisms background extinction rates


Advantages

Remote sensing can provide large-scale assessmentand measurement of recent trends.

Can be used to estimate rates of speciesendangerment based upon species-arearelationships.

Links to delivery of biogeochemical ecosystemservices are potentially strong.

Potentially strong links to ecosystem services anddirect relevance to material needs of humans.

May provide deeper understanding of ecosystemfunctioning that is helpful for management.

May provide quick and efficient compositemeasurement of state of habitats and taxa.

Most commonly available information from existingrecords at a variety of scales.

Provides easily understood data about diversity,especially through simple species counts.

Information is often directly relevant to speciesprotection legislation.

Information frequently highly relevant to theidentification of priority areas for conservation anddefinition of their legal status.

Useful historical data may exist from distributionatlases, museum specimen collection localities andother sources.

More sensitive than measurements based upondistribution alone.

Data potentially can be aggregated across species toprovide a composite index.

Clear relationship with some values and services.

Disadvantages

Objective delineation of boundaries of habitat orecosystem sometimes problematic.

Often difficult to identify ecologically importantsubdivisions of broad habitats such as forests.

Degradation of habitats and loss of component taxaof ecosystems may not be detected.

Reliability may vary across habitat types.

Methods require careful validation before wideapplication.

Relation to fates of taxa often uncertain.

Difficult to ensure that absence of records meansthat the taxon is absent.

Difficult to generalise: most taxa are undescribed andknown taxa may not be a representative sample.

Taxonomic distinctions difficult to make or of limitedvalidity for some groups (often a function of asexualreproduction).

Sometimes difficult to factor out survey effort andmethods; apparent changes in distribution may beartefacts of changes in effort or observer skill.

Changes in range, even when accurately recorded,may be insensitive as an index of overall abundance.

Labour intensive and expensive to collect the data:not practical for many taxa.

Methods and interpretation are taxon-specific.

Limited historical information compared withdistribution measurements.

Some applications require expensive acquisition ofseparate data on abundance of age/sex/size classes.

Measure

Extent ofhabitat

Ecosystemprocesses andchanges infunctioning

Listing anddistributionmapping oftaxa (especiallyspecies andsub-species)

Population sizeof selectedspecies

Table 3.2 Four broad categories of biodiversity measures, with examples of advantages and disadvantages

for comparison with contemporary losses, areproblematic to estimate.

Population size of selected speciesA fourth group of measurements concerns theabundance of individuals and changes in populationnumbers of organisms. These measurements are oftenaggregated across species to produce compositeindicators of changes in particular regions or habitats.Population surveys can provide sensitive indicators of thestatus of particular species under study, but they may alsobe expensive to acquire. It may also be difficult toextrapolate the conclusions to other species withdifferent ecological requirements.

3.4.1.2 Deciding among alternative measurementsSelecting appropriate measurements depends first andforemost on the object of the assessment and theattributes of interest. Some attributes can be measureddirectly, but others cannot and in these cases reliance mayhave to be placed on indirect (proxy or surrogate)measurements that are in some way correlated with thoseof direct interest. In this case, the precise relationship ofthe indirect measures with those measurements of directinterest will need to be determined, if necessary by pilotstudies. Unfortunately, such relationships are often quitevariable and context dependent. The resource demandsinvolved in establishing the usefulness of an indirectmeasure may sometimes be so high that they negate itsbenefits.

In some cases, measurement may not be currentlyfeasible. For example, the taxonomy may be in such astate that it is impractical to estimate species diversity inthe field. Similarly, there may not be enough informationknown to be able to effectively measure complexecosystem functions such as nutrient cycling. Solvingthese problems will require a concerted effort from thescientific community.

Among the many considerations in selectingmeasurements two are especially important. First,alternative measurements might vary in usefulness andfitness-for-purpose. Second, alternative measurementsmay vary in cost and difficulty. These two factors need tobe considered together; the best measurement is the onethat is most useful in relation to available resources. Anillustration of the decision process around these factors isshown in Box 1.

This example brings out the value of having some priorsense of the qualitative relationships between thevarious variables indicated in Figure 3.2, as well asconsidering the total amount of time and resources thatwill be available for the assessment and the results thathave already been obtained by previous efforts. Ifrepeat surveys to monitor change are anticipated, thenthe ease with which a given assessment could beexactly reproduced in future would need to be

incorporated into the evaluation of the usefulnessversus completeness function. In the example givenhere, this might reduce the advantage of the directground surveys (for which new staff might need to betrained) over the more automated processing ofsatellite imagery.

This model really does no more than formalise thedecisions to use informative data sets that can mostreadily be gathered. Yet it is important to make thisexplicit. Often the temptation is to gather the easiest toobtain data despite their limited bearing on the issue(failure to consider the relationship in Fig. 3.2a), or togather new data when strategic additions to existing datacould be more efficient (failure to consider therelationship in Fig. 3.2c). A dataset that can be readilygathered or completed may or may not be the bestoption, depending on the circumstances.

3.4.3 Devising the sampling strategyThe next step is to use the model, imperfect as it may be,to develop a sampling strategy that specifies what tomeasure, when, where and how.

A fundamental decision is whether to make estimates ofattributes of interest by extrapolating frommeasurements made upon a sample, or whether tosurvey or measure them in their entirety. Assessments ofplant and animal populations routinely use counts madefrom surveys in sample areas, such as quadrats or linetransects, which are then extrapolated to estimate thetotal population size. If characteristics of surveyed andunsurveyed areas that are correlates of populationdensity, such as vegetation type or topography, can beobtained, then these can be used to improve theprecision of the estimates and to model the distributionand abundance of the species outside the surveyedareas. Line transect methods for estimating thepopulation of particular species and some approaches tosurveying species richness extrapolate not only from thesurveyed area to other areas, but also within thesurveyed area to allow for incomplete detection ofindividuals and species within it (Buckland et al 1993;Boulinier et al 1998). These methods recognise that notall individuals or species in the surveyed area arenecessarily detected by the fieldwork and use modelsthat describe the probability of detection of individuals orspecies by the observer as a function of distance from thetransect line, time spent searching or some other proxyfor observer effort.

Extrapolations from a sample are often the only feasibleway to estimate large or widespread populations or thesize of large assemblages of species. The accuracy ofassessments based upon extrapolation from samples islikely to depend critically upon the realism of theassessor’s model of the system. Stratified randomsampling, based on prior knowledge, can greatlyincrease the precision of the resulting estimates.



As well as deciding where to survey, the questions ofwhen to survey and on how many occasions also need tobe considered. If trends over time are to be measuredcareful consideration must also be given to the extent towhich attributes fluctuate over short time intervals orshow systematic trends over longer periods. Assumptionsin the model about the causes of variation over time in theattributes being measured are important here. Forexample, trends in the population size of a large-bodiedmammal with a high mean annual survival rate and lowaverage fecundity would probably be revealed bysampling at intervals of several years. Much morefrequent sampling will be required for a small-bodiedmammal with strong intra- and inter-annual populationvariability.

When measurements of the attributes of interest havebeen identified, and a provisional sampling strategy hasbeen developed, an attempt should be made to use themodel of the system to simulate the likely outcomes ofthe assessment. This can be attempted based onplausible guesses about the true state of the attributesbeing measured. In a preliminary way, the questionshould be posed: is it likely that the study will yieldmeaningful results that will help to answer thequestions posed by the interested parties? It is alsoessential to consider the likely confidence intervals forall the parameters used, so as to determine whether theinevitable uncertainty in the results, arising both fromthe model specification and the statistical analysis (orsampling), are such that the results are likely to be toovague to be useful to interested parties. Considerationshould also be given to any technical aspects of thedata collection methods that are in doubt.Development work on the data collection methods or apilot survey to check that they work as envisaged mightbe needed.

3.5 The implementation and reporting stage

3.5.1 Data gathering, checking and storageThe collection of data is dictated by the sampling strategy.Several principles of data collection apply to a wide varietyof biodiversity assessments (listed below). Guidedprimarily by the requirements of effective analysis andreporting (Section 3.5.2 and 3.5.3), the data will also addto existing knowledge. Data stored with details of exactlyhow it was collected will therefore increase its valuecompared to other similar data for which suchinformation is not available.

• Ensure that people collecting data are adequatelytrained and follow a common protocol for collectingand recording information.

• Keep raw data for checking and re-interpretation.• Store data in its most disaggregated form.• Record precise locations of field study areas.• Record sampling effort and who collected the data.

• Record both presence and apparent absence indistribution and abundance assessments.

• Ensure that checks are carried out to keep errors inrecording and data storage at an acceptable level.

• Where possible, collect any additional, low cost datathat may be useful later (such as simple data on driverswhen collecting information on an object’s state).

• Review progress regularly to check that the data beingcollected will address the questions originally posed.For example, do the ground truth checks confirm theaccuracy of the distinction between regeneratingforest and plantations in a remote sensing survey offorest extent? If not, there may still be an opportunityto revise the methods. Regular review will also allowthe assessment to be adapted to respond to anyunforeseen circumstances, such as the appearance ofalien species.

3.5.2 Analysis, assessment and reportingThe details of the analysis and reporting of theassessment will be specific to each particular case, but wemake some recommendations on issues of widesignificance.

• The sensitivity of the conclusions of the assessment tothe underlying model and its assumptions should beexplored and reported clearly. Where appropriate,alternative conclusions arrived at from plausiblevariants of the model should be reported.

• The results of the assessment should be used to updateand improve the underlying model of the system as abasis for future research. Defects in the modelunderlying the assessment should be identified clearlyand remedies suggested.

• The survey design, the procedures used in sample areaselection, and the fieldwork and analysis protocolsshould all be described in sufficient detail to allow thesurvey to be repeated. This is especially important forcomplex, semi-automated techniques such as themapping of habitats.

• Where possible, the raw data from the assessmentshould be available to other researchers for alternativeanalyses.

• Precise survey localities should be archived so that thestudy could be repeated at the same localities ifnecessary.

• The results of the assessment should, wherever possible,be published in the peer-reviewed literature. Where thisis not possible, an attempt should be made to subjectthe outputs to other forms of external review.

3.5.3 Reporting to interested partiesScientific assessments of biodiversity should be reported tointerested parties in ways that minimise the scope formisinterpretation of the results. Reporting results in a formaccessible to scientific colleagues is important, but is unlikelyto be effective in communicating to interested partieswithout further effort. For example, when fisheriesbiologists communicate with each other they refer to F, the


Box 1: Deciding between two measures

Suppose a team is trying to estimate the area of old-growthnative forests in a region. They could either send groups ofresearchers out to map hundreds of sites on the ground, oruse satellite imagery. But satellite imagery, though developedusing some ground truth information, may not be entirelyreliable in distinguishing between younger forests, nativeold-growth, or plantations of exotic species. If this reliabilityproblem could not be overcome the team might assume thatusefulness is determined mainly by the accuracy of the twodifferent methods for making the estimates. But the decisionmay rest on other factors.

The expected usefulness of the results can be plotted againstthe completeness of each of the alternative researchprogrammes (Fig 3.2a). Both methods become moreaccurate and therefore more useful as they becomecomplete, though the rate of improvement declines as thetask nears completion. This would be expected if bothmethods produce more accurate estimates based uponprogressively larger samples. However, the satellite-basedestimates are only half as useful as the direct ground surveys,even when complete, because of their lower reliability.

Figure 3.2a Expected usefulness of results against thecompleteness of alternative measurements

Similarly, the completeness of the research programme inrelation to the effort (cost or person-years of effort)expended can be considered (Fig 3.2b). The advantage of thesatellite-based method is that it can be accomplished withmuch less effort than the direct ground-based survey.

Combining these two approaches it is possible to estimatethe relationship between usefulness and effort expended(Fig 3.2c). Suppose 20 units of effort were available and thatneither measurement approach had yet been started. FromFig 3.2b it is clear that the satellite-based survey could bevirtually completed with this level of resources, but that thesame expenditure would only complete a small fraction ofthe direct ground survey work. In this case the completesatellite-based survey obtained with 20 effort units is moreuseful than the incomplete ground survey, despite the higherreliability of direct land-based survey data (Fig 3.2c).

However, had 40 units of effort been available theusefulness of the results of the direct survey would havebeen higher. In this example, if the effort available issmall, the indirect proxy is favoured, whereas if moretime and resources are available, the direct surveymethod is preferable. Intuition alone would notnecessarily lead to this conclusion, because the answerdepends on the particular functions that relateusefulness to completeness, and completeness to effort.Different outcomes can be obtained with differentfunctions.

Figure 3.2b Completeness of the research programme inrelation to the effort or money expended

Figure 3.2c Estimated relationship between usefulness andeffort expended

An additional consideration in choosing among alternativemeasurements is the amount of information that is alreadyavailable. For example, suppose that 20 units of effort hadalready been spent on the direct ground survey before theoption of satellite-based methods became possible, and afurther 20 units of effort became available. Examination ofFig 3.2c indicates that continuing to spend the newresource on the ground survey will yield a more useful resultthan switching to the satellite-based method. However, ifonly 5 units had already been invested in the ground survey,a more useful outcome would be obtained by spending thenewly available 20 units on initiating the satellite-basedsurvey.

0

20

40

60

80

100

0 20 40 60

Ground surveys

Satellite surveys

80 100

Usf

ulne

ss

Completeness (%)

0

20

40

60

80

100

0 20 40 60 80 100

Effort

Com

plet

enes

s (%

)Ground surveys

Satellite surveys

0

20

40

60

80

100

0 20 40 60 80 100

Use

fuln

ess

Effort

Satellite surveys

Ground surveys

instantaneous rate of fishing mortality. But when theycommunicate with fishers and politicians, they usuallyconvert F to percentage mortality, which is more meaningfulto non-scientists. Reporting to interested parties should beas straightforward as possible and specifically address thepossible policy responses under consideration. Interestedparties should be informed of the relative strength of theevidence for and against a particular policy option, and theextent to which this could be altered if untestedassumptions prove erroneous.

Reporting of the results of an assessment to interestedparties will often lead to proposals to repeat theassessment or to undertake some connected piece ofresearch. This should involve re-discussing valuedattributes, aims and objectives with interested parties.The completed assessment is now part of the existingknowledge to be taken into account in designing the newstudy. The DPSIR framework may well be modified andnew stakeholders may be identified and recruited asinterested parties. With the new information it will usuallybe possible to identify important questions not identifiedwhen the original study was being planned. Experiencemay also suggest changes to methods that will improveefficiency or accuracy. Such changes in objectives andmethods should be considered carefully and with greatcaution. If the measurement of changes over time isimportant it is often essential to ensure that the originaldesign is repeated so that results can be compared

directly among assessments. However, solutions are oftenpossible in which new, more detailed recording methodsyield results that can be converted into a formcomparable to those derived from previous methods.

3.6 Conclusion

Routine implementation of this framework, in allsituations, would ensure that measures are appropriate tothe purpose to which they are being applied. As a resulteach biodiversity assessment would clearly identify: i)interested parties; ii) the attributes which those partiesvalue and are seeking to measure; iii) the extent ofexisting knowledge relevant to the assessment; iv) theassumptions used in the assessment and the limitations ofthe measure in addressing the valued attributes; v)precisely how each measure is defined; vi) the nature ofthe sampling strategy used; and vii) the data gatheringand analytical methods to be employed. Applying thisframework would also help to identify weaknesses incurrent approaches, as well as highlight major scienceand information gaps.

In the following section we demonstrate, with a series ofcases studies, how this framework can be used in anumber of circumstances, for terrestrial, freshwater andmarine systems, and at the ecosystem, species andpopulation levels.


In this section we present 11 case studies that illustratehow the framework, put forward in Section 3, can guideplanning for biodiversity measurements. We have chosenexamples to illustrate the widest possible kinds ofbiodiversity measurements, ranging from those thatencompass ecosystems to others involving geneticvariation within and among populations. While all ofthese case studies are based on real examples, in manycases we have considered hypothetical stakeholders whomight be interested in different attributes of each system.Consideration of differing stakeholder priorities illustratesthe importance of implementing full scoping phases, withprofound impacts on the subsequent choice of samplingstrategy, data gathering and analyses.

4.1 Forest cover in the Eastern Arc Mountainsof Tanzania

4.1.1 BackgroundThe forests of the Eastern Arc Mountains are prioritycandidates for biodiversity conservation. A keyrequirement is to understand how changes in habitat arearelate to changes in the delivery of ecosystem services.

4.1.2 Valued attributesMany different groups of interested parties areconcerned about forest loss in this area. Here two ofthese groups are considered: a group concernedprimarily with the maintenance of forest biodiversityand a group concerned primarily with the delivery ofecosystem services by forests to local people. Both needto measure the rate of loss of natural forest quantity andquality, now and in 2010. Those most interested inbiodiversity may also want to know about numbers ofspecies committed to extinction by habitat loss, whilethose most interested in services may need to knowabout stream flows, soil erosion, rates of harvesting wildspecies, and so on. Both groups probably want to collectadditional information on pressures and underlyingdrivers of deforestation.

4.1.3 KnowledgeThe Eastern Arc forests are among the most importantpriorities for biodiversity conservation in Africa, withexceptional levels of floral and faunal richness and endemism(Burgess et al 1998; Lovett 1998; Newmark 2002). Theireastern slopes intercept clouds arriving from the IndianOcean, with the run-off from the forests providing theprincipal water supply to major coastal settlements,including Dar es Salaam (where 3–4 million people and mostof Tanzania’s industry is located). This water catchmentfunction was the main reason for the establishment of thearea’s forest reserves by colonial authorities.

Since the 1950s, the Eastern Arc forests have beenreduced considerably in extent, with estimated rates ofloss for individual mountains running at 0.5–1.0% peryear, principally in areas not covered by forest reserves(Burgess et al 2002). However, quantitative estimates offorest loss are patchy, and estimates of changes in forestquality are almost non-existent.

The main reasons for forest loss have been clearance forfarming, both for subsistence and for sale to readilyaccessible urban populations in Dar es Salaam andelsewhere, as well as extraction of timber and non-timberforest products for local use (Burgess et al 2002;Newmark 2002). The intensity of these pressures andtheir underlying drivers vary with local population density,land tenure and access to roads.

Complete ground surveys of the extent of forestwould be impractical, however satellite imagery andaerial photographs can be analysed to yield estimatesof changes in forest cover. Studies in Brazil, Kenya andelsewhere (Pimm 1998; Brooks et al 1999) haveshown that the biological value of forests will dependnot just on their total area but on their degree offragmentation, with species-area and other theoryenabling us to predict very roughly how and whenspecies numbers change with area and isolation.Species lists of the terrestrial vertebrates and most ofthe flowering plants of Eastern Arc are reasonablycomplete and it is known which species are endemicto individual mountains or to the Eastern Arc as awhole (Burgess et al 1998; Lovett 1998; Newmark2002).

4.1.4 Model Combining species-area theory with empirical data onthe spatial extent of edge effects and on times taken forthe composition of the forest biodiversity to adjust or‘relax’ following habitat loss, should enable us toestimate the impact of changes in habitat area onspecies persistence. Far less theoretical or empiricalinformation exists on how changes in habitat arearelate to changes in the delivery of ecosystem services.However, the development of a water balance model ofthe region would be valuable. This should pay particularattention to modelling the relationship between thetype and extent of vegetation and the amount andtemporal variability of streamflow from mountaincatchments. Depending on which ecosystem servicesare of interest, models of soil erosion and harvesting ofwild species may also be necessary.

Measures of erosion and streamflow would be required,as would measures necessary to make parameterestimates for the water balance model of catchments.


4 Case studies

4.1.5 Relevant questions identified by the scopingstage• How has the extent of forest changed in recent

decades?• Where forest has disappeared, what other land cover

types have replaced it?• How do observed and expected changes in species

distribution relate to changes in forest cover?• How do changes in ecosystem service provision,

especially water supply from stream flow, relate tochanges in forest cover?

4.1.6 Sampling strategy, data gathering and analysis The core data for both groups of interested parties willlikely be time series of aerial photography or satelliteimagery to characterise the extent of forest fragments –done retrospectively at intervals, depending upon imageavailability, in order to improve and standardise estimatesof past losses. The same methods would be used tomeasure future habitat extent at intervals of not morethan five years.

If complete coverage of all Eastern Arc forests is notpossible, data collection should be carefully stratified,with disproportionate sampling effort devoted to areasexpected to have higher rates of clearance, higher levelsof endemism (for those interested in biodiversity), andhigher likely values for service delivery (because of denserhuman settlement, more rivers, steeper slopes, etc – forthose interested in ecosystem services) (see Achard et al2002 for a worked example). Collating historical imagerymay also be useful. Interpreting imagery will requireconsultation with local and international experts, as wellas a stratified programme of ground-truthing.

Those interested in biodiversity losses may want tosupplement estimates of loss of habitat quantity withground surveys to test predictions of changes in speciesoccurrence. These should again be stratified, focusingdisproportionately on areas of high loss and endemism,and on taxa that exhibit high endemism and which arerelatively easily surveyed and identified. Those interestedin changes in service delivery may be able to get someinformation (e.g. on erosion) from aerial or satelliteimagery, but will probably also need to collect aconsiderable amount of field data (e.g. on water flows,harvesting rates, etc.). Stratification criteria here mightinclude proximity to villages, slope, aspect, and groundcover. Historical data on service delivery may be extremelyuseful too, though at present little such information iscompiled in an accessible format.

Both the interpretation of remote imagery and groundsurveys could be expanded at relatively low cost toprovide information on pressures and drivers behindforest loss – through mapping land uses, settlement andinfrastructure, and through social surveys documentingreasons for conversion, household incomes, demographicpatterns, land tenure systems, and so on.

4.1.7 Science gaps • Very little is known, either empirically or theoretically,

of how estimates of changes in the delivery ofecosystem services relate to more readily assessedchanges in habitat area – yet this is of crucialimportance to many stakeholders, and may be farmore likely to result in effective responses to pressuresthan will information on species losses.

• Data on historical levels of ecosystem delivery (such asstreamflow) is very limited – yet, in many cases suchinformation is likely to have been gathered byorganisations such as museums and colonialadministrations. A major effort to collate these dataand make them accessible may be very worthwhile.

4.2 Global extent of open freshwaterhabitats

4.2.1 BackgroundMost freshwater ecosystems support substantial levels ofbiological diversity. Natural processes and humandemand for water for drinking and irrigation threatenthese ecosystems. A fundamental requirement is for adynamic, descriptive model of the quantity, quality anddiversity of freshwater ecosystems worldwide.

4.2.2 Valued attributesNo substitute is known for freshwater; it is an absolutenecessity for all non-marine life. Freshwater ecosystemsprovide a range of services, including the provision ofreliable water supplies for human consumption,irrigation, sustainable freshwater fisheries, nutrientcycling, detoxification of wastes, recreational pursuits,and aesthetic pleasure. All of these are products ofhealthy ecosystem functioning, which is driven andsustained by the activities of a great diversity of aquaticlife forms. Stakeholders include all individuals of thehuman species, whether in rural or metropolitan areas.

Ponds, lakes and swamps in particular tend to bebiologically productive, and this productivity is usuallyassociated with substantial biological diversity. Thus alarge aquatic biodiversity is typically supported within arelatively small geographical area. Freshwater ecosystemsalso support species such as dragonflies, hippopotamus,coarse fish, and bulrushes that do not exist in otherecosystem types, and many lake ecosystems are home toendemic species. It is self-evident that the persistenceglobally of numerous unpolluted freshwater ecosystemsof diverse types is necessary for the conservation offreshwater biodiversity worldwide.

4.2.3 KnowledgeMore than 99% of the fresh water in the biosphere islocked up in polar ice and glaciers, or hidden from viewbelow ground, but all fresh waters, whether still orflowing, groundwaters or bogs, are connected with each


other through dynamic interactions (e.g. above andbelow ground exchange), so that impacts on oneecosystem type eventually filter through to otherecosystem types.

It is generally acknowledged that the world’s waterproblems continue to worsen, but the deterioration is notalways directly linked to anthropogenic factors. LakeChad has shrunk to one twentieth of its size 40 years ago,but this may have more to do with persistent droughtthan with direct human impact. The causes of otherthreats are more obvious. Since 1950, the number oflarge dams worldwide has increased from about 5,000 toroughly 45,000, with a proportionate increase in thedegree of ecosystem alteration and destruction. In SouthAmerica, the projected multinational Hidrovia Dam will (ifit goes ahead) turn the Paraná and Paraguay Rivers into abarge canal that will drain the Pantanal – one of theworld’s largest and most biodiverse wetlands.

Where human demand for fresh water exceeds supply,ecosystem degradation usually follows. The US (sevenstates including California) and Mexico, remove almost allof the water that flows down the Colorado River, leavingvirtually nothing to service the deltaic ecosystem on thefringe of the Gulf of Mexico.

Problems associated with excessive irrigation are mostclearly demonstrated in the Murray-Darling Basin – awatershed that accounts for one seventh of the area ofAustralia. Extensive irrigation has brought the soils’naturally occurring salts into the root zones and rivers,leading to basin salinisation. Most of the Basin’s rivers nowexceed World Health Organisations’ guidelines for drinkingwaters. On the other side of the continent, some 450species of plants, insects and birds have been officiallyrecognised as being under threat in Western Australiaalone. One of the biggest contemporary problems is theover-pumping of ground waters, especially across largeareas of the middle and Far East, and in India, where theGanges runs dry for part of each year.

In the UK, about 90% of the total number of standingbodies of fresh water have areas of one hectare or less,and there may be 750,000 of these in England and Walesalone. It is estimated that they are decreasing in numberby several thousand per year, largely as a result of newdrainage schemes and urbanisation.

4.2.4 ModelLoss of and damage to freshwater ecosystems threatensbiodiversity worldwide, but quantitative data at theglobal scale are lacking. The fundamental requirement isfor a dynamic, descriptive model of the quantity, qualityand diversity of freshwater ecosystems worldwide. At thecore of this model is a world map, possibly sub-dividedinto biogeographic regions, and, at a range of finerscales, superimposed layers of quantitative informationfor:

• Availability of unpolluted surface waters relative tothat in undisturbed catchments at the same latitude

• Level of salinisation of surface waters• Time required to recharge aquifers (assuming current

levels of abstraction)• Probability of catastrophic flooding, including that

associated with political/military activities (e.g. theTigris-Euphrates watershed)

• Distribution of pressures and drivers arising fromhuman use.

With the incorporation of time-based data, the mostuseful products of the model will be measured rates ofchange in extent and quality of water in sampleecosystems.

4.2.5 Relevant questions identified by the scopingstageThe aims are to provide global time series data for the realextent, quality and diversity of freshwater ecosystems andto use these to derive rates of change in these parametersand the services they provide to stakeholders. Causes ofloss of freshwater habitats would also be identified.

4.2.6 Sampling strategy data gathering and analysisAerial photography and remote sensing techniques arenow very advanced, capable of recording with very highresolution. They can be deployed for accurate mapping oflakes, rivers and floodplains, and can measure subtlewater level changes over time in all of these. A majoradvantage over traditional mapping is that the data arepotentially available in real time. It is now possible to usedata from airborne or satellite based spectrometers toestimate aquatic chlorophyll a concentrations. Waterstatus can be assessed using a number of indicatorsincluding Secchi disk depth, turbidity and stream flowrates. The main problem is that coverage of the world, atan appropriate level of detail, is far from complete,although it is excellent for North America and parts ofEurope. For areas of the world where data are notavailable or where maps are currently inadequate, remotesensing is probably the only feasible approach togathering data for small (< one hectare), invariablyephemeral, water bodies in large continental areas.

For the Western Palaearctic, it should be possible toproduce the first maps within a few years, especially if theEuropean Union Water Framework Directive and otherlarge-scale projects can drive the project. Of principalinterest are trends over time, and it should be possible toanswer the question by the year 2010, as to how theglobal status of freshwater ecosystems is changing.

4.2.7 Science gaps• What will happen if some of the most extensive

wetlands in the world are drained is simply not known. • Some habitat types (e.g. farmyard ponds) are much

more common globally than others (e.g. the deep riftvalley lakes of East Africa). Loss or degradation of a


single major lake may result in a significant number ofspecies extinctions, yet there is no internationallyagreed hierarchical list of vulnerable freshwaterecosystems.

• Local communities often have low awareness of thefreshwater ecosystems in their neighbourhoods,although voluntary organisations in the UK and othercountries have raised the public profile of freshwaterponds in particular. In the USA, the scientificmanagement and conservation of water resources isincreasingly becoming the prerogative of the localcommunity (e.g. the Catskills Watershed agreement,which involves local people, forest owners, New YorkCity and State, the Environment Protection Agency,and environmentalists), who collectively recognise thevalue of ecosystem services (e.g. water purification byforested catchment) and are supporting an enhancedwatershed protection programme for New York City’sdrinking water supply.

4.3 Trophic integrity of marine ecosystems

4.3.1 BackgroundMarine ecosystems have been exploited for hundreds ofyears. Fisheries tend to target the largest and mostvaluable species. But intensification of the industry anddepletion of the large bodied and higher trophic levelspecies has led to the targeting of increasingly smallerspecies, further down the food chain.

4.3.2 Valued attributesIntact, pristine ecosystems often support high levels ofbiomass of large-bodied and higher trophic level species(Odum 1969; Christensen & Pauly 1998). By contrast,disturbed, exploited and polluted ecosystems can becharacterised by the absence or rarity of such species,and by dominance of small-bodied species with highrates of population turnover, usually from lower trophiclevels (Odum 1969; Christensen & Pauly 1998).Compared to undisturbed ecosystems, they may havesimplified food webs and can be prone to blooms ofalgae, microbes and plankton (Jackson et al 2001).Relatively intact marine ecosystems may be able tosustain a level of exploitation of their animal populationsand can act as sources of recruits for fisheries in otherareas. Stakeholders include those who depend uponfishing for their livelihoods and those who value marinebiodiversity.

4.3.3 KnowledgeRecent analyses of fisheries and historical data suggestthat over the last 500 years, marine ecosystems haveundergone major losses in biomass of larger-bodied andhigher trophic level species (Jackson 1997; Jackson et al2001). In parts of the world where large-scaleexploitation extends far back in time, such lossesoccurred so long ago that few people today appreciate

that the species were ever common in these places. Forexample, at the time of Columbus’ discovery of theCaribbean, there might have been as many as 33 millionturtles there (Jackson 1997). Early explorers and settlerskilled many of these and other large-bodied animals tosupply their food, and for commodities such as oil andfur. Throughout the world, other organisms includingsea otters, seals, manatees, dugongs, and cetaceanshave met similar fates.

The intensification of fishing has continued the process ofecological extinction of species from marine food webs.Fisheries tend to target the largest and most valuablespecies first and as each is depleted they then move on toothers that are smaller and less desirable. Pauly et al(1998) call this phenomenon ‘fishing down marine foodwebs’ and have used world catch statistics from the Foodand Agriculture Organisation of the United Nations toquantify the phenomenon.

4.3.4 ModelThe model is based upon a simplified description of thetrophic relationships among marine animals and plants. Italso includes relationships between the level of humanexploitation of a given species and the abundance ofother species at higher trophic levels.

4.3.5 Relevant questions identified by the scopingstageThe main aim is to quantify changes over time in theaverage trophic level of marine animals being exploited byfisheries.

4.3.6 Sampling strategy, data gathering and analysisThe relevant data concern time series of landings of fishtaken by fisheries in defined areas. Pauly et al (1998)assigned an integer value to the trophic level of eachspecies in landings, based on the fraction of its diet thatcomes from each trophic level. For example, if 50% of anomnivorous species’ diet consists of algae and 50% comesfrom species that are exclusively herbivorous, it will have atrophic level of 1.5 [0.5 x 1 (algae) + 0.5 x 2 (herbivores)].This approach deals effectively with the problem that mostspecies have diets that consist of foods taken from a varietyof trophic levels. Using this information, it is possible toestimate the average trophic level of animals in fisherylandings by multiplying the proportion in the catch by thetrophic level of the species. Changes in this measure overtime could then be examined.

Trophic levels of landings have been declining in nearly allregions of the world since the 1950s (Pauly et al 1998). Insome areas, such as the North Atlantic, declines havebeen steep while in others they have been lesspronounced, such as the Mediterranean. Slower declinesin the Mediterranean reflect a more extended history ofexploitation, where mean trophic levels of catches werealready low by the 1950s. Globally, the trend has been aloss of 0.1 trophic levels per decade.


Pauly et al’s (1998) approach has been criticised onseveral grounds. First, the analyses involved somenecessary approximations since catches are oftenreported in aggregate units, such as ‘mixedgroundfish’. Trophic levels had to be assigned to theseunits on the basis of probable rather than actual speciescomposition. Second, some argue that changes introphic composition of landings may have more to dowith changing tastes than change in availability ofspecies. In the 1950s, few people ate shrimp, which aredebris eaters, but these detritivores are now consideredamong the most desirable of seafood. Third, reportedlandings are not the same as catches, since they do notaccount for species caught but discarded as unwantedbycatch, or if fishermen have gone over their quota.

More recent analyses have attempted to deal with someof these problems. For example, Pinnegar et al (2002)estimated trends of decline in mean trophic level in theNorth Sea since 1982. They used fishery survey data, soremoving the problem of unreported catches andreducing the problem of aggregation of species intobroader categories. Like Pauly et al (1998), they found aclear trend of decline in mean trophic level. They alsoexamined price trends for species in relation to trophiclevel, and found that high trophic level species hadexperienced faster increases in price than those fromlower down food webs, indicating that demand for thesespecies remains high. This suggests that the view thatlower trophic level species are being caught today onlybecause of changing tastes is false. While consumershave clearly developed tastes for species like shrimp andlobster in recent years, high trophic level species remainhighly desirable.

Few scientists now doubt that ‘fishing down marine foodwebs’ is real, and most consider that Pauly et al’s (1998)measure represents a robust indicator of underlyingtrends of species depletion. It provides a valuable proxymeasure of loss of biodiversity and its ecosystem leveleffects.

4.3.7 Science gaps• The quantitative relationship between landings and

catches is relatively poorly known for most parts of theworld.

At regional and global scales, applying the measure toFood and Agriculture Organisation (FAO) data willcontinue to provide a credible and useful synoptic pictureof regional and global trends. However, at a smaller scale,countries could adopt approaches similar to thatemployed by Pinnegar et al (2002) and use fishery surveydata directly to calculate year-on-year measures of trophiclevel of catches. The ecosystem level effects of fisherymanagement approaches designed to rebuild depletedstocks should soon become apparent through systematicapplication of this measure.

4.4 Meiofaunal indicators of pollution

4.4.1 BackgroundMany inshore coastal and estuarine sites are vulnerable toindustrial pollution. Meiofauna (mud- and sand-livingorganisms including nematodes, arthropods, andannelids) are especially sensitive, but their localbiodiversity is often difficult to characterise because of alack of available specialist taxonomic knowledge. ModernDNA-based methods offer the prospect of simple andunambiguous surveys of biodiverse but taxonomicallydifficult groups.

4.4.2 Valued attributesInshore habitats support a large animal diversity, notablywading birds and meiofauna. They also provide theresources required for a range of recreational pursuits.Monitoring of pollutant impact on coastal sites istherefore important for maintenance and protection, andalso for assessing remediation and recovery processes.

Many large industrial concerns use tidal rivers as part oftheir waste management strategy. While currentdischarges are carefully monitored and controlled,historical discharge, and also accidental discharge, canresult in unknown effects on ecosystems under threat. Forexample, in the Firth of Forth, there are several largeindustrial/marine establishments (Grangemouth, RosythDocks and Europarc), as well as significant inputs fromcities and towns and associated domestic and lighterindustrial activity. Confounding this pattern ofanthropogenic pollutant input are several ‘natural’sources of coal-measure derived material, in the form ofgas seeps and oil-shale seeps. Monitoring the health ofthe Firth’s shores requires measurement of biologicalresponses and recoveries from both sources ofdisturbance.

4.4.3 KnowledgeMeiofauna are very sensitive to disturbance, and after amajor pollution event, may take years to recovercompletely. On the other hand, their short life cycles andability to colonise, means that individual species’responses may be relatively rapid. One barrier to the useof meiofaunal organisms in monitoring is their sheerabundance and diversity: it is rare to find a workercapable in the many different phyla involved, and rarer tofind one with time on his/her hands. Nematodes inparticular are often neglected in meiofaunal surveysowing to the small size of each individual, the relativepaucity of easy morphological characters and thedifficulty in identifying juvenile and other stages.

Meiofaunal taxa occur in many phyla, including, amongothers, the nematodes, arthropods, tardigrades andannelids. For each of these groups, at least some DNAsequence data is in a publicly accessible database


(GenBank www.ncbi.nlm.nih.gov/Genbank/GenbankOverview.html), having been provided by researchersworldwide working on phylogenetic or populationgenetic questions. These data can be used to developpotentially unique markers for species. DNA sequencing isbecoming a common technique and its cost is droppingsignificantly. In addition, the technology for isolation ofsequenceable DNA from small specimens has beendeveloped to a stage where it is possible to implement iton a large-throughput scale.

Molecular bar-coding methods involve isolating aninformative segment of the genome and sequencing it foreach specimen, then using the sequence to place anindividual in a molecular taxonomic unit (Floyd et al2002). These methods may be both sensitive, in that theycan distinguish between closely related ‘species’, anduniversal, insofar as they are applicable across the rangeof biological diversity. The resulting data consist of DNAsequences with associated ecological and othermetadata. These data can be archived and sharedbetween sites by the use of Internet-available databases.

4.4.4 ModelThe meiofaunal assemblage of a particular site is anintegrated response to both current and historicalconditions, overlaid with stochastic and seasonal/climaticmetapatterns. It is assumed that the distribution andabundance of individual molecular taxonomic units areinfluenced by the abundance of food, predators and thechemical composition of water and sediment.

4.4.5 Relevant questions identified by the scopingstageThe main aim is to build up a spatial-temporal map ofmeiofaunal species diversity and the relative dominanceof individual molecular taxonomic units in the area ofinterest, in this case the shores of the Firth of Forth. Thesedata can also be used retrospectively to describe changesover time and relate them to external pressures anddrivers.

4.4.6 Sampling strategy, data gathering and analysisThe survey requires a dedicated sample acquisition andprocessing team. Small sediment samples are taken atsites distributed along the tidal shore, based on asampling strategy that takes into account height abovelow water and substrate diversity as well as geographicalspread. Samples are sieved for meiofauna and asubsample taken for individual molecular analysis. Theremainder of the sample is archived, part as a standardfixed ‘museum’ deposition, and part as a total meiofaunalpreparation. These permit future morphological surveysor DNA-chip based surveys. Individual meiofauna DNA isextracted, and a segment amplified and sequenced.Sequences are stored in a relational, online database.

Each sample can be assessed for taxon diversity (i.e. thenumber of different molecular sequences), taxon relative

abundance, taxon biology, and thus ecology. Biologicalidentifiers are added to sequences by comparing themwith a database of sequences from well-identifiedspecimens. Identity allows attribution of the knowntaxon’s biology to the sequenced specimen, while closesimilarity will permit less robust annotation. Over time,with integration of the space and time-mapped sequencebiodiversity data with known abiotic factors (particularlythe point sources of pollutants) a map of meiofaunalresponse to the environment can be established. Re-assessment of diversity at intervals permits seasonal aswell as longer term (e.g. associated with global warming,or the release from pollutant pressure) trends to beestablished. The data also allow simple comparison withother sites sampled in a similar manner.

Data analysis tools developed for molecular bar-codesurveys of terrestrial species (Floyd et al 2002),incorporating sensitive discriminators for sequencingerror and taxon identification can be adapted to thesurvey. All data and analyses can be readily sharedbetween sites as raw sequence, or as annotatedmolecular taxonomic unit measures.

4.4.7 Science gaps• Very few DNA sequences currently exist from properly

identified meiofaunal specimens, in public databases.These are essential to establish the relationship betweenhistorical knowledge based on morphology andLinnaean taxonomy, and the new molecular taxonomy.

• The isolation of meiofauna individuals and thegeneration of molecular sequence are currently labour-intensive, but could be robotically automated, perhapsin one (or a few) environmental genomic taxonomycentres.

• To simplify the process, once a large database ofrelevant sequences is available, it would be possible todesign a DNA oligonucleotide microchip thatrepresented all relevant taxa. The chip could beproduced in large quantities, and samples screenedfrom bulk meiofaunal DNA samples by hybridisationfor a fraction of the already small sequencing cost.

• A long-term goal is the development of methods thatcan be universally applied to all inshore habitats. Auniversal protocol would yield congruent datasets,permitting analyses of ecological patterns on a largescale.

4.5 Global state of plant biodiversity at thespecies level

4.5.1 BackgroundCompleting a global assessment of the conservationstatus of plant species diversity by 2010 is one of theachievable goals identified as priorities for action on aglobal scale (NRC 1980; Ehrlich 2002). Its completion isalso highlighted in the Global Strategy for Plant


Conservation, adopted by the Conference of the Partiesto the CBD in 2002. Such an assessment will enable us toobtain a clear picture of those plants under threat(currently estimated as 22–47% of species), and todevelop action plans to safeguard the most threatened ofthose species (Pitman et al 2002; Pitman & Jørgensen2002).

4.5.2 Valued attributesPlants are the primary producers and key structuralelements of most ecosystems on land. Total plant diversityis often correlated with diversity of other groups oforganisms (Williams et al 1998). The valued attribute isglobal plant species richness. Many different groups ofinterested parties wish to know about the state of plantdiversity at the global level, ranging from those interestedin maintaining the use value of many thousands ofspecies to conservationists wishing to maintain overalllevels of diversity. Plant diversity is relatively well knownand thus a global synthesis represents a more achievableglobal target for plants than for other larger and less well-studied groups, such as beetles or fungi (Erhlich 2002).Making this information usable by non-specialists andtransferring it to the field is an important goal, particularlyfor areas of high biodiversity where expertise may belacking.

4.5.3 KnowledgeProgress towards a global overview of the state of plantdiversity has been relatively rapid. In 1980 no SouthAmerican country had complete Floras or checklists.Today, checklists have been compiled for Ecuador(Jørgensen & León-Yánez 1999), Peru (Brako & Zarucchi1993) and Argentina (Zuloaga et al 1994; Zuloaga &Morrone 1999), demonstrating that such syntheses arepossible and can be completed in a timely manner.Species level treatments like these are the starting pointfor more formal and detailed conservation assessments(Valencia et al 2000), which are lacking for the majority ofthe world’s plant species.

Despite this progress and given that vascular/seed plantsare one of the better-known major groups of organismscurrent knowledge is surprisingly patchy. As nocomprehensive checklist of plants of the world existsthere can be no agreed figure for total species number.Estimates range from 270,000 (Prance 2000; Scotland &Wortley 2003) to 420,000 (Govaerts 2001; Bramwell2002). Complete species level checklists with fullsynonymy and formal conservation assessments for eachaccepted species are available for only a tiny proportion ofplant species, for example the conifers (c. 600 species.),the cacti (c. 900 species) and some groups of orchids (c.500 species). For other groups, published checklists havebeen made with synonymy and geographical informationat country level (e.g. the International Legume Databaseand Information Service (ILDIS) www.ildis.org), but theinformation is generally not sufficiently detailed to serveas the basis for formal species level conservation

assessments. For groups such as Sapotaceae,Euphorbiaceae, Magnoliaceae, and Fagales, informationis only available on the numbers of species and in whichcountries they occur, but not which are most threatenedand in need of immediate action to conserve them. Anestimated 30% of plant species are included in globalsynonymised checklists of this sort. For the vast majorityof plant species there is no comprehensive list of acceptednames and synonymy. Instead, users rely on theInternational Plant Names Index (www.ipni.org), whichlists all published names including many which shouldcorrectly be placed in synonymy. The databasemaintained at the Missouri Botanic Garden (TROPICOS;http://mobot.mobot.org/W3T/Search/vast) presentssynonymy information for many taxa, but this is patchyand reflects areas of current staff and project interest (i.e.Mesoamerica, Ecuador, Bolivia, Iridaceae, Poaceae etc.).

4.5.4 ModelThe model assumes that an informed collation ofnomenclatural information from authoritative publishedsources (Floras and monographs) will result in a listingthat approximates to a globally comprehensive survey ofplant species. It further assumes that a large proportion(80–90%) of the taxonomic and nomenclaturalinformation published in the recent literature iscongruent and can be presented as a consensus listingwith relative ease. Specialist taxonomic input is assumedto be a very limited resource that is best directed in atargeted fashion towards i) refining and correcting draftlistings and ii) presenting defensible resolutions oftaxonomically controversial groups.

With respect to the utility of the eventual product, themodel assumes that the majority of plant species arealready known to science so that, once synonymy is takeninto account, known diversity is a reasonable proxy fortotal diversity for most plant groups and (to a lesserextent) areas. This assumption is more likely to hold truefor groups where the baseline taxonomic information isquite good, for example Fabaceae (legumes). It should betested for groups for which the available taxonomy ispatchy and dated, for example Celastraceae.

Implicit in the whole undertaking is the key assumptionthat the results based on those plant species that areknown will be representative of the true status of all plantspecies, whether known to science or not. Somesynonymy will remain undetected indefinitely and tens ofthousands of plant species remain to be described. Thus afully complete and definitive checklist is not attainablewithin a timescale consistent with the urgency of theneed to protect as many plant species as possible,whether known to science or not.

4.5.5 Relevant questions identified by the scopingstage• How many plant species are known only from one

country/state?


• Which plant species should be considered as prioritiesfor conservation assessment and action?

• How can the rate at which the conservation status ofindividual plant species is assessed be increased onehundred fold, so as to focus attention on the mostthreatened species before they become extinct?

• This plant has been referred to by at least two differentnames – which is correct?

• This plant appears to have changed name more thanonce over the past few years – how will this affect theaccessing of all the relevant literature for planningresearch or a conservation strategy?

• How can a meaningful comparison of these specieslists be made from adjacent reserves in neighbouringcountries?

• How many plant species are there?

4.5.6 Sampling strategy, data gathering and analysisSampling for a global checklist should be conducted on ataxonomic rather than a regional basis and based onknowledge already in the published literature. Herbariumcollections should be used where available, particularlyfor poorly known groups and at local scales (Funk et al1999). Decisions must be attributed; it will be necessaryto trace synonyms by contributor if not from the literature(see examples in W3FM – Flora Mesoamericana;www.mobot.org/mobot/FM).

Step 1: Create a searchable database for disseminationvia the Web. The result would be a ‘first pass’ list withaccepted names and place of publication; synonyms withplace of publication (and attribution of taxonomicdecision); geographical distribution by country(Taxonomic Databases Working Group (TDWG) level 3,with literature or herbarium specimen ‘vouchering’); anda preliminary assessment of conservation status for eachaccepted species. These assessments would be: i) notassessed – assumed not threatened, ii) assessed and noapparent threat, iii) preliminary assessment indicatessome level of threat, iv) data deficient, v) formal IUCNassessment completed.

Step 2: (in the specialist community) Send lists totaxonomically focussed referees; add links to on-linespecimen databases (many of which exist, but are notlinked up). Make revisions and add herbarium data asadvised by referees.

Step 3: Make the global checklist available on-line in aformat that can be annotated and checked by users of theinformation and accessible to non-specialists.

4.5.7 Science gaps• The effect of synonymy remains one of the most

difficult factors to assess. Estimates of overall rates ofsynonymy for vascular plants vary widely (see Section2.2.1) (Govaerts 2001; Bramwell 2002; Scotland &Wortley 2003). Evidence from recent monographs mayshed some light on the problem but a large proportion

of the total diversity is accounted for by large generathat tend to be neglected by monographers.

• Distribution data compiled from the literature andfrom herbarium vouchers, are unlikely to prove anadequate basis for a preliminary assessment ofconservation status in certain groups that tend to beunder collected (such as palms and cacti). For thesegroups robust alternative means of documentingdistribution must be established.

4.6 Genetic diversity in wild lentils

4.6.1 BackgroundIn 30 years time, the world’s human population will havegrown significantly to 8.5 billion according to someestimates and substantial, sustainable increases in foodsupply will be needed. Conservation and sustainable useof plant genetic resources should form a foundation uponwhich improvements in sustainable agriculturalproductivity can be built. (Hawkes et al 2000; Maxted et al1997).

4.6.2 Valued attributesWild species are an important component of agriculturalbiodiversity. Species that are relatives of crop plantscontain valuable genes for crop improvement, and wildspecies of plants may be important nutritionally andculturally to people in many parts of the world, wherethey serve as food in times of famine, provide vitamins,minerals and nutrients, and provide income for cash-poorhouseholds (Hawkes et al 2000; Maxted et al 1997). Inthe case of wild species of lentils, it will be necessary tobreed better adapted crop varieties in the future(Ferguson et al 1998c). Natural populations of lentilspecies conserved in regions of greatest diversity will beparticularly valuable when their genetic diversity resultsfrom adaptation to changing abiotic factors. Wild speciesof lentil form an important component of ecosystemsthat are vulnerable to climate change, and to reducedrainfall in particular (Ferguson et al 1998a). Wild lentilgermplasm is represented in ex situ seed bank collectionsbut the diversity conserved does not accurately reflect thegenetic diversity found in nature (Ferguson et al 1998a).

4.6.3 KnowledgeThe genus Lens consists of 6 wild taxa that are relatives ofthe cultivated lentil – L. culinaris (Ferguson et al 2000).Their distribution is the Mediterranean basin, with onespecies spreading into Central Asia. All species exist insmall, disjunct populations, predominantly in undisturbedrocky or pine forest habitats. These factors make thesespecies vulnerable to genetic erosion unless effectiveconservation in protected areas can be achieved.Approximately 800 seed accessions are conserved in genebanks – the majority at the International Center forAgricultural Research in the Dry Areas (ICARDAwww.icarda.cgiar.org/) where ‘passport data’ are


maintained. Other data are available from herbariumspecimens. Together these data sets are sufficient todetermine accurate geographical distributions andecological preferences for all species. Each species ishighly inbreeding; heterozygotes are extremely rare. Onaverage, 89% of genetic diversity within the speciesoccurs among (as opposed to within) populations, butwithin-population variation is significantly different fordifferent populations, and is therefore an importantconsideration for conservation (Ferguson 1998b).

4.6.4 ModelThe model assumes that there is genetic variation withinthe wild species of lentils and that it may vary within andamong populations in different parts of the geographicalrange. It further assumes that measures of genetic distanceobtained from a sample of molecular markers canrepresent this variation in a quantitative and reliable way.

4.6.5 Relevant questions identified by the scopingstage The main aim is to map the geographical distribution ofLens species and the genetic variation within thesespecies. It is feasible to do this for four of the six wild taxaof lentils (L. culinaris subspecies orientalis, subspeciesodemensis, L. ervoides, L. nigricans). From the map itshould be possible to identify the most important wildpopulations for the conservation of genetic resources(Ferguson et al 1998a).

4.6.6 Sampling strategy, data gathering and analysis An assessment of the distribution of the genetic diversityof each species both within and among populations wasmade using molecular markers (RAPD). This approachwas chosen because RAPD markers are easily applicableto species where little or no genomic information isavailable, and they are relatively simple and inexpensive.

Geographical distribution of the genetic variation in thefour taxa was assessed by calculating genetic distancebetween population samples, cluster analysis, thencalculating gene diversity as well as ‘number of clustersper sub-region’ (NCSR). Further analysis was undertakento determine whether diversity was directly associatedwith ecological and geographic range, so that areaswhere further sampling or collection should beundertaken, in a recurrent biodiversity assessment, couldbe identified.

Seed samples of each species were obtained from thegene bank accessions held at ICARDA (141 in total).Exploration and seed collection activities provided adataset and material for assessing the correlationbetween genetic diversity and ecogeographic range tohighlight future conservation priorities. While seedcollection from each population was randomised, thelocations of these populations was not, but was guidedby ecogeographic information from various sources.Sampling (by ICARDA) was concentrated in the Fertile

Crescent of the Middle East, but samples were alsocollected throughout the total geographical range of wildlentils from Portugal to Uzbekistan. It was not possible todetermine the likely total number of populations inexistence for each species, although previous pilot studiesgave indications that the Fertile Crescent had the highestnumbers of populations. For every sample, geographicalco-ordinates of the collection site were available. Bothassessments of diversity were undertaken on all samples.

The two measures of diversity (gene diversity and NCSR)sometimes gave conflicting results. To meet the objectiveof maximising conserved variation, the use of NCSR wasused. Some areas of high diversity coincide with highplant population densities and are therefore good targetsfor reserve establishment and in situ conservation (listedbelow). In one case (L. culinaris subspecies odemensis)however, genetic diversity is distributed widely and will bemore problematic to conserve in situ.

• Lens culinaris subspecies orientalis: two centres ofdiversity are identified (west and north Jordan andsouthern Syria; southeast Turkey and northwest Syria)and should be prioritised for in situ conservation.

• Lens culinaris subspecies odemensis: genetic diversityis localised in small pockets of distinct germplasm, andsix of the eight subregions studied exhibited uniquegermplasm. For adequate conservation, many smallreserves throughout the distributional range wouldneed to be established.

• Lens ervoides: a centre of diversity associated with highpopulation numbers (density) is found on the coast ofSyria and is a good target for reserve establishment.

• Lens nigricans: a clear centre of diversity with a highpopulation density exists in western Turkey that shouldbe a target for in situ conservation.

Areas of high and unique genetic diversity are located foreach taxon in Turkey, Syria and Jordan; outside of thesecountries, much less genetic diversity, total and unique, isfound. This indicates that although the majority ofexisting ex situ conserved material has been sampled fromTurkey, Syria and Jordan, it still under-represents thegenetic diversity from these countries. Conversely thediversity found in peripheral countries is already wellrepresented in existing collections.

4.6.7 Science gaps• New data are needed for two taxa (not included in the

study) – Lens culinaris subspecies tomentosus was onlyrecently separated taxonomically from subspeciesorientalis, so populations and samples with definitiveidentification were not included. Lens lamottei isregarded as a cytotype of L. nigricans, and insufficientpopulations/material could be identified for the study.

• Although the majority of existing ex situ conservedmaterial has been sampled from Turkey, Syria andJordan, it still under-represents the genetic diversityfrom these countries.


• In situ genetic reserves have recently been establishedfor wild lentils in Turkey, Syria, Lebanon, Palestine andJordan, but even these multiple reserves cannotrepresent the total genetic diversity found in nature.ICARDA continues to routinely sample genetic diversityfor wild lentils at the geographical centre of theirdiversity.

• Other molecular markers such as SSRs, AFLPs or SNPscould provide much more transferable information,and could be developed as specific indicators ofbiodiversity loss for future assessments in this group ofspecies.

4.7 Freshwater fish of conservation concern inthe UK and Mexico

4.7.1 BackgroundAround 40% of the 25,000+ species of fish are found infreshwater habitats. This figure is all the more impressivegiven that available fresh water, in lakes and rivers,accounts for less than 0.01% of water on the Earth(Nelson 1994).

4.7.2 Valued attributesFreshwater fish are of significant value to a wide variety ofinterested groups. Some species, for example the bony-tongued pirarucu, Arapaima gigas, and the herring-likeLimnothrissa miodon, support substantial fisheries.Others, including the Atlantic salmon, Salmo salar, areiconic game species and are associated with a significantrecreational industry. Freshwater species are important infish farming and underpin the valuable home aquaristtrade. In addition, they have been extensively used forinvestigations of behaviour, ecology, physiology, geneticsand evolution. Finally, freshwater fish are useful indicatorsof pollution and have come to symbolise a healthyecosystem. The return of salmonids to the Thames inLondon was hailed as a breakthrough in environmentalmanagement.

4.7.3 KnowledgeAlthough freshwater fish have been the focus of a largebody of scientific research and are well documented intemperate regions it is estimated that Amazonia and otherlarge tropical systems contain significant numbers ofunrecorded species. What is certain is that the threats facingfreshwater fish species worldwide are disproportionatelyhigher than those for terrestrial vertebrates. More than 20percent of the world’s known 10,000 freshwater fishspecies have become extinct or endangered in recentdecades. In the United States 343 fish species (36% percentof the fauna) are at risk of extinction and 27 species havealready been lost (Moyle & Leidy 1992).

Two examples are used to highlight the issue. In the UKrelict populations of whitefish (genus Coregonus) arefound in the English Lake District, Scotland, North Wales

and Northern Ireland (Winfield et al 1996). The pollan, C.autumnalis, is still fished commercially in Lough Neaghbut the population is at risk from poor water quality andcompetition from non-native roach, Rutilus rutilus. Thevendace (C. albula) and whitefish (or schelly, powan orgwyniad) (C. lavaretus) are already protected under theWildlife & Countryside Act 1981. Both surviving UKvendace populations and four of seven whitefishpopulations occur in stillwaters of the English LakeDistrict, where they and their habitats have been subjectto considerable conservation research and managementover the last few years.

The second example, the Goodeidae, are an endemicgroup of livebearing Mexican fish. Thirty-five species arecurrently recognised. Of these 19 species are endangered(eight critically) and two are already extinct in the wild(though the possibility of reintroduction exists as somestocks are maintained in culture). Habitat loss andfragmentation – a result of the increasing urbanisation ofCentral Mexico – invasive species and pollution are theprincipal factors in the loss of native fish. Mexicanbiologists are currently quantifying these threats (De laVega-Salazar et al in press a, b).

4.7.4 ModelA simulation model is used to carry out populationviability analysis (PVA) on each of the fish species. Thissubdivides each population into component sub-populations where necessary. Effects on demographicrates of exploitation, predation or competition fromintroduced species and habitat loss and degradation areincluded in the model and population-level outcomes,such as probabilities of extinction are assessed.

4.7.5 Relevant questions identified by the scopingstage • What is the probability of extinction and expected

population size within a specified time period underexisting and forecast conditions?

• How would specified changes to management affectthe probability of extinction and expected populationsize?

• Would reintroductions be likely to increase populationviability?

• Can a minimum viable population size be identifiedbelow which the population becomes vulnerable tostochastic extinction?

• What do local resource managers need to know toimplement effective conservation programmes?

4.7.6 Sampling strategy, data gathering and analysisIn commercially exploited species (such as the pollan)sampling can be integrated with the normal fisheriesactivities. Where vulnerable populations and species areinvolved non-destructive sampling should be usedwherever possible. The goal is to monitor changes inspecies status over time and evaluate threats. Datagathered will include population level data for as many


species at as many sites as possible, habitat quality datafor rivers and lakes and the presence and abundance ofinvasive species. Since inferences on changes in theviability of populations and species require repeatedsampling it will not be possible to examine all taxa at thesame degree of resolution. One approach is to focus onkey groups that are already reasonably well documented.These include the Lake District whitefish and goodeids inthe genera Skiffia and Zoogoneticus.

All data gathered must be appropriately archived in rawand processed form. Forced and unforced changes insampling techniques must be taken into account duringdata interpretation. Since freshwater fish throughout theworld are affected by the same anthropogenic impactsthere is considerable scope for collaborative projects thatanalyse the rate of biodiversity loss and draw up commonguidelines for the preservation of stocks and species(Moyle & Leidy 1992).

4.7.7 Science gaps• Population level data such as size, structure and

recruitment over all species’ ranges are not uniformlyavailable for most species of freshwater fish.

• Presence (and abundance) of invasive species coupledwith field investigations (supported where possible bylaboratory studies) of interactions between invasiveand target species will be critical to monitoringchanges in relation to the 2010 target.

• Close monitoring of environmental variables includingeutrophication, siltation, pesticide spills anddisturbance will allow environmental data to becorrelated with biological effects.

• The performance (including behavioural andreproductive success) of hatchery or zoo-reared fish priorto any reintroduction to the wild must be evaluated.

4.8 Barndoor skate

4.8.1 BackgroundThe barndoor skate, Dipturus laevis, is a large-bodied,wide-ranging marine fish found in coastal regions anddeeper waters from North Carolina to the Grand Banksoff Newfoundland. It was abundant up until the 1950s,but since then has undergone severe declines to the pointwhere it was considered vulnerable to extinction (Casey &Myers 1998). Fisheries have been the clear driver of itsdecline, and its recovery is probably hampered bymortality when it is caught as unwanted bycatch ofcommercial fisheries aimed at other species.

4.8.2 Valued attributes Two different communities of people are considered inthis example. The first is concerned with speciesconservation. They include the people who proposed thatthe species should be listed under the US EndangeredSpecies Act. Their primary value is a reduced risk of

extinction, though of course they will probably also beconcerned about any negative consequences of recoveryprogrammes for local fisheries. The goal would be for thespecies to attain a large enough population size to makeit unlikely to become extinct within a time period ofcenturies. This value has two implications for anassessment. First, it implies a risk-averse attitude toassessment whereby an over-estimate of population sizeand viability is clearly unacceptable, whereas some under-estimate would not be as bad. Second, precise estimatesof population size and extinction risk would be ideal, butit might also be acceptable to have sampling strategy thatinvolves monitoring a small part of the range so long asthis could be representative.

The second community is involved in fisheries. They mightbe less concerned about extinction risk per se, but wantthe species to show sufficient recovery to avoid harshmeasures being imposed on their fishing activities as aresult of concerns about skates being taken in bycatches.Ultimately they might like there to be a sufficient numberthat a barndoor skate fishery could be opened. In contrastto species conservationists, they may not mind over-estimates of population size. However, they would bevery concerned about over-estimates of extinction risk,which could lead to unnecessary curtailment of theirfishing. People who manage the local fisheries on theirbehalf might therefore be content with index or a sample,but the level of accuracy required would be greater if theresults started to suggest a need for reduced fishing.

4.8.3 KnowledgeBarndoor skates are known to inhabit shallow waters aswell as deeper regions that are currently beyond the reachof fishing boats. Little is known about the biology of thisspecies, though educated guesses are possible based onrelated species. Thus, it probably matures at about age11, and it may produce about 50 eggs per year.

4.8.4 Model A simulation model of the population could be developedand used to produce a population viability forecast.Ideally an age- or stage-structured model would be used.This should also have a spatial component, to reflectspatial variation in densities and to inform decisions aboutpotential area closures. It would be important to modelthe effects drivers and pressures, such as habitatalteration, fishing pressure, and bycatches ondemographic rates. There might be a specific need to lookat size and age structure in bycatch alongside informationon size-related breeding structure.

4.8.5 Relevant questions identified by the scopingstage• What is the current population size and recent

population trend of the barndoor skate?• What is the probability of extinction and expected

population size within a specified time period underexisting and forecast conditions?



Figure 4.1 Abundance of the barndoor skate from the Gulf of Maine to southern New England (Dulvy et al 2003)

Figure 4.2 Same time series as in Figure 4.1, restricted to the period 1980–2000 to clarify recent trends in abundance ofthe barndoor skate. a) Statistically significant increase in mean abundance (regression: P = 0.002). b) Same data as in a)but with 95% confidence intervals, showing an overlap with zero abundance in most years

0

1

2

3

1960 1970 1980 1990 2000

Year

Barn

door

ska

te s

urve

ybi

omas

s in

dex

(kg

per

traw

l)

0

0.05

0.1

0.15

1980 1990 2000

a.

Barn

door

ska

te in

dex

(kg

per

traw

l)

Year

-0.05

0.05

0.15

0.25

0.35

1980 1990 2000

b.

Barn

door

ska

te in

dex

(kg

per

traw

l)

Year


• How would specified changes to management affect theprobability of extinction and expected population size?

4.8.6 Sampling strategy, data gathering and analysisThis case requires estimates of population sizes andtrends. These may need to be quite precise to satisfy thosewho wish to estimate risk of extinction, whereas lessprecise indices may suffice for those concerned mainlywith overall trends. This species illustrates the typicaldifficulties of sampling rare organisms, especially in thesea, where surveys can be very expensive, and it is difficultto stratify surveys according to habitats because habitatsare not well known.

Dedicated surveys using trawlers can be used to estimateabundance if catches per unit effort can be scaled up topopulation sizes by estimating catchability. Observers onfishing boats can help estimate bycatches in commercialfisheries. Habitat surveys and environmental monitoringcan be used to match fish survey data to habitats, andthereby scale up to regional estimates. Note that the long-lived nature of this species means that fish surveys can berepeated over longer time periods than would be necessarywith species that have shorter generation times.

The recent data in Figure 4.1 seem to indicate a slightupward trend in the numbers of this species, based onsurveys. This is brought out more clearly by a reanalysis inFigure 4.2a. Based partly on this increase, it wasconcluded that there was ‘no evidence that they were indanger of extinction or likely to become endangeredwithin the foreseeable future’ (NEFSC 2000). Thisinterpretation of the data as well as the discovery of thefish in deeper waters than previously considered led tothe denial of a proposal to list this species under the USEndangered Species Act. However, consideration of thevariation around the estimates from the surveys suggestsa less optimistic interpretation (Figure 4.2b). The 95%confidence intervals include zero abundance in most ofthe last 15 years (Dulvy et al 2003). Furthermore, thesedata include adults and juveniles combined, whereas itwould be more informative to examine population trendswithin age classes, especially adults.

These results show how difficult it can be to give clearanswers about trends in biodiversity when individuals arerare. Unless a massive and expensive survey programme islaunched, it is unlikely that biologists will be able to saywith any certainty whether or not this species isrecovering by 2010.

4.8.7 Science gaps• Existing survey data is severely lacking in precision.

This is inevitable for marine surveys of rare fish species. • It is very difficult to forecast population trends of

species like the Barndoor Skate, when little basicinformation about life histories and behaviour isknown, especially if the data do not distinguishbetween adults and juveniles.

• A frequent problem with marine species is lack ofinformation about habitat requirements.

4.9 Trinidadian Guppy

4.9.1 Background The Trinidadian Guppy, Poecilia reticulata, has become amodel system for testing theories in evolutionary biology(Houde 1997). The species is native to Trinidad andTobago and NE South America and is widely distributedamong freshwater habitats there. Caryl Haskins, workingin the 1940s and 1950s, was the first person to observethat populations of guppies differed in a predictablemanner and that this variation is correlated with theseverity of the predation risk (Haskins et al 1961).Subsequent researchers have shown that a shift in theintensity of predation leads to heritable changes in arange of traits. In other words the system can be used todemonstrate evolution in action over tractable timescales.For example, when a population is released from severepredation males become more colourful, females havefewer but larger offspring and the intensity of schoolingbehaviour declines within 10 to 100 generations (Endler1995; Magurran 1998; Reznick et al 1990). Moreover,recent work has revealed that the degree of heritablevariation in populations is correlated with molecularvariation. This variation is superimposed upon markedgenetic divergence amongst populations, itself a legacyof geological events.

4.9.2 Valued attributesThe primary value of this system lies in the opportunities itoffers to evolutionary biologists. At present more than 20different groups of scientists (from North America andEurope) use the species to answer questions in naturaland sexual selection. Over 250 papers have beenpublished on the system since 1981. However, the guppysystem is also of value to Trinidadians since income isgenerated through scientific research and because it is avehicle for knowledge transfer from visiting to localbiologists. Finally, the wild fish provide genetic materialfor the ornamental fish trade.

4.9.3 KnowledgeThe guppy is not an endangered species in theconventional sense – population sizes may be large andthe species is found in every freshwater habitat from clearmountain streams to turbid brackish pools. However, therich diversity of populations is threatened. One of thegreat strengths of the guppy system is that contrastsbetween predation regimes (and other ecologicalvariables) are replicated across rivers and between guppyclades or subspecies. This greatly enhances the power ofthe system in hypothesis testing. But, like many otherfreshwater systems, Trinidadian Rivers are subject topollution – including industrial, domestic and agriculturaleffluent, disturbance, water abstraction, exotic


introductions and periodic flooding (exacerbated byforest clearance). Furthermore, scientists can impact thevery biological diversity they come to study by over-collecting at key sites, and moving fish between localities.Artificial transplants have proved to be an important toolfor deducing evolutionary rates but at the same time leadto irreversible changes to guppy population genetics andto the ecology of the manipulated streams.

4.9.4 ModelA simulation model of the fish population is used to carryout population viability analysis (PVA). This approach isespecially relevant since researchers are primarilyinterested in the small, low-fecundity populations foundnear the upper altitudinal range of the species. Suchpopulations typically experience large fluctuations in sizeand sex ratio. The guppy system can thus be used tomodel the consequences of various biotic and abioticfactors on vulnerable populations. The model can also beexpanded to include the effects of gene flow on heritablelocal adaptation.

4.9.5 Relevant questions identified by the scopingstage• How do biotic and abiotic factors affect the size and

distribution of the guppy population and its long-termviability?

• How does gene flow among sub-populations affectdemographic and evolutionary processes?

• How do fates of populations experiencing different ratesof immigration vary? Since the balance between femalechoice and sexual coercion varies amongst populationsmodels of gene flow must in turn be informed bymodels of sexual selection and sperm competition andinformation on individual mating tactics.

• Does molecular variation reflect heritable variation?The guppy system provides important opportunitiesfor further testing this.

4.9.6 Sampling strategy, data gathering and analysisThe goal must be to collect consistent information on awide range of populations. As many as possible of thefollowing items should be recorded on each occasion.Ideally there should be a stratified sampling programmeto cover the diversity of guppy habitats but in practicesamples are likely to be restricted to the sites that guppybiologists habitually visit. Sampling should also bereplicated over time so that the temporal dynamics ofpopulations can be assessed.

• Grid reference of site• Population size • Sex ratio• Age/size structure of population• Presence (and if possible abundance) of other fish

species and key invertebrates• Basic habitat descriptors – river width, depth,

temperature, cover• Water quality descriptors – pH, turbidity, biochemical

oxygen demand (BOD), nutrients, productivity,pollutants

• Description of any impacts such as quarrying, sewageoutfall, pesticide use

• Mark-recapture studies to estimate mortality risk,migration rate.

• Molecular analysis to include nuclear andmitochondrial DNA markers

• Record of precise number of fish collected from and/orintroduced to each site.

Many of these data are already routinely collected but arestored in field notebooks and individual computers. Thechallenge therefore is to ensure that they are entered in asingle database and that all researchers collect anddeposit information in a standard manner. The database,as described above, would provide rich opportunities forcomparative analyses of populations using standardapproaches in ecology and evolution. Its main purposehowever would be to generate guidelines for theconservation of guppy populations in Trinidad. Theseguidelines would be developed through consultationwith Trinidadian biologists and international researchers.

4.9.7 Science gaps• To date few studies have been able to disentangle

genetic diversity and population size when assessingpopulation viability. This system could be used toaddress this important science gap.

4.10 Brown argus butterfly

4.10.1 BackgroundBrown argus butterflies (Aricia) are members of a smallgroup of related European butterflies whose taxonomicstatus in Britain is still a matter of controversy. Geneticand ecological work has investigated fragmentedpopulations of this group in order to understand theirtaxonomic status and the role of adaptive evolution inmaintaining genetic diversity at the species boundary.

4.10.2 Valued attributesBrown argus butterflies are locally distributed andstrongly associated with chalk downlands and limestonegrasslands, where one of the main larval food-plants, therockrose Helianthemum nummularium, growsabundantly (Thomas & Lewington 1991). High speciesdiversity of these calcareous grasslands is maintainedunder traditional grazing regimes: knowledge of theecology of brown argus and other butterfly species iscritical for grassland management that is designed toconserve both plant and animal species.

Approximately 30–40% of so-called ‘species-level’diversity of European butterflies consists of closely relatedsister species that are either partially (parapatric) or fully(allopatric) geographically separated. These sister species


either do or potentially could interbreed. Forconservationists, this leads to the question of how distinctforms are maintained in areas of contact between relatedspecies.

4.10.3 KnowledgeOriginally, Aricia from Scotland and northern Englandwas classified as the only endemic British butterflyspecies, Aricia artaxerxes. Later, northern and southernBritish forms were amalgamated as subspecies within theEuropean Aricia agestis. However, most recent treatmentsseparate them again (e.g. Thomas & Lewington 1991).Based on genetic data from mitochondrial DNA, bothspecies as currently envisaged have wide distributions inEurope (Aagaard et al 2002), so that the northern BritishA. artaxerxes is no longer viewed as an endemic Britishspecies, in spite of some differences in colour patternfrom the European forms.

The two species differ mainly in ecology: northern A. artaxerxes is single-brooded, flying in June and July,while southern A. agestis is typically double-brooded,flying in May and early June, with a second brood inAugust (Thomas & Lewington 1991). The number ofbroods is an adaptation to prevailing temperature: twobroods are achieved in the south, but only one brood ispossible in the shorter northern growing season.

In North Wales, both single-brooded and double-broodedforms occur across a patchwork, or ‘metapopulation’, ofdifferent limestone grassland sites. Recent mitochondrialDNA studies have shown that both brood types havesouthern mitochondrial DNA haplotypes, suggesting thatall forms in north Wales belong to Aricia agestis (Aagaardet al 2002; Wynne et al submitted). However, it may bethat the area represents an ancient hybrid zone betweenthe two taxa, and the northern mitochondrial haplotypehas been simply lost by genetic drift. Thus genetic andecological studies in this area would be of great interest tounderstand both the taxonomic status of a potentiallyendemic British taxon, and also the adaptive evolutionthat may have led to their divergence.

4.10.4 ModelConsideration of how genetic diversity is maintained andcan be conserved in this system requires identification of:• Whether single- and double-brooded forms differ

genetically.• Whether the forms are ‘adaptive’, in the sense of

occurring in habitats where the number of broods peryear matches the local microclimate.

• The extent to which genetic differences betweendifferent populations depend on their connectednessto other populations.

• How landscape-scale patterns of genetic variationreflect metapopulation dynamics.

• How a species achieves distinct adaptations (one ortwo generations per year) in the face of continuousenvironmental variation (temperature).

4.10.5 Relevant questions generated by the scopingstageA key need is to characterise and understand climaticadaptation of Aricia butterflies in terms of the numbers ofgenerations per year. The taxonomic status of the twoforms is also of interest, because two nominal speciesoccurring in the UK have previously been recognised onthe basis of whether they have one or two generationsper year. Understanding this system will show howdistinct forms are maintained in areas of contact betweenrelated taxa, and how extinction/colonisation eventsaffect this adaptive response in a taxonomically relevantcharacter.

4.10.6 Sampling strategy, data gathering andanalysisAll Aricia populations across mainland North Wales aremapped and visited repeatedly, and population densitiesare assessed. Populations are always either single-brooded or double-brooded; no population includedindividuals from both brood types (Wilson et al 2002;Wynne et al submitted). Population sizes and distancesbetween populations are recorded to give an estimate ofhow connected populations are to other populations andgroups of populations. Temperature is monitoredcontinuously, and these data are used to develop astatistical model that provides an estimate of the thermalenvironment for each patch suitable for the butterflies.Samples of key populations are studied genetically usingseven enzyme loci.

Double-brooded populations are typically found in thewarmer habitat patches, while single-brooded populationsare found in cooler habitats, as expected. Genetically,populations of the two brood types are not clearlydifferentiated in enzyme alleles; instead there is an overalleffect of physical distance on genetic distance. Caterpillarsfrom single- and double-brooded populations differ in theirresponses to day length in captivity: the caterpillars ofsingle-brooded populations enter an over-winteringdiapause when exposed mid-summer day lengths, whereasdouble-brooded caterpillars continue to grow.

However, the match between the number of broods andthe thermal environment is not perfect. Populations nearto large populations of a particular brood structure tendto show the brood structure of those populations, ratherthan the optimal brood structure for the particularthermal microenvironment in which they were found.These mismatched populations also have generally lowerdensities, as expected if poorly adapted to the localclimate (Wynne et al submitted). These maladaptivepopulations are almost certainly a result ofextinction/recolonisation turnover, which, according tosimulations, is expected to be much faster than localadaptation. The likelihood of rapid population turnover isalso suggested by genetic results: isolated populations aremore differentiated genetically than populations near tolarge networks of populations, suggestive of drift caused


by colonisation bottlenecks. Perhaps surprisingly, localconnectivity of populations acts to prevent localadaptation over distances of tens of kilometres, versustypical individual movements of much less than 1km(Wynne et al submitted). This appears to be due to theimportance of rare long-distance movements inrecolonising extinction-prone populations.

4.10.7 Science gaps• Long-term monitoring (say 20 years or more) of very

large numbers of populations is required to test andrefine estimates of colonisations and extinctions basedon simulations. Such data are rare, but essential forunderstanding the population biology and adaptabilityof such species.

• As well as documenting newly colonised habitatpatches, it would also be extremely useful to knowfrom which source populations successfulcolonisations originated. Genetic studies will hardlyhelp in this matter, since the populations studied herewere only weakly differentiated, and colonization maycause a bottleneck in population size that can radicallychange gene frequencies.

• New molecular markers characteristic of A. artaxerxesand A. agestis will be needed to assess haplotypesacross the species range, focusing on N. Wales.Recently a sex-linked marker has been found whichhas characteristic haplotypes in these two forms(Wynne et al submitted). Polymorphism in some N.England single-brooded populations suggests thatancient hybrid zones may indeed exist.

• To clinch understanding of climatic adaptation byAricia in N. Wales would require finding the genesresponsible for switching from single to double-brooded phenology. Independent evolution of single-broodedness in N. Wales and in A. artaxerxes in N.England and Scotland should be detectable viasequence differences at these loci. Because of thedifficulty of isolating such genes, it will be a long timebefore this particular science gap is filled.

4.11 The tiger in India

4.11.1 BackgroundDuring the 1970s, the Indian government made acommitment to protect the tiger in India. ‘Project Tiger’was launched, and focussed on the management of thespecies in 27 Project Tiger Reserves distributed across thecountry. Thirty years on the programme has been anational success; over 50% of the world’s tigers still live inIndia, one of the most densely populated counties onEarth with almost a billion people inhabiting a deeplyfragmented habitat and facing high levels of poverty.

4.11.2 Valued attributesProject Tiger is a continuing commitment with a budgetapproved every five years by the Indian Federal

Government. The main objective of the project is toensure maintenance of a viable population of tigers inIndia for scientific, economic, aesthetic, cultural andecological values. As one of the main threats to tigers ishabitat loss a second objective of the project is to protectareas of national biological importance that will also be ofbenefit, education and enjoyment of the people. Thegovernment and Indian biologists therefore wish to knowwhether, and where, the population of tigers isincreasing, stable or declining.

4.11.3 KnowledgeTigers continue to face a variety of threats, even withinthe reserves. Habitat deterioration, prey depletion,poaching for trade, and direct persecution by people areall continuing problems. However, tigers have relativelyhigh fecundity, and the potential to recover fromdepletion as long as there is adequate prey and suitablehabitat. Political will and commitment for Project Tigerhave declined over time (Thapar 1999). Some biologistsnow question the methods used to monitor tigers acrosstheir range (Karanth 1999; Karanth et al 2003).

Over the last 30 years the status of the tiger has beenassessed by a regular nation-wide ‘pugmark census’. Overa period of no more than a couple of weeks, thousands ofgovernment staff search tiger habitats and gather plastercast imprints from the left hind foot pugmark of eachtiger track located. A central database, individualrecognition from the pugmarks and reconciliation acrossgeographical areas result in an estimate of the tigerpopulation number for the country as a whole. Thismethod is therefore an attempt to census the entirepopulation in India. The variation in numbers betweeneach census forms the basis for future management andfunding of the programme.

Field studies on wild tigers, in recent years, havegenerated new knowledge about their ecology andbehaviour, and methods for population estimation havebecome more sophisticated. These advances can informthe choice of methods used in the Indian national tigercensus.

Tigers may be quite polygynous with overlap in theranges of breeding females within male territories,common routes followed by adjacent breeding males,and both female and male transient individuals that moveacross breeding territories. In the census therefore, notonly must tracks from each individual tiger be located,regardless of the accessibility of the area, but trackersmust also be able to distinguish between differentindividuals who may be following the same tracks. Thismust pose difficulties in all habitats but will be especiallyproblematic in areas where the substrate is inappropriate,and be almost impossible in areas that are especiallyremote or inaccessible. Hence, is it appropriate toundertake the full census and to estimate the size of theentire tiger population when there are so many


uncertainties, especially as the method used does notallow any estimate of the uncertainty associated with theresulting population number? (Karanth et al 2003).

4.11.4 Model The tiger is widely distributed across India, but dependson local reserves for its persistence. Local extinctionwithin reserves, the loss of migration routes and thefragmentation of the habitat could all lead to its rapiddecline. Absolute measures of population numbers aretherefore less informative for responsive conservationplanning than effective monitoring of the range extent,and of the trends in population size in key reserves.

4.11.5 Relevant questions identified at the scopingstageThree goals for tiger monitoring have therefore beenstated (Karanth & Nichols 2002; Karanth et al 2003):

• How is the countrywide distribution of the specieschanging over time? This can help to assess thenegative effects of habitat fragmentation and localextinction versus the positive outcomes ofconservation actions leading to range expansion.

• What are the observed trends in local abundance inindividual reserves? This will help managers to assesswhether the status locally is improving, deterioratingor stable. This measure allows conservation actions,and where necessary, more detailed monitoring to beappropriate targeted.

• What is the absolute abundance of tigers, especially atpriority sites?

4.11.6 Sampling strategy, data gathering andanalysis At the national level (>300,000 km2 of tiger habitat), thepresence or absence of tigers could be recorded annuallyby geo-referenced, statistically rigorous sampling surveysrecording presence or absence of tiger tracks and othersigns. This would be substantially less effort than thepugmark census but yield annually a map of the extent oftiger populations with associated uncertainly levels.

Within reserves, managers need to monitor localpopulations to assess the effectiveness of conservationmeasures. But estimating population density may beprohibitively difficult and expensive, especially on theannual basis that is recommended for management(Karanth et al 2003). Instead, an index of relative densitycould be developed, which might not be easily translatedto population abundance measures but would berelatively cheap, objective and replicable, and provideuseful information for managers. The index might be

derived from standardised encounter rates of sign, suchas the number of tiger track sets or tiger scatsencountered per 10km walked, or the proportion of 1kmlong trail sections in which tiger sign was detected.

Where resources allow, or where the low-intensitymethods above suggest more knowledge is required,absolute numbers of tigers, and their age and sex ratiosmay need to be estimated. Land transect surveys andcamera trap capture-recapture surveys are possiblemethods to use here.

4.11.7 Science gapsThis example highlights the importance of choosing theappropriate monitoring method for the context. Limitedfinancial and skill resources can be used to better effectwhen information on the species and relevant statisticaltechniques are incorporated into the design of themonitoring programme.

Further information on the breeding and ranging behaviourof different age and sex individuals, an improvedunderstanding of predator-prey relationships, and bettertechniques for monitoring prey abundance are all needed toimprove tiger monitoring and thereby its conservationmanagement (Seidensticker et al 1999).

4.12 Case study conclusions

The 11 case studies presented above show that theframework presented in Section 3 can be used to guide awide variety of biodiversity measurements. It does nomore than represent best science practice but its value isthat it helps promote careful dialogue between scientistsand stakeholders during the preliminary scoping phase, ithelps uncover assumptions about the nature of thesystem (the initial ‘model’), and this information, in turn,informs the sampling strategy, data gathering andanalysis. All of the case studies found major science gapsthat ideally, would be filled in order to provide completepictures of the state of biodiversity in each study.However, in summarising prior knowledge andarticulating conceptual models of the system, it oftenbecomes apparent that a great deal is known about eachcase, which could usefully inform practical decisionsabout what to sample, where to sample, and how often.Use of the biodiversity measurement framework wouldtherefore help those engaged in biodiversity assessmentmake the best use of the time and resources available inseeking to provide answers to the questions being askedby the interested parties.


Clear evidence and widespread scientific consensusindicate that losses of biodiversity have accelerated overthe last two centuries as a direct and indirectconsequence of human population growth,unsustainable patterns of resource consumption andassociated environmental changes, such aseutrophication and the effects of alien invasive species.Effective methods of measuring biodiversity are neededto monitor its condition and to measure progress towardthe target established at the World Summit onSustainable Development in 2002 of achieving ‘asignificant reduction in the current rate of biodiversity lossby 2010’. Despite data insufficiencies, substantialprogress has been achieved in developing andimplementing conservation techniques and policies toprotect biodiversity. However, no sound basis currentlyexists for assessing performance against the WSSDobjectives. Without internationally agreed measures itwill be impossible to determine whether losses ofbiodiversity are declining or accelerating. It will thereforebe impossible to assess the success of mitigating actions.

A broad suite of measures is necessary to monitorchanges in biodiversity through time, and to assess thesuccess of conservation and sustainability initiatives. Suchmeasures need to be implemented cost-effectively andmust be scientifically sound. However, a significantimpediment to assessing the overall state of nature, nowand in the future, is the extraordinarily limited knowledgeof many aspects of the biodiversity with which we sharethe planet, and upon which we all depend for criticalgoods and services. Most significantly, of all the speciespresent on earth, possibly only one in ten are known toscience. The fates of organisms that have not yet beenscientifically described cannot be measured. Likewise,how ecosystems function cannot be understood untilmore is known about the organisms of which they arecomprised. Several key deficiencies in knowledge exist:

• For most species that have been described little ornothing is known of their distribution, ecology,population sizes or evolutionary history.

• Knowledge of biodiversity is most limited and patchyfor the very geographic areas and biomes wherespecies diversity is greatest – principally in the tropics,and next to nothing is known of the biodiversity of thedeep sea.

• Reliable estimates of current conservation status areavailable for only a tiny proportion of species and arebiased toward certain groups (birds and mammals) andcertain parts of the world (temperate regions).

• Understanding of trends in the state of biodiversity ishampered by the absence of reliable baseline data formost groups and habitats as well as inconsistencies inmethodology over space and time.

• Knowledge of the benefits to people from biodiversity

remains limited, especially in terms of the value ofeconomic services.

• Analyses are needed of the gains and losses incurred asrelatively pristine habitats are transformed andbiodiversity is lost.

Measures of biodiversity vary in scale and purpose andcan extend beyond the species level to encompass entirehabitats and ecosystems or focus on details ofpopulations and genes. No one measure is best for allpurposes. A broad suite of measures is necessary to meetparticular needs but the sheer multiplicity of currentmeasures contributes to the difficulty of building publicawareness and understanding. Selecting appropriatemeasures requires a careful consideration of the purposeof the assessment as well as the tradeoffs betweenusefulness, completeness and required effort in terms oftime and resources. Although commonly used,measurements of extinction rates for species are aninherently poor way to monitor biodiversity loss.Population data, although expensive to collect, canprovide a more sensitive measure of a particular speciesover the short term and, depending on the specieschosen, can often have considerable resonance with thepolicy-makers and the public.

Detailed measures of many key aspects of biodiversity arelimited by resources, but effective sampling strategies andapplication of new technologies could transform thecurrent knowledge of changes in habitat types, patternsand rates of delivery of ecosystem services, distributionsof specific taxa and changes in population abundance.

For many geographic areas, important information onhistorical status of habitats, species and ecosystems residesin museums, libraries and informal records. Making thisinformation available as a basis for assessing trends andestablishing time series would greatly increase the value ofdata being collected now. Transferring taxonomicinformation into an accessible form, for example throughthe use of the Internet, appropriate information technologyor as user-friendly guides, will greatly enhance conservationefforts. A list of taxonomic experts, who are able to respondeither remotely or directly, could also support situationswhere advice is required urgently.

Despite the difficulties of measuring biodiversity, andcurrent inadequacies, enough is known about the state ofglobal biodiversity to say with confidence thatunprecedented rapid losses of biodiversity are occurring.What are needed are better measures of rates of losstogether with information on the geographic areas,habitats and groups of organisms where these losses areconcentrated. Better measures, based on sound science,will help assess success in managing biodiversity andpreventing further losses.

5 Conclusions and recommendations


We therefore make the followingrecommendations:

• The framework for biodiversity assessment presentedin this report should be applied routinely by thosecommissioning, funding and undertaking biodiversitymeasurements. As the case studies presented in thisreport (Section 4) demonstrate, the framework can beused for terrestrial, freshwater and marine systems,and at the ecosystem, species and population levels.We also believe it is applicable to situations rangingfrom large, long-term studies to instances where arapid response is required, and can accommodatedifferences in the timescales of stakeholder interests.Application of the framework would help ensurestakeholder involvement and that measures are fit forthe purpose to which they are being applied. It wouldalso help to identify weaknesses in some currentapproaches as well as major science and informationgaps.

• Urgent emphasis on synthesis is needed by thescientific community to make otherwise scattered datamore readily available and more useful. This needs tobe accompanied by a more favourable attitudetowards such projects by funding bodies and morewidespread use of web-based technology for moreeffective dissemination of information. Synthesis willquickly reveal key gaps in knowledge, which shouldthen be addressed by the development of realistic newprogrammes capable of delivering substantialimprovements in knowledge of otherwise poorlyunderstood geographic areas, habitats and groups oforganisms. Such programmes must be implementedurgently with realistic goals for completion. It is crucialthat they are completed in the course of the next threeto seven years.

• The scientific community should focus on thedevelopment of data gathering and analyticaltechniques to provide biodiversity information that isboth relevant and organised for efficiency. This willinvolve: consideration of sampling strategies (both

sample sizes and appropriate stratification);assessment and integration of the relevant drivers ofchange, including input from the social sciences;better information on the values ascribed tobiodiversity by different stakeholders; effectivedeployment of new techniques from moleculargenetics, bioinformatics, remote sensing and e-science; as well as consideration of the role ofvolunteers and informal methods of data gathering.

• We recommend enhancing levels of taxonomictraining and linking such training more directly to theongoing measurement and management ofbiodiversity. Increasing scientific and technical capacityin countries with high biodiversity is crucial. It isespecially important to increase the number ofprofessional taxonomists for key groups of organisms,and to ease the problems of identifying a broad rangeof organisms in the field by the more effective use ofappropriate information technology. Low costapproaches to facilitate identification can also beextremely effective. Maximising the efficiency withwhich the information generated by systematists istransferred and made useful to biologists in the field iscrucial.

• The international community and intergovernmentalorganisations should undertake a review of currentprogrammes for biodiversity status assessment,especially at a global level. Existing monitoringprogrammes that are already contributing to, ordelivering, robust, global assessments of biodiversitymust continue. Where possible they should beenhanced and extended. Across both new and existingprogrammes, there should be a particular focus onestablishing a baseline and rates of change so thatprogress towards reducing rates of biodiversity loss by2010 can be measured. Expanding existing monitoringprogrammes and developing new assessments willrequire a marked increase in funding as well as adegree of co-ordination and co-operation amongNGOs, academics, and governmental andintergovernmental agencies.


Name of submitter Organisation

Dr Tundi Agardy Sound Seas USADr Mark Avery Royal Society for the Protection of Birds (RSPB)Dr Harald Beck University of MiamiDr Andrea Belgrano University of New MexicoDr Colin Bibby Birdlife International, Biodiversity Conservation Information System (BCIS)Mr David Brackett Species Survival CommissionDr Ben ten Brink Research for Man and Environment (RIVM)Dr Colin Catto Bat Conservation TrustDr Geoffrey Chapman Personal submission Professor Andrew Clarke British Antarctic SurveyDame Barbara Clayton University of SouthamptonMr Jonathan Cowie Institute of BiologyDr Ben Delbaere European Centre for Nature Conservation (ECNC)Dr Keith Duff English NatureDr Nicholas Dulvy The Centre for Environment, Fisheries and Aquaculture Science (CEFAS)Dr Sam Fanshawe Marine Conservation Society (MCS)Dr Alan Feest Bristol UniversityDr Gustavo Fonseca Center for Applied Biodiversity Science (CABS) – Conservation International (CI)Dr Ron Fraser Society of General MicrobiologyDr Nick Gotelli University of VermontProfessor Jeremy Greenwood British Trust for Ornithology (BTO) Dr Richard Gregory Royal Society for the Protection of Birds (RSPB)Ms Melanie Heath Birdlife International Mr Colin Hedley Country Landowners Association Professor Mac Hunter Society of Conservation BiologyMs Janet Hurst Society of General MicrobiologyMr John Jackson Natural History MuseumMr David Mansell-Moullin International Petroleum Industry Environmental Conservation Association (IPIECA) –

Energy and Biodiversity InitiativeMs Rebecca May World Wide Fund for Nature (WWF)Dr Dorian Moss Centre for Ecology and Hydrology (CEH) – Monks WoodMr Kalemani Mulongoy Principal Officer, Secretariat of the Convention on Biological Diversity, Scientific,

Technical and Technological MattersProfessor Norman Myers Personal submissionMs Louisa Ogry Nakanuku Environmental Information Service (EIS) NamibiaDr Adrian Newton United Nations Environment Programme – World Conservation and Monitoring Centre

(UNEP–WCMC)Professor Aharon Oren International Committee on Systematics of ProkaryotesDr Kent Redford Wildlife Conservation InstituteMr Nick Reeves Chartered Institution of Water and Environmental ManagementProfessor John Richards University of NewcastleDr Paul Rose Joint Nature Conservancy Commission (JNCC)Dr Bernhard Schmid Institut fuer UmweltwissenschaftenDr Martin Sharman European Commission Dr Mark Shaw National Museums of Scotland Dr Phil Shaw Scottish Natural Heritage (SNH)Dr Andrew Sier UK Environmental Change Network (UK ECN)Sir David Smith FRS Linnean SocietyDr Malcom Smith Countryside Council for Wales (CCW)Mr Richard Smithers Woodland TrustDr Jorge Soberon Conabio

Annex A List of submissions to call for views

The working group sought the views from a variety of organisations and individuals. The working group is grateful toall who responded; they are identified below.

Dr Bruce Stein NatureServeDr Simon Stuart IUCN Species Survival CommissionProfessor Albert van Jaarsveld Centre for Environmental Studies, University of Pretoria Dr Allan Watt Centre for Ecology and Hydrology (CEH) Banchory



Annex B List of participants in the consultation meetings

To inform the study and discuss initial findings, the working group hosted two workshops, 25 November and 13December 2003, inviting UK and international academics, policy makers and representatives from industry,conservation and non-governmental organisations (NGO). The working group are grateful to all those who attended;they are listed below.

Name of submitter OrganisationProfessor Michael Akam FRS Zoology Museum, CambridgeMr Jonathan Bass Centre for Ecology and Hydrology (CEH) DorsetProfessor Richard Bateman Natural History MuseumDr Harald Beck University of MiamiDr Colin Bibby BirdLife International, Biodiversity Conservation Information System (BCIS) Professor Ian Boyd University of St AndrewsMr David Brackett Species Survival CommissionDr Ben ten Brink Research for Man and Environment (RIVM)Dr Thomas Brooks Conservation International (CI)Mr Neil Burgess World Wide Fund for Nature – US (WWF)Mr Martin Capstick Department for the Environment Food and Rural Affairs (DEFRA)Dr Kevin Charman English NatureMr Mark Collins United Nations Environment Programme – World Conservation and Monitoring

Centre (UNEP–WCMC)Ms Mireille de Heer UK Permanent Representation to the European UnionMr Mark Diamond Environment AgencyDr Keith Duff English Nature Dr Nick Dulvy The Centre for Environment, Fisheries and Aquaculture Science (CEFAS)Dr Mark Eaton Royal Society for the Protection of Birds (RSPB)Professor Dianne Edwards FRS Cardiff University Dr Alan Feest Bristol UniversityDr Brian Ford-Lloyd University of Birmingham Dr Ed Green United Nations Environment Programme (UNEP)Dr Michael Green Norfolk Broads AuthorityProfessor Jeremy Greenwood British Trust for Ornithology (BTO)Dr Andy Hector Imperial CollegeDr Colin Hindmarch Institute of BiologyProfessor Chris Humphries Natural History MuseumMr John Jackson Natural History MuseumDr K A Joysey Linnean SocietyDr Val Kapos United Nations Environment Programme – World Conservation and Monitoring

Centre (UNEP–WCMC)Dr Terry Langford Linnean SocietyProfessor Nigel Leader-Williams The University of Kent at CanterburyDr Kyrre Lekve University of OsloMr Jonathan Loh World Wide Fund for Nature – UK (WWF)Dr Nigel Maxted University of BirminghamDr Dave Roberts Natural History MuseumDr E J Milner-Gulland Imperial CollegeDr Dorian Moss Centre for Ecology and Hydrology (CEH) Monks WoodProfessor Norman Myers Fellow of Oxford UniversityDr Adrian Newton World Conservation Monitoring Centre (WCMC)Dr Mark O’Connell Wildfowl and Wetlands Trust (WWT)Dr Andy Purvis Imperial CollegeProfessor Paul Raffaelli University of York Dr Paul Raven Environment Agency Professor John Richards University of NewcastleDr Paul Rose Joint Nature Conservancy Society (JNCC)Dr Jane Sears Royal Society for the Protection of Birds (RSPB)Dr Mark Shaw National Museums of Scotland


Annex C Contributors of additional case studies

Dr Andrew Sier UK Environmental Change Network (UK ECN)Dr Malcolm Smith Countryside Council for Wales Mr Richard Smithers Woodland TrustDr Jorge Soberon ConabioDr Jean-Luc Solanot Marine Conservation Society (MCS)Dr Alistair Taylor Natural History MuseumDr Kate Trumper House of Commons, Committee SpecialistDr Paul Williams Natural History Museum

We are very grateful to the following people who provided invaluable assistance in utilising and developing theframework through authoring specific case studies for the report.

Professor Nigel Maxted and Professor Brian Ford-Lloyd Genetic diversity at the species level in wild lentilsUniversity of Birmingham

Professor James Mallet – UCL and Brown argus butterflyProfessor Chris Thomas – University of Leeds

Dr Mark Blaxter Meiofaunal indicators of pollutionUniversity of Edinburgh


Over the past several decades the fate of biodiversity hasbecome an important political issue that has beenaddressed in a variety of ways including through thedevelopment of international and national policy. ThisAnnex provides an overview of the policy situationinternationally, and in the UK, along with a briefconsideration of its relation to the NGO community andindustry.

D.1 International agreements and theConvention on Biological Diversity

A gradual increase in environmental awareness throughthe 1950s and 1960s resulted in the StockholmDeclaration at the United Nations Conference on thehuman environment, which outlined the need for ahealthy natural environment to ensure the well-being ofhumanity (United Nations 1972). In parallel, importantearly steps toward global protection of certain aspects ofbiodiversity were enshrined in the Convention onInternational Trade in Endangered Species of Wild Faunaand Flora (CITES) in 1973, which led from the WorldConservation Union (IUCN) resolution in 1963 to protectcertain species from the threats of unregulated trade.

In the following decades it became clear that speciesconservation needed to be integrated into a broaderecological and social context. This was emphasised in theWorld Conservation Strategy (IUCN 1980) and the WorldCharter for Nature (United Nations 1982). Mostsignificantly this broader view culminated in theBrundtland report (World Commission on Environmentand Development 1987), which introduced the conceptof ‘Sustainable Development’.

Subsequently the Convention on Biological Diversity(CBD) emerged from the Earth Summit, in Rio in 1992.The CBD went beyond existing conventions, such as theRamsar Convention on Wetlands (Ramsar 1971) andCITES, by emphasising the importance of protectingbiological diversity within the paradigm of sustainabledevelopment. The CBD has three main objectives: i) theconservation of biological diversity; ii) the sustainable useof the components of biological diversity; and iii) the fairand equitable sharing of benefits arising from theutilisation of genetic resources.

The CBD is designed to accommodate change andevolution. Scientific assessments, to assist this process,are provided by the Subsidiary Body on Scientific,Technical and Technological Advice (SBSTTA), whichreports back to the decision making biannual Conferenceof the Parties (COP). Regular assessments of the CBD,

incorporating expert scientific advice, enable the Partiesto set priorities for international work and strengthennational implementation.

Two international mechanisms support the work of theCBD. The Global Environment Facility (GEF) facilitates thedistribution of financial aid to biodiversity andconservation projects in developing countries, and theClearing House Mechanism (CHM) provides a virtualmeeting place for the exchange of information andknowledge, along with advising on the development ofstandards, formats and protocols.

The objectives of the CBD in relation to biodiversityconservation and monitoring are set out in Articles 6 and7 of the Convention. In recognition of the differentconditions and capabilities of the Parties, the objectivesare broad, requiring each country to develop their ownstrategy for implementation.

In 2001, a review of the current status of globalbiodiversity was published in the first edition of the GlobalBiodiversity Outlook (CBD 2001). Requested by the COPof the CBD, the report also analysed progress towards thethree objectives of the CBD. It highlighted that theabsence of clear objectives for the CBD, such as havingdefined targets, specific sites for conservation or lists ofprotected species, has generated problems in developingreporting standards for national achievements. Problemsof coordination have also frustrated efforts to set andmonitor global targets. The report emphasised that thelack of biodiversity information was one of the majorimpediments not only for reporting achievements butalso for the creation of meaningful targets against whichprogress could be measured. At COP 6 in April 2002, theparties adopted the Global Strategy for PlantConservation, thereby agreeing specific global targets forconservation (albeit for a single group of organisms) forthe first time.

One response to insufficiencies in the quantity and qualityof global biodiversity information was the endorsementof a programme of work for The Global TaxonomyInitiative (GTI) at COP 6 in 2002. The GTI specifies thateach Party should set up National Focal Points. In the UKthis role has been taken by the Natural History Museum.National Focal Points are responsible for linking withregional centres of taxonomic expertise around thecountry and where appropriate sharing their informationwith other countries. Similarly, in response to the EarthSummit in 1992, the OECD introduced the GlobalBiodiversity Information Facility (GBIF), which aims toprovide the infrastructure for an international mechanismto make biodiversity data and information universally

Annex D International agreements: policy responses to biodiversity loss


accessible. Various other initiatives are working towardsweb-based inventories of life on earth, for example, theCatalogue of Life by Species 2000, All Species andCODATA (Annex F).

D.2 At the European Union level

Responding to concerns about the loss of biodiversity, apan-European agreement, the Bern Convention on theConservation of European Wildlife and Natural Habitats,came into force in 1979. Since the adoption of the CBD,Europe has also introduced two strategies to help itsimplementation at a Member State level. They are thePan-European Biological and Landscape DiversityStrategy (PEBLDS) which strengthens existing Europeanconventions and the European Community BiodiversityStrategy (EC 1998), which introduces four sectoralAction Plans, namely Conservation of NaturalResources, Agriculture, Fisheries, and Economic Co-operation and Development. In addition, the Natura2000 programme, which is supported legislatively bythe European Birds (EEC 1979) and Habitats Directives(EEC 1992), outlines habitats and species for protectionand has been an important guide in creating priorityconservation areas within Europe. As these SpecialAreas of Conservation (SACs) have been identified anddesignated by Member States, a successful programmewill require the co-ordination and exchange of a highlevel of information between countries.

A limitation to the potential success of a pan-Europeanstrategy is that an over-arching framework does not existfor the production of formal and regular reports of thetrends and state of European biodiversity. The EuropeanBiodiversity Monitoring and Indicator Framework (EBMI-F)potentially provides a good basis for monitoring withinthe framework of the PEBLDS, but the EuropeanEnvironment Action Programme (EEAP), in its sixth actionplan for the period until 2010, highlights the need forextra funding and research for basic data collection.Without data on trends and the state of biodiversity, theEEAP stresses that the major environmental institutionswould be restricted in carrying out their work (EC 2001).

D.3 The United Kingdom

The UK ratified the CBD in 1992 and published its ownbiodiversity strategy in Biodiversity: the UK Action Plan(DoE 1994). The 59 objectives in the plan cover four mainareas: i) the development of action plans for key speciesand habitats; ii) monitoring systems; iii) access toinformation through biodiversity databases; and iv) publicawareness and involvement in contributing toconservation efforts.

Progress towards the targets for conservation andrecovery of both species and habitats was outlined in a

review of the UK Biodiversity Strategy and Action Plan (UKBiodiversity Group 2001). However, this found that therewere still declines in some of the identified priority speciesand one of the priority habitats. There was also a concernthat biodiversity issues and policy were not fullyintegrated into central and local government practices orfully factored into decisions throughout the business andindustrial sectors. The difficulty of assessing biodiversitystatus without access to sufficient information to makeinformed decisions was also noted.

In part, information shortages derive from a lack ofresources (both skills and finance) for the identification,description and classification of species. A decline insupport for systematic biology research within the UK wasfirst highlighted in the House of Lords Select Committeeinquiry into Systematic Biology Research (House of Lords1992), which expressed concern over its decline in theUK. It proposed a number of measures to rectify thesituation such as increased funding and further help tothe relevant institutions. Ten years later a similar study bythe House of Lords (House of Lords 2002) found that,despite several subsequent initiatives, the number ofscientists involved with systematic biology, in particulartaxonomy, had continued to decline. It stressed thatimprovement would only arise through a collaboration ofgovernment and scientists determined to address,prioritise and fund taxonomy.

More positively, recognition of the broad responsibility andpotential international contribution of UK science andconservation activities overseas has come through theDarwin Initiative. Established in 1992, this programmedraws on UK expertise in biodiversity and provides funds forprojects that create partnerships with organisations andscientists in less developed countries to conserve biodiversityand promote the sustainable use of natural resources. ThePrime Minister, Tony Blair, announced at the World Summiton Sustainable Development in 2002 that the UK intendedto increase funding for the Darwin Initiative to £7 million perannum by 2005. This new funding will be used to enhanceproject programmes, while also developing the capacity ofhost countries to be able to continue the projectsthemselves. To this end, it will also fund scholarships tobroaden the professional knowledge and experience inoverseas participants.

D.4 NGOs and industry

Without the awareness, participation and commitment ofbusiness and society, policies developed to combatbiodiversity loss and habitat decline will not be successful.Current political best practice is becoming more competentin involving all sectors earlier in decisions, and also inmaking use of the breadth of resources and experiencefound outside of government. For example, many NGO-ledinitiatives in the area of biodiversity management are nowtaken up and used by governments.


Industry and business are starting to understand theconcepts of sustainable resource use and the potentialcommercial value of biodiversity, both as a resource andin terms of generating consumer goodwill. In somecases, business has taken the lead in creatingbiodiversity action plans and developing indicators tomeasure and limit the impact of their activities. Suchaction plans have a two-fold effect, both makingbusiness practice more sustainable, and encouragingstakeholder participation in the company’s activities.The global coverage of many businesses means thatmuch could be achieved by promoting conservationwithin their activities. In some countries the scope for

business to undertake concerted conservation actionmay far outweigh that of national governments. It istherefore vital that responsibility is taken and thatpartnerships are forged in these areas.

For the activities of business and NGOs to be successful,baseline measurements and appropriate means formeasuring and monitoring biodiversity are essential.These will facilitate refinement of strategies and tactics,and the international co-ordination of conservationefforts. They will also facilitate the ongoing monitoring ofbiodiversity and provide a basis for recognising wheresuccess has been achieved.


Aquatic Macrophytes A plant that grows or has part of its life cycle in an aquatic systemBAP Biodiversity Action PlanBiome The classification of certain physical and chemical characteristics of an environment. Often

characterised by the dominant forms of plant life and the prevailing climate. Examplesinclude temperate forests and deserts

BOD Biochemical Oxygen DemandBTO British Trust for Ornithology CBD Convention on Biological DiversityCEFAS The Centre for Environment, Fisheries and Aquaculture ScienceCITES Convention on International Trade in Endangered Species of Wild Fauna and FloraCOP Conference of the PartiesCytotype Exhibiting identical cytological features (i.e. chromosomes) as those originally described for

the taxonDEFRA Department for Environment, Food and Rural AffairsDPSIR Driver-Pressure-State-Impact-Response FrameworkEBMI-F The European Biodiversity Monitoring and Indicator FrameworkEco-geographic Involving both the entire ecological and geographic range of a species.Eutrophication The process whereby water bodies become severely oxygen deficient as a result of

decomposing organic matter. Increased organic matter levels are often caused by pollutionFAO Food and Agriculture OrganisationGBIF Global Biodiversity Information SystemGEF Global Environmental FacilityGene bank A centre or institution that manages genetic resources, in particular, maintaining ex situ or in

situ collectionsGermplasm Often synonymous with ‘genetic material’, it is the name given to the total genetic

variability, represented by germ cells or seeds, available to a particular population oforganisms

Germplasm, ex situ Maintaining the genetic variability of a population in a different environment or geographic conservation location than where it evolved i.e. botanical gardens, breeding orchards, cold storage of

seed or pollen (seed banks)Germplasm, in situ Maintaining the genetic variability of a population in approximately the same geographic conservation and ecological conditions under which it evolvedGeo-referencing The task of establishing the relationship between page coordinates on a planar map and

known real world coordinatesGIS Geographic Information SystemGTI Global Taxonomy InitiativeHaplotype A set of closely linked genetic markers present on one chromosome which tend to be

inherited togetherICARDA Centre for Agricultural Research in the Dry AreasIPNI International Plant Names IndexIUCN World Conservation UnionMA Millennium Ecosystem AssessmentNCSR Number of Clusters per Sub RegionNERC Natural Environment Research CouncilNGO Non-Governmental OrganisationNomenclature Assignment of names to taxaOECD Organisation for Economic Co-operation and DevelopmentOligonucleotide A sequence of nucleotides. Nucleotides being organic molecules that constitute the

building blocks of genetic material such as DNA.PEBLDS Pan-European Biological and Landscape StrategyPhylogeny The relationship between groups of organisms that share a common ancestryProtists Microscopic, unicellular eukaryotes. As distinct from prokaryotes they have a membrane-

bound nucleus. PVA Population Viability AnalysisRSPB Royal Society for the Protection of BirdsSAC Special Areas of Conservation

Annex E Abbreviations and glossary


SBSTTA Subsidiary body on Scientific, Technical and Technological Advice (SBSTTA) for the CBDSynonyms The same taxa or species inadvertently given different names by different peopleSystematics The organisation of biological information using the combination of taxonomic science and

nomenclatureTaxa Groups of organisms distinguished through the science of taxonomy. A species is the

primary taxonomic unit (singular taxon)Taxonomy The science of classifying organisms into groups using information that reflects their natural

and evolutionary relationships Trophic level The level in the food chain at which an organism feeds. Primary producers such as

phytoplankton or grass, using photosynthesis to convert sunlight into biomass, are on thefirst trophic level.

TDWG Taxonomic Database Working GroupUNCED United Nations Conference on Environment and DevelopmentUNEP United Nations Environment ProgrammeWEHAB Water, Energy, Health, Agriculture and Biodiversity – 5 key thematic areas raised by Kofi

Annan for WSSD 2002WSSD World Summit on Sustainable Development Johannesburg 2002.WWF World Wide Fund for Nature


Aagaard K, Hindar K, Pullin AS, James CH, Hammarstedt O,Balstad T, Hanssen O (2002). Phylogenetic relationships inbrown argus butterflies (Lepidoptera: Lycaenidae: Aricia)from north-western Europe. Biol. J. Linn. Soc.

Achard F, Eva HD, Stibig HJ, Mayaux P, Gallego J, RichardsT, Malingreau J-P (2002). Determination of deforestationrates of the world’s humid tropical forests. Science 297,999–1002

Attfield R (1981). The good of trees. Journal of ValueInquiry 15, 35–54

Ayensu E, Claasen DV, Collins M, Dearing A, Fresco L,Gadgil M, Gitay H, Glaser G, Juma C, Krebs J, Lenton R,Lubchenco J, McNeely JA, Mooney HA, Pinstrup-Andersen P, Ramos M, Raven P, Reid WV, Samper C,Sarukhan J, Schei P, Tundisi JG, Watson RT, Xu GH, ZakriAH (2000). Biosphere management: Some tools of thetrade – Response. Science 287, 234–235

Baird Callicott J (1995). Animal liberation: a triangularaffair. Environmental Ethics 2, 311–338

Balmford A, Bruner A, Cooper P, Costanza R, Farber S,Green RE, Jenkins M, Jefferiss P, Jessamy V, Madden J,Munro K, Myers N, Naeem S, Paavola J, Rayment M,Rosendo S, Roughgarden J, Trumper K, Turner RK (2002).Economic reasons for conserving wild nature. Science297 (5583), 950–953

Balmford A, Green R, Jenkins M (2003). Measuring thechanging state of nature. Trends in Ecology and Evolution18 (7), 326–330

Baum JK, Myers RA, Kehler DG, Worm B, Harley SJ,Doherty PA (2003). Collapse and conservation of sharkpopulations in the Northwest Atlantic. Science 299(5605), 389–92

Biological Flora of the British Isles. In The Journal ofEcology British Ecological Society/Blackwell Sciencewww.britishecologicalsociety.org/articles/publications/journals/ecology/

BirdLife (2000). Threatened Birds of the World. BirdLifeInternational and Lynx Editions: Cambridge, UK andBarcelona, Spain

Boulinier T, Nichols JD, Sauer JR, Hines JE, Pollock KH (1998).Estimating species richness: the importance of heterogeneityin species detectability. Ecology 79, 1018–1028

Brako L & Zarucchi JL (eds.) (1993). Catalogue of theflowering plants and gymnosperms of Peru. Monogr.Syst. Bot. Missouri Bot. Gard. 45, 1–1286

Bramwell D (2002). How many plant species are there?Plant Talk 28, 32–33

Brooks TM, Pimm SL, Oyugi JO (1999). Time lag betweendeforestation and bird extinction in tropical forestfragments. Comserv. Biol. 13, 1140–1150

Buckland ST, Anderson DR, Burnham KP, Laake JL (1993).Distance sampling – Estimating abundance of BiologicalPopulations. Chapman & Hall: London

Burgess ND, Fjeldså J, Botterweg R (1998). Faunalimportance of the Eastern Arc Mountains of Kenya andTanzania. Journal of East African Natural History Soc. 87,37–58

Burgess ND, Doggart N, Lovett JC (2002). The Ulugurumountains of eastern Tanzania: the effect of forest loss onbiodiversity. Oryx 36, 140–152

Butynski TM (2001). Africa’s Great Apes. In Great Apesand Humans: The Ethics of Coexistence (eds. Beck B,Stoinski T, Hutchins M, Maple TL, Norton BG, Rowan A,Stevens EF, Arluke A). Simithonian Institution Press:Washington, DC

Casey J & Myers RA (1998). Near extinction of a large,widely distributed fish. Science 281, 690–692

CBD (2001). Global Biodiversity Outlook. Secretariat ofthe Convention on Biological Diversity / UNEP: Montreal,Canada

Christensen V & Pauly D (1998). Changes in models ofaquatic ecosystems approaching carrying capacity. Ecol.Applic. Suppl. S8, S104–S109.

CITES (1973). Convention on International Trade inEndangered Species of Wild Fauna and Flora. WashingtonDC

Costanza R, D’Arge R, DeGroot R, Farber S, Grasso M,Hannon B, Limburg K, Naeem S, O’Neill RV, Paruelo J,Raskin RG, Sutton P, van den Belt M (1997). The value ofthe world’s ecosystem services and natural capital. Nature387, 253–260

Daily GC (ed.) (1997). Nature’s Services: SocietalDependence on Natural Ecosystems. Island Press:Washington, DC

De La Vega-Salazar MY, Avila-Luna EG, Macías-Garcia C(a in press). Threatened fishes of the world: Zoogoneticustequila (Webb and Miller 1998) (Goodeidae).(Environmental Biology of Fishes).

Annex F References and relevant websites


De la Vega-Salazar MY, Avila-Luna EG Macías-Garcia C (bin press) Ecological evaluation of local extinction: the caseof two genera of endemic Mexican fish, Zoogoneticusand Skiffia (Biodiversity and Conservation)

DEFRA (Department for Environment Food and RuralAffairs) (2002). Wild Bird Index. In Achieving a betterquality of life. Review of progress towards sustainabledevelopment. Government Annual Report 2002. DEFRA:UK

DOE (Department of Environment) (1994). Biodiversity:The UK Action Plan (HMSO): London, UK

Dulvy NK, Metcalfe JD, Glanville J, Pausson MK, ReynoldsJD (2000). Fishery stability, local extinctions, and shifts incommunity structure in skates. Conservation Biology 14,283–293

Dulvy NK, Sadovy Y, Reynolds JD (2003). Extinctionvulnerability in marine populations. Fish and Fisheries 4,25–64

Endler JA (1995). Multiple-trait coevolution andenvironmental gradients in guppies. Trends Ecol. Evol. 10,22–29

Erhlich PR (2002). Human natures, nature conservationand environmental ethics. BioScience 52, 31–43

EC (European Commission) (1998). Communication fromthe commission to the Council and to the Parliament on aEuropean Community Biodiversity Strategy COM (98) 42:Brussels, Belgium

EC (European Commission) (2001). Sixth EnvironmentAction Programme COM(2001)31 Environment2010:Our Future – Our Choice: Brussels

EEC (1979). European Birds Directive. Council Directive79/409/EEC of 2 April 1979 on the conservation of wildbirds: Brussels

EEC (1992). European Habitats Directive. CouncilDirective 92/43/EEC of 21 May 1992 on the conservationof Natural Habitats and of Wild fauna and Flora: Brussels

FAO (2001). Global Forest Resources Assessment 2000:Main Report. FAO: Rome

Ferguson ME, Ford-Lloyd BV, Robertson LD, Maxted N,Newbury HJ (1998a). Mapping the geographicaldistribution of genetic variation in the genus Lens for theenhanced conservation of plant genetic diversity.Molecular Ecology 7, 1743–1755

Ferguson ME, Newbury HJ, Maxted N, Ford-Lloyd BV,Robertson LD (1998b). Population genetic structure of

Lens taxa as revealed by isozyme and RAPD analysis.Genetic Resources and Crop Evolution 45, 549–559

Ferguson ME, Robertson LD, Ford-Lloyd BV, Newbury HJ,Maxted N (1998c). Contrasting genetic variation amongstlentil landraces from different geographical origins.Euphytica 102, 265–273

Ferguson ME, Maxted N, van Slageren M, Robertson LD,(2000). A re-assessment of the taxonomy of Lens Mill.(Leguminosae, Papilionoideae, Vicieae). Botanical Journalof the Linnean Society 133, 41–59

Floyd R, Abebe E, Papert A, Blaxter M (2002). Molecularbarcodes for soil nematode identification. Mol. Ecol. 11,839–850.

Freiwald A, Wilson JB, Henrich R (1999). GroundingPleistocene icebergs shape recent deep-water coral reefs.Sediment. Geol. 125, 1–8

Funk VA, Zermoglio MF, Nasir N (1999). Testing the use ofspecimen collection data and GIS in biodiversityexploration and conservation decision making in Guyana.Biodiversity and Conservation 8, 727–751

Gaston KJ & May RM (1992). Taxonomy of taxonomists.Nature 356, 281–282

Godfray HCJ (2002). How might more systematics befunded? Antenna 26, 11–17

Govaerts R (2001). How many species of seed plants arethere? Taxon 50, 1085–1090

Gregory RD, Wilkinson NI, Noble DG, Robinson JA, BrownAF, Hughes J, Proctor DA, Gibbons DW, Galbraith CA(2002). The population status of birds in the UnitedKingdom, Channel Islands and the Isle of Man: an analysisof conversation concern 2002–2007. British Birds 95,410–450

Hammond PM (1995). The current magnitude ofbiodiversity. In Global Biodiversity Assessment (ed.Heywood VH) 113–128. Cambridge University Press:Cambridge

Haskins CP, Haskins EF, McLaughlin JJA, Hewitt RE (1961).Polymorphism and population structure in Lebistesreticulatus, an ecological study. In Vertebrate Speciation(ed. Blair WF), 320–395: University of Texas Press: Austin

Hawkes JG, Maxted N, Ford-Lloyd BV (2000). The ex situconservation of plant genetic resources. Kluwer:Dordrecht.

Houde AE (1997). Sex, color and mate choice in guppies.Princeton University Press: Princeton


Houlahan JE, Findlay CS, Schmidt BR, Meyer AH, KuzminSL (2000). Quantitative evidence for global amphibianpopulation declines. Nature 404, 752–755

House of Lords, Select Committee on Science andTechnology (1992). Systematic biology research, 1stReport. HL Paper 22 – I: London, UK

House of Lords, Select Committee on Science andTechnology (2002). What on Earth? The threat to thescience underpinning conservation, 3rd Report. HL Paper118 (i): London, UK

IUCN (1980). World Conservation Strategy: LivingResource Conservation for Sustainable Development.International Union for Conservation of Nature andNatural Resources: Gland, Switzerland

IUCN (1997). IUCN Red List of threatened plants.International Union for Conservation of Nature andNatural Resources: Gland, Switzerland

IUCN (2002). 2002 IUCN Red List of threatened species.(See web link below)

Jackson JBC (1997). Reefs since Columbus. Coral ReefsS16, S23–S32

Jackson JBC, Kirby MX, Berger WH, Bjorndal KA, BotsfordLW, Bourque BJ, Bradbury R, Cooke R, Estes JA, HughesTP, Kidwell S, Lange CB, Lenihan HS, Pandolfi JM,Peterson CH, Steneck RS, Tegner MJ, Warner RR (2001).Historical overfishing and the collapse of coastalecosystems. Science 293, 629–638

Jenkins DG, Grissom S, Miller K (2003). Consequences ofprairie wetland drainage for crustacean biodiversity andmetapopulations. Conservation Biology 17, 158–167

Jørgensen PM & León-Yánez S (eds.) (1999). Catalogue ofthe vascular plants of Ecuador. Missouri BotanicalGarden: St. Louis.

Karanth KU (1999). Counting tigers with confidence. InRiding the tiger: tiger conservation in human-dominated landscape (eds. Seidensticker J, Christie S,Jackson P), 350–353. Cambridge University Press:Cambridge

Karanth KU & Nichols JD (2002). Monitoring tigers andtheir prey: a manual for researchers, managers andconservationists in tropical Asia. Centre for WildlifeStudies: Bangalore, India

Karanth KU Nichols JD, Seidensticker J, Dinerstein E,Smith JLD, McDougal C, Johnsingh AJT, Chudawat RS(2003). Science deficiency in conservation practice: themonitoring of tiger populations in India. AnimalConservation 6 (2).

Loh J (ed.) (2002). Living Planet Report. World Wide Fundfor Nature (WWF): Gland, Switzerland

Lovett J (1998). Importance of the Eastern Arc mountainsfor vascular plants. Journal of East African Natural HistorySoc. 87, 59–74

Magurran AE (1998). Population differentiation withoutspeciation. Phil. Trans. R. Soc. Lond. B. 353, 275–286

Master L, Stein A, Kutner L, Hammerson G (2000).Vanishing Assets: Conservation status of US species. InPrecious Heritage: The status of biodiversity in the UnitedStates (eds. Stein B, Kutner L, Adams J) 93–98. OxfordUniversity Press: New York

Matthews E (2001). Understanding the FRA 2000. ForestBriefing No 1. World Resources Institute: Washington DC

Maxted N, Ford-Lloyd BV & Hawkes J G (eds.) (1997).Plant genetic conservation: The in situ approach.Chapman and Hall: Reino Unido

May RM, Lawton JH, Stork NE (1995). Assessingextinction rates. In Extinction Rates (eds. Lawton, JH&May, RM), 1–24: Oxford, UK

May RM (1999). The dimensions of life on earth. In Natureand Human Society (ed. Raven P). National Academy ofSciences Press: Washington DC

Moyle PB & Leidy RA (1992). Loss of biodiversity in aquaticecosystems: evidence from fish faunas. In ConservationBiology (eds. Fiefler PL & Jain SK), 127–169, Chapman &Hall: London

Myers N & Knoll AH (2001). The biotic crisis and the futureof evolution. Proc Natl Acad Sci. 98 (10), 5389–92

NRC (National Research Council) (1980). Researchpriorities in tropical biology. National Academy ofSciences: Washington, DC

NEFSC (2000) 30th Northeast Regional Stock AssessmentWorkshop (30th SAW): Stock Assessment ReviewCommittee (SARC) consensus summary of assessments.Northeast Fisheries Science Centre: Woods Hole

Nelson BW, Ferreira CAC, da Silva MF & Kawasaki ML(1990). Endemism centres, refugia and botanicalcollection density in Brazilian Amazonia. Nature 345,714–716

Nelson JS (1994). Fishes of the world. Wiley-Interscience:New York

Newmark WD (2002). Conserving biodiversity in EastAfrican forests. A study of the Eastern Arc mountains.Springer: Berlin.


O’Neill J (2001). Sustainability: Ethics, politics and theenvironment. In Environmental Ethics and Philosophy(eds. O’Neill J, Turner R, Bateman I). Edward Elgar,Cheltenham.

Odum EP (1969). The strategy of ecosystemdevelopment. Science 164, 262–270

Packer MJ & Macdonald DW (2002). Business andbiodiversity: existing threats – emerging opportunities. InBusiness & Biodiversity: The Handbook for CorporateAction; Earthwatch Europe, IUCN and WBCSD.

Paterson GLJ (1993). Patterns of polychaete assemblagestructure from bathymetric transects in the RockallTrough, NE Atlantic Ocean. PhD disseration, University ofWales.

Pauly D, Christensen V, Dalsgaard J, Froese R, Torres F(1998). Fishing down marine food webs. Science 279,860–863

Pimm SL & Askins RA (1995). Forest losses predict birdextinctions in eastern North America. Proc Natl Acad Sci.92 (20), 9343–7

Pimm SL (1998). The forest fragment classic. Nature 393,23–24

Pinnegar JK, Jennings S, O’Brien CM, Polunin NVC(2002). Long-term changes in the trophic level of theCeltic Sea fish community and fish market pricedistribution. Journal of Applied Ecology 39, 377–390

Pitman NCA & Jørgensen PM (2002). Estimating the sizeof the world’s threatened flora. Science 298, 989

Pitman NCA, Jørgensen PM, Williams RSR, León-YánezS, Valencia R (2002). Extinction-rate estimates for amodern Neotropical flora. Conservation Biology 16,1427–1431

Prance GT, Beentje H, Dransfield J, Johns R (2000). Thetropical flora remains undercollected. Annals of theMissouri Botanical Garden 87, 67–71

Preston C D, Pearman D A & Dines T D (2002). New Atlas ofthe British and Irish Flora. Oxford University Press: Oxford

Ramsar (1971) Convention on Wetlands of InternationalImportance especially as Waterfowl Habitat. Ramsar,Iran

Raven PH (2002). Science, sustainability, and the humanprospect. Science 297, 954–958

Reznick DN, Bryga H, Endler JA (1990). Experimentallyinduced life-history evolution in a natural population.Nature 346, 357–359

Royal Society (2002). Response from the Royal Society tothe House of Lords Inquiry into Systematics andBiodiversity. Document 33/02

Scotland RW & Wortley AH (2003). How many species ofseed plants are there? Taxon 52, 101–104

Seidensticker J, Christie S, Jackson P (1999). Riding thetiger: tiger conservation in human-dominatedlandscapes: Cambridge University Press: Cambridge

Solow AR, Mound LA, Gaston KJ (1995). Estimating therate of synonymy. Syst. Biol. 44, 93–96

Spalding MD, Ravilious C, Green EP (2001). World Atlas ofCoral Reefs. Prepared at the World ConservationMonitoring Centre. University of California Press:Berkeley

Stokstad E (2001). UN report suggests slowed forestlosses. Science 291, 2294

Taylor PW (1981). The ethics of respect for nature.Environmental Ethics 3, 197–218

Thapar V (1999). The tragedy of the Indian tiger: startingfrom scratch. In Riding the tiger: tiger conservation inhuman-dominated landscape (eds. Seidensticker J,Christie S, Jackson P) 297–306. Cambridge UniversityPress: Cambridge

Thomas JA, Lewington R (1991). The Butterflies of Britainand Ireland. Dorling Kindersley: London

Turner RK, Lorenzoni I, Beaumont N, Bateman IJ,Langford IH, McDonald AL (1998). Coastal managementfor sustainable development: Analyzing environmentaland socio-economic changes on the UK coast. TheGeographical Journal 164, 269–281

Turner RK, Paavola J, Cooper P, Farber S, Jessamy V,Georgiou S (2002). Valuing nature: lessons learned andfuture research directions. Centre for Social andEconomics Research on the Global Environment (CSERGE)Working Paper EDM 02–05: University of East Anglia,Norwich

UK Biodiversity Group (2001). Millennium biodiversityreport. Sustaining the variety of life: 5 years of the UKBiodiversity Action Plan. DETR (Department ofEnvironment, Transport and the Regions): UK

United Nations (1972). Report of the United NationsConference on the Human Environment, Stockholm,5–16 June 1972. United Nations: New York

United Nations (1982). World Charter for Nature. UNGeneral Assembly 37th Session (UN/GA/RES/37/ 7): NewYork


United Nations (2002). World Summit on SustainableDevelopment: Johannesburg Plan of Implementation.United Nations: New York

Valencia R, Pitman N, León-Yanéz S & Jorgensen PM (eds.)(2000). Libro rojo de las plantas endémicas del Ecuador.Pontificia Universidad Católica del Ecuador: Quito

Valiela IO, Bowen JL, York JK (2001). Mangrove forests:one of the world’s most threatened major tropicalenvironments. BioScience 51, 807–815

Williams PH, Gaston KJ & Humphries CJ (1998). Mappingbiodiversity value world-wide: combining higher-taxonrichness from different groups. Proceedings of the RoyalSociety, Biological Sciences 264, 141–148

Wilson EO (1999). The Diversity of Life. Norton: New York

Wilson RJ, Ellis S, Baker JS, Lineham ME, Whitehead R,Thomas CD (2002). Large scale patterns of distributionand persistence at the range margins of a butterfly.Ecology 83, 3357–3368

Winfield IJ, Adams CE, Fletcher JM (1996). Recentintroductions of the ruffe (Gymnocephalus cernuus) tothree United Kingdom lakes containing Coregonusspecies. Annales Zoologici Fennici 33, 459–466

WCED, World Commission on Environment andDevelopment (1987) Our Common Future (theBrundtland Report). Oxford University Press: Oxford, UK

Wynne IR, Wilson RJ, Burke AS, Simpson F, Mallet J, PullinAS, Thomas CD (submitted). The Effect ofMetapopulation Processes on the Spatial Scale ofAdaptation Across an Environmental Gradient. Amer.Nat. Also available at:http://abacus.gene.ucl.ac.uk/jim/pap/wynne03.pdf

Zuloaga FO & Morrone O (eds.) (1999). Catálogo de lasplantas vasculares de la República Argentina.Monographs in Systematic Botany from the MissouriBotanical Garden 74: St. Louis

Zuloaga FO, Nicora E, Rúgolo de Agrasar ZE, Morrone O,Ciadella AM (1994). Catálogo de la familia Poaceae en laRepública Argentina. Monographs in Systematic Botanyfrom the Missouri Botanical Garden 47: St. Louis

Relevant Websites

2002 IUCN Red List of Threatened Specieswww.redlist.org

2002 IUCN Red List Update for Birdswww.birdlife.net/action/science/species/globally_tbu/gtbu_redlist.html

All Specieswww.all-species.org

(CITES) Convention on International Trade on EndangeredSpecies of Wild Fauna and Flora www.cites.org/

(CBD) Convention on Biological Diversitywww.biodiv.org

(CODATA) – Committee on Data for Science andTechnology – International Council for Science (ICSU)www.codata.org

Global Taxonomy Initiativewww.biodiv.org/programmes/cross-cutting/taxonomy/default.asp

International Plant Names Indexwww.ipni.org

Millennium Ecosystem Assessmentwww.millenniumassessment.org

Missouri Botanical Gardenswww.mobot.org/

Species 2000www.sp2000.org/

UK Biodiversity Websitewww.ukbap.org.uk

United Nations World Summit on SustainableDevelopment: Plan of Implementationwww.un.org/esa/sustdev/documents/WSSD_POI_PD/English/POIToc.htm

Keeping science open: the effects of intellectualproperty policy on the conduct of science (38 pages,April 2003, 02/03)

Comments on the Government’s response to theHouse of Lords report on systematic biology andbiodiversity (‘What on Earth?’) (2 pages, April 2003,06/03)

Human reproductive cloning: A statement by theRoyal Society (2 pages, January 2003, 1/03)

Economic instruments for the reduction of carbondioxide emissions (44 pages, November 2002, 26/02)

Scientist support for biological weapons control(November 2002, 1 page)

Submission to FCO green paper on strengtheningthe biological and toxin weapons convention (6pages, November 2002, 25/02)

Climate change: what we know and what we needto know(19 pages, 22/02, August 2002, ISBN 0 85403 581 8)

Infectious diseases in livestock (8 page summary,19/02, July 2002, ISBN 0 85403 580 X and 160 pagedocument, 15/02, July 2002, £25 ISBN 0 85403 579 6)

Response to the DEFRA consultation documentManaging Radioactive Waste Safely (8 pagestatement, 12/02, May 2002)

The health effects of depleted uranium munitionsSummary (8 page summary of Parts I and II 6/02, March2002 ISBN 085403 5753)

The health hazards of depleted uranium munitionsPart II (150 page document 5/02, March 2002 £17.50ISBN 0 85403 574 5)

Genetically modified plants for food use and humanhealth – an update (20 page document, 4/02, February2002, ISBN 0 85403 576 1)

Statement of the Royal Society’s position on the useof animals in research (2 page statement, 3/02, January2002, ISBN 0 85403 5737)

Response from the Royal Society to the House ofLords Inquiry into Systematics and Biodiversity (5pages, January 2002, 33/02)

Response to the Policy Commission on the future offarming and food (4 page response to the PolicyCommission, 23/01, October 2001)

The role of land carbon sinks in mitigating globalclimate change (2 page summary, 11/01, July 2001,ISBN 0 85403 561 3 and 32 page document, 10/01, July01, ISBN 0 85403 562 1)

Transmissible spongiform encephlopathies (11 pagestatement 8/01, June 2001, ISBN 0 85403 559 1)

The health hazards of depleted uranium munitions,Part I (2 page summary 07/01, May 2001, ISBN 0 854035575 and 88 page document 06/01, £17.50 ISBN 085403 3540)

The use of genetically modified animals (46 pagedocument 5/01, May 2001, ISBN 0 85403 556 7)

The Science of Climate Change (2 page joint statementfrom 16 scientific academies, May 2001)

The future of Sites of Special Scientific Interest(SSSIs) (21 page document 1/01, February 2001)

The role of the Renewables Directive in meetingKyoto targets (12 page document 11/00, October 2000ISBN 00 85403 5486)

Developing a National strategy for science (8 pagedocument 10/00, July 2000 ISBN 0 85403 5451)

Transgenic plants in world agriculture (2 pagesummary 09/00, July 2000 and 19 page full report 08/00,July 2000, ISBN 0 85403 5443)

Further copies of these reports can be obtained from: Science Advice Section The Royal Society6–9 Carlton House Terrace London SW1Y 5AG

The full text, or summary, of these reports can be found on the Royal Society’s web site (www.royalsoc.ac.uk)

Other recent Royal Society reports

measuring biodiversity for conservation/media/royal_society... · measuring biodiversity for...

Documents