Business & Finance

Use and misuse of the IUCN Red List Criteria in projecting climate change impacts on biodiversity

Description
Recent attempts at projecting climate change impacts on biodiversity have used the IUCN Red List Criteria to obtain estimates of extinction rates based on projected range shifts. In these studies, the Criteria are often misapplied, potentially
Published
of 10
9
Published
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Transcript
    1 Use and Misuse of the IUCN Red List Criteria inProjecting Climate Change Impacts on Biodiversity H. Resit AKÇAKAYA, 1 Stuart H. M. BUTCHART, 2 Georgina M. MACE, 3 Simon N.STUART, 4 and Craig HILTON-TAYLOR  5   1 Applied Biomathematics, 100 North Country Road, Setauket, NY 11733, USA 2 BirdLife International, Wellbrook Court, Girton Road, Cambridge, CB3 0NA, UK  3 Institute of Zoology, Zoological Society of London, Regents Park, London, NW1 4RY, UK  4 IUCN/SSC Biodiversity Assessment Unit, Center for Applied Biodiversity Science,Conservation International, 1919 M Street, NW Suite 600, Washington, DC 20036, USA 5 IUCN Red List Programme, IUCN/SSC UK Office, 219c Huntingdon Road, Cambridge,CB3 0DL, UK  Author Posting. © Blackwell Publishing Ltd 2006. This is the author's version of the work. It is postedhere by permission of Blackwell Publishing Ltd for personal use, not for redistribution. The definitiveversion was published in Global Change Biology  , 12, 2037–2043 , doi:http://dx.doi.org/10.1111/j.1365-2486.2006.01253.x  Abstract  Recent attempts at projecting climate change impacts on biodiversity have used the IUCNRed List Criteria to obtain estimates of extinction rates based on projected range shifts. Inthese studies, the Criteria are often misapplied, potentially introducing substantial bias anduncertainty. These misapplications include arbitrary changes to temporal and spatial scales;confusion of the spatial variables; and assume a linear relationship between abundance andrange area. Using the IUCN Red List Criteria to identify which species are threatened byclimate change presents special problems and uncertainties, especially for shorter-livedspecies. Responses of most species to future climate change are not understood well enoughto estimate extinction risks based solely on climate change scenarios and projections of shiftsand/or reductions in range areas. One way to further such understanding would be to analyzethe interactions among habitat shifts, landscape structure and demography for a number of species, using a combination of models. Evaluating the patterns in the results might allow thedevelopment of guidelines for assigning species to threat categories, based on a combinationof life history parameters, characteristics of the landscapes in which they live,   and projectedrange changes. Introduction Recently observed responses by many species to climate change (Hughes 2000, Parmesan andYohe 2003, Root et al. 2003) have led to efforts to quantify the impact of future climatechange on biodiversity. One measure of the impact is a predicted increase in species’extinction rates. A common approach to investigating this involves bioclimatic modeling,which projects future distributions of species under the assumption that the current climaticconstraints that define a species’ distribution reflect its environmental preferences, and will,therefore, be retained under climate change (thus, often resulting in shifts in species' ranges    2towards poles or higher elevations). There are several problems and uncertainties with thisapproach (Pearson & Dawson 2003, Buckley & Roughgarden 2004, Hampe 2004, Thuiller  et al. 2004, Araujo et al. 2005), which are widely acknowledged. In this paper, we focus on adifferent problematic aspect of this approach: that of linking the results of bioclimatic models(shifts and reductions in species' ranges) to extinction rates, using an approach loosely basedon the IUCN Red List Criteria (IUCN 2001), as has been carried out in a number of recent publications (Thomas et al. 2004, Bomhard et al. 2005, Shoo et al. 2005a, Thuiller  et al.  2005).The  IUCN Red List of Threatened Species (IUCN 2004)   is widely regarded as the mostauthoritative list of globally threatened species (Lamoreux et al. 2003, Rodrigues et al. 2006).The IUCN Red List Criteria allocate species to categories of extinction risk using simplequantitative rules based on population sizes and population decline rates, and range areas andrange declines (Table 1). The Criteria recognize that there are major differences betweenspecies, and the circumstances leading to their extinction risk, and therefore, specify closelyhow the data from species should be used in order to compare each case with the specifiedquantitative thresholds (IUCN 2001, S&PS 2005).All recent studies we have reviewed that use the IUCN Red List Criteria to attempt toquantify likely extinctions from climate change have misapplied the Criteria, even if somehave explicitly acknowledged that this is the case. Although we do not doubt the centralconclusions of these studies that climate change will result in significantly increasedextinction rates, these misapplications could nevertheless introduce substantial bias anduncertainty to projections of climate change impacts on biodiversity. We first discuss themisapplication of the Criteria and the biases this introduces. Secondly, we discuss the particular problems and uncertainties associated with applying the IUCN Red List Criteria tospecies threatened by climate change, and propose some broad approaches for how these can be overcome. Common Mistakes in Using the Red List Criteria The main misuses of the criteria involve quantitative estimates of extinction risk, temporaland spatial scales, spatial resolution, and assumptions about species-area relationships. Quantitative estimates of extinction risk  Four of the IUCN Red List Criteria (A-D) are based on the size and rate of decline of the population and/or geographical range, while the fifth (Criterion E) is based on quantitativemodels of extinction risk such as Population Viability Analyses (Table 1). Thus, onlyCriterion E includes quantitative thresholds for the risk of extinction. Because other criteriado not include such thresholds, the risk-based thresholds of Criterion E should not be used toinfer an extinction risk for species assessed as threatened under any of the criteria A to D(S&PS 2005). The reasons for this are that, given the variation among species, it is not possible to validate the equivalence of the thresholds in different criteria, and the factors builtinto an evaluation under E, or under A-D may not be incorporated in the alternative criterion. Temporal scale  The most common mistake involves the time periods over which decline rates and extinctionrisks are to be calculated. The IUCN Red List Criteria assess population declines over a period of 10 years or 3 generations (whichever is longer) up to a maximum of 100 years into    3the future (IUCN 2001). Several studies arbitrarily change this time scale. For example,Thomas et al . (2004) use time scales of 50 years (for CR and EN) and 100 years (for VU) toassess declines in future ranges of species, stating that the srcinal timescales "are not suitedto evaluate the consequences of slow-acting but persistent threats". Similarly, Shoo et al .(2005a) use a time scale of 100 years to assign species to extinction risk categories based on projected declines in population size. Thuiller  et al . (2005) use an arbitrary time scale of 80years to assess declines in future ranges of species (they also incorrectly state that the IUCNRed List Criteria use a 20-year time scale). Bomhard et al . (2005) do take generation timeinto account, but they use three different time scales for three subcriteria (60 years for A2 20years for A3, and 80 years for A4), whereas all subcriteria require the same time scale (if notmore than 100 years).It is misleading and incorrect to use these arbitrary time frames, especially when thegeneration times of the species being assessed vary. For example, Thuiller  et al . (2005) study1,350 European plants, and assume that they all have the same generation time. Generationtimes for such a large collection of species would vary considerably; the time periods for estimating declines under the IUCN criteria would range from 10 years for very short-lived plants to 100 years for long-lived trees (or even longer under Criteria A2 and A4). The biasintroduced by these arbitrary time frames depends on the generation time of the speciesinvolved, and its size and direction cannot be estimated without this information. In theThomas et al study, the animal species considered have relatively short generation times onaverage, so the overall effect of increasing the timescale is likely to have exaggeratedestimates of extinction risk.Although the issue of "slow-acting but persistent threats" is an important one (discussed below), the ad-hoc decision to arbitrarily change time scales while keeping the same declinethresholds as in the IUCN Criteria cannot be justified, even if they are changed in a way to produce conservative results. The time scales and decline thresholds in the IUCN Criteriawere set against a common standard to provide broad consistency between Criteria and toallow comparisons across taxonomic groups (IUCN 2001). When time scales are changedand the thresholds are kept the same, the resulting set of rules loses this consistency, andcannot then be referred to as the IUCN Red List Criteria. Spatial scale  The IUCN Red List is explicitly a global assessment of projected extinction risk for species.Applying the IUCN Red List Criteria at sub-global scales requires special considerations(IUCN 2003). When climate change impacts on species are assessed at a continental or smaller spatial scale, the projections for many species often exclude part of the species' range.Models based only on part of a species’ range cannot be used to assess global risk for thatspecies, as they do not take into account the dynamics across the entirety of a species' range.For example, the analysis by Thuiller  et al . (2005) is restricted to data from western Europe.For some species, this may include all or most of the range, but most species are likely tohave ranges that extend into eastern Europe, western Asia, and northern Africa. Ignoring thisfactor is likely to increase estimates of extinction risk. Spatial measures and resolution  IUCN Red List Criterion B refers to two spatial measures related to the distribution of thespecies: Extent of Occurrence (EOO) and Area of Occupancy (AOO). These measures aredefined in a specific way (and at a specific resolution for AOO) in order to provide a    4consistent standard, based on the thresholds set for the threat categories. Broadly, EOO is therange (e.g. a minimum convex polygon encompassing all sites of present occurrence) andAOO is the area within EOO that is occupied by the species (IUCN 2001).Thomas et al . (2004) compared the size of the projected future distribution area against theIUCN’s thresholds for AOO (see Table 1). However, the area that they model is not AOO asdefined in the IUCN Criteria. It may actually be closer to EOO, for which the thresholds aremuch higher (Table 1). In general, the modeled distributions must be processed to estimateEOO and/or AOO, as defined by the IUCN Red List Criteria and related guidelines, before thethresholds specified under Criterion B can be applied to the sizes of these predicted areas.Otherwise, as in the case of arbitrary changes to temporal scales, it results in inconsistencies.In the case of Thomas et al . (2004), the results are likely to have been conservative(underestimating potentially threatened species, because the thresholds for EOO are higher than those for AOO), which the authors acknowledge.Spatial resolution is an important issue even when the projected areas are not compared to theCriterion B thresholds. For example, the spatial resolution used by Thuiller  et al . (2005) (50x 50 km, projected to 10’ grid) may be adequate for making broad projections of biodiversityimpacts, but is too coarse for making predictions about specific species, particularly inmountainous areas, because it cannot incorporate the large variation in species' distributionsand in microclimate brought about by high topographic relief. The use of a coarser resolutionmay underestimate threats for such species if smaller scale declines are not projected at thecoarser resolution. Relationship between distribution and population size  Biodiversity impacts simulated in the studies referred to above are often based on (or evenlimited to) changes in projected range of the species, ignoring other aspects of the species’ biology. For example, when Criterion A is used to predict species extinction rates based onchanges in range area (as was done by Bomhard et al . 2005 and Thuiller  et al . 2005), a linear relationship between abundance and range area is assumed. In practice, although there isusually a positive and sometimes linear relationship between abundance and area amongspecies, the relationship is rarely linear within individual species, and theory suggests it isunlikely that a species’ abundance will decline at the same rate as its range area (Lawton1993, Rodríguez 2002). However, the assumption is especially problematic in the context of climate change impacts. This is because several additional factors are likely to change (and inmany cases exacerbate) the effects of climate change on species viability beyond the effects predicted by range shifts or changes in available habitat area (thus, methods that do notincorporate these factors may underestimate extinction risks). These factors include non-uniform spatial distribution of a species throughout its range (Shoo et al. 2005b), interactionof range shifts with fragmentation and land-use (Benning et al. 2002), isolation (e.g. Honnay  et al. 2002), increased frequency of extreme weather events (McLaughlin et al. 2002),increased spatial correlation of local temperatures, with an associated increase in spatialcorrelation of local population dynamics (Post & Forchhammer 2004), diseases (Burrowes et al . 2004), and other biotic interactions (Arnott & Ruxton 2002, Frederiksen et al. 2004).Although these factors are often discussed in the literature (including by the authors of thestudies we have discussed), bioclimatic models rarely incorporate them. Because thesefactors involve demographic processes (dispersal, population fluctuations, and trends in vitalrates) and the spatial structure of the landscape, they are ignored by assessments that are    5 based solely on projected range area changes. These factors are incorporated into the IUCNRed List Criteria through the use of variables such as “severely fragmented”, “continuingdecline”, “population reduction”, “extreme fluctuations”, as well as by explicit calculation of extinction risk in Criterion E. These options are further discussed in the next section. Red-listing Species Threatened by Climate Change Although analyses of overall extinction rates resulting from climate change can provideimportant information about potential impacts on biodiversity, we believe that a more practical use of the IUCN Red List in this context is to identify those species that are or may become threatened by climate change. The IUCN Red List Criteria are used to categorizespecies according to their risk of extinction, based on information on population size, habitat,range, fragmentation, fluctuations, threats, and trends in population size, habitat and range. Inthis section, we discuss specific criteria that can be used to red-list species threatened byclimate change, and special problems and uncertainties that are associated with the use of these criteria. The following are only a summary of preliminary guidelines on validapproaches; further details are currently being developed.   Population reduction  Criterion A3 is used to list species when there is a population reduction projected or suspected to occur in the future, based on an index of abundance appropriate to the taxon, or adecline in occupied habitat, range or habitat quality. For example, recently the IUCN SSCPolar Bear Specialist Group recommended listing this species as VU under Criterion A3 dueto global warming (Wiig 2005). Although this is the most straightforward way to red-listspecies threatened by climate change, there are two issues to consider. First, this criterionrequires a projected reduction in abundance . Although this population reduction may be based on a projected decline in occupied habitat, range, or habitat quality, any assumedrelationship between abundance and habitat/range must be justified, as discussed above.Second, for short-lived species, Criterion A has a limited time horizon (3 generations or 10years, whichever is longer). Because climate change can affect species in long time horizons(50+ years), three generations will often be too soon for the impacts of climate change to beapparent on these species, even if past greenhouse gas emissions have already determined(and have made inevitable) climate change effects in longer time horizons. In most cases,however, uncertainties and lack of knowledge of environmental trends and demographiccharacteristics of species make it very unreliable to make predictions of extinction risk over  periods longer than 3 generations. In cases where sufficient data exist, and assessors areconfident that longer-term predictions are justifiable, Criterion E can be used (see below). Inaddition, recent studies demonstrate that climate change can affect species viability muchfaster than implied by range shifts, when it interacts with other factors (Pounds et al . 2006).When such factors are incorporated into the assessment, a 3-generation time scale may besufficient to project impacts on many short-lived species.   Restricted distribution  Criterion B is used to list species with restricted ranges or in restricted habitats, which arealso currently undergoing declines or extreme fluctuations, are severely fragmented, or existin few locations. Location is defined as a geographically or ecologically distinct area inwhich a single threatening event can rapidly affect all individuals of the taxon present. Thesize of the location depends on the area covered by the threatening event. Thus, when climate
Search
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x