Data uncertainty and the selectivity of extinction ... - Wiley Online Library

A Journal of Conservation Biogeography

Diversity and Distributions, (Diversity Distrib.) (2012) 18, 1211–1220

BIODIVERSITY RESEARCH

Data uncertainty and the selectivity of extinction risk in freshwater invertebrates Lucie M. Bland1,2*, Ben Collen1, C. David L. Orme2 and Jon Bielby1

1

Institute of Zoology, Zoological Society of London, Regent’s Park, London, NW1 4RY, UK, 2Division of Biology, Imperial College London, Silwood Park, Ascot, SL5 7PY, UK

ABSTRACT Aim To investigate the impact of different treatments of the IUCN Data Defi-

cient (DD) category on taxonomic and geographical patterns of extinction risk in crayfish, freshwater crabs and dragonflies. Location Global. Methods We used contingency tables to evaluate taxonomic and geographical

selectivity of data deficiency and extinction risk for three invertebrate taxonomic groups (crayfish, dragonflies and damselflies, and freshwater crabs) based on their IUCN Red List status. We investigated differences in patterns of data deficiency and extinction risk among taxonomic families, geographical realms and taxonomic families within geographical realms for each of the three groups. At each level, we evaluated the impact of uncertainty conferred by the conservation status of DD species on extinction risk patterns exhibited by that group. We evaluated three scenarios: excluding DD species, treating all DD species as non-threatened and treating all DD species as threatened.

Diversity and Distributions

Results At the global scale, DD species were taxonomically non-randomly dis-

tributed in freshwater crabs and dragonflies, and geographically non-randomly distributed in all three taxonomic groups. Although the presence of under- or over-threatened families and biogeographical realms was generally unchanging across scenarios, the strength of taxonomic and geographical selectivity of extinction risk varied. There was little consistent evidence for taxonomic selectivity of extinction risk at sub-global scales in freshwater crabs and dragonflies, either among biogeographical realms or among scenarios. Main conclusions Global patterns of taxonomic selectivity and geographical

*Correspondence: Lucie M. Bland, Institute of Zoology, Zoological Society of London, Regent’s Park, London, NW1 4RY, UK. E-mail: [email protected]

selectivity were generally consistent with one another and robust to different treatments of DD species. However, sub-global scale conservation prioritization from these types of data sets will require increased investment to make accurate decisions. Given the current levels of data uncertainty, the relative importance of biological characteristics and threatening processes in driving extinctions in freshwater invertebrates cannot be easily determined. We recommend that DD species should be given high research priority to determine their true status. Keywords Crayfish, data deficient, dragonflies, extinction risk, freshwater crabs, IUCN.

With current species extinction rates exceeding geological background rates by several orders of magnitude, it is now clear that we are facing an extinction crisis comparable to mass extinctions of the paleontological past (May et al.,

1995; Millenium Ecosystem Assessment, 2005; Butchart et al., 2010). However, species are not equally at risk of extinction (Purvis et al., 2000a). Extinction risk has been found to be non-randomly distributed among many groups, including families of birds (Bennett & Owens, 1997), amphibians (Stuart et al., 2004; Bielby et al., 2006) and mammals (Purvis

ª 2012 Blackwell Publishing Ltd

DOI: 10.1111/j.1472-4642.2012.00914.x http://wileyonlinelibrary.com/journal/ddi

INTRODUCTION

1211

L.M. Bland et al. et al., 2000b). This phenomenon, termed taxonomic selectivity of extinction risk, has not only been observed in extant taxa but also in historical and paleontological patterns of extinction. For example, over the last 75 million years, some echinoid genera have been more likely to go extinct than others (McKinney, 1997). The phylogenetically clumped nature of threat has severe implications for the loss of biodiversity and evolutionary history. Non-random extinction risk results in the more rapid loss of higher taxa and phylogenetic diversity than is predicted by random extinction (Nee & May, 1997; Purvis et al., 2000a). This is because the loss of all constituent species of clade and their evolutionary history is more likely under non-random extinction than under random extinction (Russell et al., 1998). Extinction risk is also known to be non-randomly distributed across geographical areas; prevalence of extinction risk is higher where threatening processes such as habitat degradation, overexploitation, invasive species and diseases are more intense (Kerr & Currie, 1995; McKinney, 1997). Furthermore, taxonomic and geographical selectivity are not independent, as evolutionary diversification within regions produces phylogenetic proximity that is often correlated with geographical proximity (Brooks & McLennan, 1993). Geographical scale also affects patterns of taxonomic selectivity as the concordance between extinction risk at a local and global level can be low (Purvis et al., 2005). Moreover, the persistence of taxonomic patterns of selectivity observed at the global scale at smaller spatial scales (where threatening processes are more homogeneous) would indicate the importance of species biology in determining susceptibility to extinction risk (Bielby et al., 2006). Studies of global extinction risk have primarily focused on birds, mammals and amphibians (Owens & Bennett, 2000; Purvis et al., 2000b; Cardillo et al., 2004, 2005, 2008; Davies et al., 2006; Cooper et al., 2008; Lee & Jetz, 2010), using the IUCN Red List categories as a proxy for extinction risk. However, despite their diversity and ecological importance, invertebrates remain under-studied, especially from a macroecological perspective (Diniz-Filho et al., 2010). Fortunately, comprehensive and sampled IUCN Red List assessments have now been made for a range of invertebrate groups (Baillie et al., 2008). The availability of new distribution maps for invertebrate taxa and data on extinction risk offers considerable scope for macro-ecological research, and the development of conservation strategies that integrate broad-scale data for these species. In particular, invertebrate taxa from freshwater habitats provide an opportunity to study patterns of threat in freshwater ecosystems, which are largely neglected in global conservation priority-setting, yet have been shown to be some of the most threatened ecosystems on Earth (Millenium Ecosystem Assessment, 2005; Revenga et al., 2005). The imperilment of freshwater systems also has direct links with human well-being and water security (Millenium Ecosystem Assessment, 2005; Vorosmarty et al., 2010), and a

1212

representative picture of freshwater species conservation status is necessary for successful integrated water management and climate change adaptation (Strayer & Dudgeon, 2010). Because global priorities for biodiversity conservation to date have been largely biased towards vertebrate species and terrestrial ecosystems, understanding the drivers of extinction risk in freshwater invertebrates will contribute to a more accurate picture of biodiversity as a whole. However, the high levels of data uncertainty in IUCN Red List assessments for freshwater invertebrates could bias the results of broad-scale studies based on these assessments. The Data Deficient (DD) category is assigned to a species ‘when there is inadequate information to make a direct, or indirect, assessment of its risk of extinction based on its distribution and/or population status’ (Butchart & Bird 2010; IUCN, 2001). To date, all invertebrate taxa with systematic risk assessments show high proportions of DD species: 35% of dragonflies and damselflies (Clausnitzer et al., 2009), 49% of freshwater crabs (Cumberlidge et al., 2009) and 21% of crayfish (Samways & Bo¨hm, 2010) are currently listed as DD. Vertebrate groups are typically better known, with only 1% of birds, but 15% of mammals and 19% of reptiles listed as DD (Collen et al., 2009; Hilton-Taylor et al., 2009). Data Deficient species are often simply excluded from calculations of the overall conservation status of a taxon. Within groups that have a high proportion of DD species, this approach disguises considerable uncertainty in the perception of the distribution of threat, both taxonomically and geographically. If the distribution of DD species is itself non-random among families and geographical regions, treating DD species as either all threatened or non-threatened could dramatically alter observed patterns of extinction risk. Given the limited resources available for conservation, it is important to use all available information to prioritize taxa and geographical regions effectively (Leader-Williams et al., 2010). Disentangling the effects of the distribution of DD species from the observed distribution of threat is therefore crucial to obtaining a more accurate picture of biodiversity. In this study, we aim to investigate the non-randomness of extinction risk in three taxa of freshwater invertebrates: crayfish, freshwater crabs and dragonflies, and focus on four questions: 1. Is there evidence for taxonomic and geographical selectivity of data deficiency in freshwater invertebrates? 2. Is there evidence for taxonomic and geographical selectivity of extinction risk in invertebrates under different treatments of DD species (DD species excluded, DD species considered non-threatened, and DD species considered threatened)? 3. What are the effects of geographical scale on the taxonomic selectivity of data deficiency and extinction risk? 4. Are there differences in the selectivity of data deficiency and extinction risk among invertebrate taxa, and among vertebrate taxa?

Diversity and Distributions, 18, 1211–1220, ª 2012 Blackwell Publishing Ltd

Extinction risk selectivity in freshwater invertebrates METHODS Data We gathered species data from three recent assessments of freshwater invertebrate taxa: crayfish (all 590 species, IUCN, 2010), freshwater crabs (all 1281 species, Cumberlidge et al. 2009) and dragonflies (a randomly selected sample of 1500 of 5680 species, Clausnitzer et al. 2009). We followed the taxonomy listed by IUCN to identify the number of threatened (VU, EN and CR categories), DD and non-threatened (LC and NT categories) species in each group (Table 1). From the published assessments, we recorded the taxonomic family and the biogeographical realm (Olson et al., 2001) of each species. We also used the justification for listing the species as DD to assign each species to one of seven categories (Table 2): taxonomic uncertainty, known only from the type series or locality, old records (prior to 1995), few records (< 5 records), unknown habitat or provenance, unknown threats and unknown population trends. Analyses We tested for non-randomness in the distribution of both data deficiency and extinction risk within each invertebrate taxon, using Fisher’s exact tests on the number of species in each category because of the low diversity of some families and realms. First, we tested for global taxonomic non-randomness in the prevalence of DD and threatened species among families. We did not conduct analyses at the genus level because of the small size of some of the genera in the taxa considered. Second, we tested for global geographical non-randomness in the prevalence of DD and threatened species among realms. Taxonomic and geographical selectivity are likely to be strongly non-independent because of the strong biogeographical structure of families across realms and differences in threat pressure between realms. Third, we therefore tested whether there is evidence for taxonomic selectivity of data deficiency and extinction risk within a realm, using data only from realms that contained more than 30 species (crayfish: Nearctic; freshwater crabs: Afrotropical, Indomalayan and Neotropical; dragonflies: Afrotropical, Australasian, Indomalayan, Neotropical and Palearctic). In each case, we tabulated the number of DD or threatened species against the number of non-DD and non-threatened species in each family or realm. We investigated the presence of non-random extinction risk under three scenarios representing uncertainty about the conservation status of DD species: DD species excluded, all DD species considered threatened and all DD species considered non-threatened. Despite testing distinct hypotheses, our analyses used different combinations and subsets of the same underlying tables. We therefore used Benjamini & Hochberg’s (1995) correction for multiple hypothesis tests across all the Fisher’s exact tests for each taxonomic group. All tests were conducted using R version 2.12.0 (R Development Core Team, 2010).

Table 1 Number of threatened species, DD (Data Deficient) species and non-threatened species in crayfish, freshwater crabs and dragonflies

Taxon

Number of threatened species

Number of DD species

Number of non-threatened species

Crayfish Freshwater crabs Dragonflies

146 202 135

125 632 527

461 649 973

Table 2 Number of species of crayfish, freshwater crabs and dragonflies assigned to each justification for the use of the DD (Data Deficient) category. Species listed under ‘old records’ have only been recorded prior to 1995; species listed under ‘few records’ are known from five records or less

Justification Type series or locality Few records Unknown habitat or provenance Old records Unknown population trends Unknown threats Taxonomic uncertainty

Freshwater crabs

Dragonflies

8 18 0

424 181 8

247 217 168

2 117

179 16

84 48

98 18

25 15

42 40

Crayfish

We interpreted significant associations in these tests by examining the magnitude and size of the difference between observed and expected species number. We display observed and expected numbers of DD and threatened species for the 29 dragonfly families in Appendix S1 in Supporting Information. RESULTS Taxonomic and geographical selectivity of data deficiency We found no evidence that the prevalence of DD species differed among families of crayfish (Table 3). However, the prevalence of DD species differed among families of dragonflies and freshwater crabs. Three families of freshwater crabs (Gecarcinucidae, Potamidae and Pseudothelphusidae) showed a higher proportion of DD species than other crab families (Table 3). All three groups exhibited geographical structure in the prevalence of DD species, with different realms showing higher numbers than expected in each group (Table 4). The reasons given for the assignment of a species to the DD category (Table 2) revealed some major differences across taxonomic groups, which may explain their differences in data deficiency selectivity.


1213

L.M. Bland et al. Taxonomic and geographical selectivity of extinction risk Threat was unevenly distributed across families of crayfish (Table 3) and biogeographical realms (Table 4). Overall, global patterns of taxonomic and geographical selectivity in crayfish were robust to the different treatments of DD species and consistent with each other, which were expected given the clumped geographical distribution of crayfish families. In freshwater crabs and dragonflies, the strength of the association between threat status and families (Table 3) or biogeographical realms (Table 4) varied with the treatment of DD species: the association was weaker when DD species were considered non-threatened. In both taxa, the distribution of threatened species across realms was identical when DD species were either excluded or considered non-threatened. The distribution varied in certain realms when DD species were considered threatened, although differences between the expected and observed levels of threat were small. The geographical distribution of threat across all scenarios was strikingly similar between freshwater crabs and dragonflies, except in the Australasian realm, which contains few freshwater crab species. Effect of geographical scale on the taxonomic selectivity of data deficiency and extinction risk

No significant taxonomic selectivity in data deficiency or extinction risk was detected in the Afrotropical realm. In the Indomalayan and Neotropical realms (Table 5), data deficiency was not randomly distributed across families, and selectivity of extinction risk varied depending on the scenarios considered. Selectivity of extinction risk was only significant in the Indomalayan realm when DD species where excluded or considered non-threatened, whereas in the Neotropical realm selectivity was only significant when DD species were excluded or considered threatened. In these cases, the distribution of threat among families was congruent across scenarios. We investigated the presence of taxonomic selectivity of dragonflies in the Afrotropical, Australasian, Indomalayan, Neotropical and Palearctic realms (Table 6). The distribution of DD species and threatened species under the three scenarios was not significantly different from random in the Australasian realm. Data deficiency was non-randomly distributed in the remaining realms, and we generally detected no selectivity of extinction risk when DD species were excluded or considered non-threatened. On the other hand, taxonomic selectivity was highly significant when DD species were considered threatened. DISCUSSION

Only the Nearctic realm was suitable for the investigation of the taxonomic selectivity of crayfish at the sub-global level, and showed no significant selectivity in data deficiency or extinction risk. For freshwater crabs, the Afrotropical, Indomalayan and Neotropical realms were suitable for the investigation of taxonomic selectivity at the sub-global level.

Taxonomic and geographical selectivity of data deficiency Multiple factors affect information availability for any given species and directly contribute to observed patterns of data deficiency among taxa. One potential concern could be a

Table 3 Global taxonomic selectivity of data deficiency and extinction risk in crayfish, freshwater crabs and dragonflies. The observed and expected distributions of DD (Data Deficient) or threatened species among families are indicated along with the results of the Fisher’s exact tests Data Deficiency Family

Observed

Crayfish

ns

Astacidae Cambaridae Parastacidae

3 91 31 ***

Gecarcinucidae Parathelphusidae Potamidae Potamonautidae Pseudothelphusidae Trichodactylidae

28 116 305 31 144 8

Dragonflies

***

ns, non-significant. P < 0.05; **P < 0.01;

1214

Trend

Observed

Expected

Trend

***

Freshwater crabs

*

Expected

DD species excluded

2.2 82.9 39.9

+ +

3 71 72

+ + +

9 77 53 28 31 4 ***

DD species threatened

Observed

Observed Expected

Expected

Trend

*** 2.4 95.3 48.3

+ +

** 26.5 142.9 249.1 65.5 124.5 23

DD species non-threatened

3 71 72

*** 2.6 96.9 46.6

+ +

* 8.1 54 62.2 31.9 33.7 12.1

+ +

9 77 53 28 31 4 *

Trend

6 162 103

4.7 180 86.3

+

35.2 188.8 328.8 86.6 164 30.6

+ + +

+

*** 8.5 45.7 79.7 21 39.8 7.4

+ + +

37 193 358 59 175 12

+

***

***

P < 0.001.


Extinction risk selectivity in freshwater invertebrates Table 4 Geographical selectivity of data deficiency and extinction risk in crayfish, freshwater crabs and dragonflies. The observed and expected distributions of DD (Data Deficient) or threatened species among biogeographical realms are indicated along with the results of the Fisher’s exact tests Biogeographical realms Crayfish

Data Deficiency Observed

DD species excluded

Expected

Trend

***

Afrotropics Australasia Nearctic Neotropics Palearctic

***

Afrotropics Australasia Indomalaya Neotropics Palearctic

32 21 426 152 5

Dragonflies

***

Afrotropics Australasia Indomalaya Nearctic Neotropics Oceania Palearctic

76 84 185 13 132 15 57


*

Expected

Trend

***

4 17 81 19 6

Freshwater crabs

Observed

+

1.5 32 81.2 9.5 2.6

+ +

2 70 65 10 2

+ +

0.9 42.3 95.3 8.2 1.9

+ +

+

Observed

Observed

+ +

2 70 65 10 2

+ +

Expected

Trend

28 2 136 35 1

32.1 3.7 114.6 45.4 6.2

+ +

22 21 55 2 23 0 10

28 2 136 35 1

1.7 37.7 95.2 11.2 3

+ +

+ + +

22 21 55 2 23 0 10

Trend

6 87 146 29 8

3.2 69.7 176.4 20.8 5.5

+ + + +

*** 21.3 5.1 124.7 46.9 3.9

+ +

*** 21.1 14.7 26 19 35.8 0.7 15.6

Expected

***

*

*** 84.3 69.7 136.5 56.9 144.4 7.3 62.9


***

** 67.2 16.3 392.7 147.6 12.3


60 23 561 187 6

88.4 21.1 516.9 194.1 16.3

+ +

*** 19.9 16.5 32.3 13.5 34.1 1.8 14.9

+ + +

98 105 240 15 155 15 67

104.4 86.4 168.6 70.4 178.6 9.1 77.7

+ +

+

***

P < 0.001.

Table 5 Taxonomic selectivity of data deficiency and extinction risk of freshwater crabs within biogeographical realms. The observed and expected distributions of DD (Data Deficient) or threatened species among families are indicated along with the results of the Fisher’s exact tests. No significant taxonomic selectivity in data deficiency or extinction risk was detected in the Afrotropical realm. There was no significant taxonomic selectivity of extinction risk when DD species were considered threatened in the Indomalayan realm or considered non-threatened in the Neotropical realm Data deficiency*** Indomalaya

Observed

Expected

Gecarcinucidae Parathelphusidae Potamidae

28 94 304

28.9 137.4 259.8

DD species excluded**

DD species non-threatened**

Trend

Observed

Expected

Trend

Observed

Expected

Trend

9.5 59.8 66.7

+

9 75 52

9.2 43.9 82.9

+

+

9 75 52

Data deficiency***

DD species excluded*

DD species threatened***

Neotropics

Observed

Expected

Trend

Observed

Expected

Trend

Observed

Expected

Trend

Pseudothelphusidae Trichodactylidae

144 8

128.1 23.9

+

31 4

25.7 9.3

+

175 12

157.6 29.4

+


*

***

P < 0.001.

difference in assessor attitude across taxa or regions, resulting in different patterns of data deficiency, and therefore impacting our results. IUCN’s quantitative criteria for the assignation of Red List categories ensures that in threatened

categories at least, subjectivity of assessors plays a minor role in differences across taxa and regions. The case is less clear for DD species listings, where this category marks a lack of information, or understanding, on a given taxon’s status.


1215

L.M. Bland et al. Table 6 Taxonomic selectivity of data deficiency and extinction risk of dragonflies within biogeographical realms. There was no significant taxonomic selectivity in data deficiency or extinction risk in the Australasian realm

Afrotropics Indomalaya Neotropics Palearctic

Data Deficiency

DD species excluded



*** *** *** ***

ns ** ns ns

ns ns ns ***

*** *** *** ***

DD, Data Deficient; ns, non-significant. * P < 0.05; **P < 0.01; ***P < 0.001.

The attitude of the assessors involved in the crayfish, freshwater crab and dragonfly assessments towards DD listings may vary; however, clear guidelines were used to assign this category, all the assessments we used were coordinated by one of the authors (B.C.), and were passed through the IUCN verification system, which should go some way to minimizing any potential effect on our results. The Nearctic and Australasian realms are well-studied centres of crayfish diversity, containing 91% of all crayfish species. The other realms, despite containing 9% of species, contain 23% of the DD species. In these realms information on population trends is especially scarce and there is little understanding of the effects of threats on crayfish populations. While DD crayfish typically lack detailed information on the impact of threats and trajectory of population trends, most DD freshwater crabs are only known from one or two geographical locations, with little or no information on their extent of occurrence, ecological requirements and population size (Table 2). Moreover, DD freshwater crab species are concentrated in species-rich clades and regions, such as the family Potamidae and the Indomalayan realm. Freshwater crabs exhibit high levels of endemicity, have restricted ranges and often occupy remote habitats (Cumberlidge et al., 2009), which may be the reason why, in association with limited monitoring effort, many freshwater crab species are assessed as DD. While some DD freshwater crabs may be naturally rare and therefore more likely to be classified in a non-threatened category, there is still a high chance that some species, which have not been observed in decades (e.g. Rouxana papuana from Indonesia has not been observed in over a century) and/or have had their habitat transformed by human activity (e.g.Thaipotamon siamense from Thailand), may well be extinct. For such species, the only recourse is to initiate targeted surveys to confirm status. Freshwater crabs and dragonflies exhibit similar patterns in the selectivity of data deficiency. First, the geographical distribution of data deficiency in freshwater crabs and dragonflies is consistent with the commonly observed tropical biodiversity data gap (Collen et al., 2008). Second, most DD dragonfly species are only known from a very few locations and specimens, as for freshwater crabs. The lack of information about DD dragonflies is particularly alarming, as records for 168 species are from unknown provenance; these species will be especially difficult to re-assign to non-DD categories in the future.

1216

Taxonomic and geographical selectivity of extinction risk The three taxonomic groups under consideration not only show considerable differences in patterns of selectivity of extinction risk, but also in the influence of data uncertainty on these patterns. Australasian crayfish species (parastacids) remained over-threatened and Nearctic species (mostly cambarids) under-threatened in all scenarios, indicating that genuine differences in extinction risk exist between the two groups. Freshwater crayfish in the Australasian realm are particularly exposed to the threats of sedimentation because of agricultural and forestry effluents, habitat destruction, bush fires, droughts and over-exploitation. Australian crayfish also appear to be extremely susceptible to the effects of climate change, including increasing temperatures, alterations to hydrological regimes and loss of suitable highland habitat (Chiew & McMahon, 2002; Hughes, 2003). The strength of extinction risk selectivity varied across scenarios in freshwater crabs and dragonflies, with selectivity being reduced when DD species were considered nonthreatened. This is because of the overall lower prevalence of threat in all families when DD species are considered non-threatened. The freshwater crab families Gecarcinucidae, Parathelphusidae and Trichodactylidae were consistently over or under-threatened across all scenarios of risk distribution, whereas patterns varied across scenarios for the remaining families. Similarly, the reliability of extinction risk trends among scenarios varied among biogeographical realms. Our results indicate that the current understanding of patterns of threat in freshwater crabs is heavily influenced by data uncertainty. Additional work to determine the true extinction risk of these species is therefore needed before the results of conservation prioritization using these data can be viewed with confidence. Formulating hypotheses concerning the relative roles of biological traits and threatening processes in determining extinction risk in the taxon is also problematic. Semi-terrestrial species, stenotopic species and island endemics are thought to be more susceptible to anthropogenic habitat disturbance (Cumberlidge et al., 2009), but evidence is scarce and mostly derived from other taxa such as amphibians (Sodhi et al., 2008). The extent to which there is congruence in the predictors of extinction risk across freshwater taxa is therefore a useful avenue of further research.


Extinction risk selectivity in freshwater invertebrates Dragonflies showed very strong geographical non-randomness of extinction risk in all scenarios and qualitatively consistent trends across biogeographical realms. Species in the Indomalayan realm were more threatened than expected by chance regardless of how DD species were assigned, which may be due to the high number of endemic species in the Indomalayan islands and large-scale logging of lowland forests (Clausnitzer et al., 2009), while the Nearctic and Neotropical realms were consistently under-threatened. Temperate species, including Nearctic species, have suffered declines in the second half of the 20th century but many are recovering because of improved water management (Kalkman et al., 2008). These species also tend to have wider distributions than their tropical counterparts and may be more able to recover from local-scale population decline (for a global review of threats affecting dragonflies, see articles in Clausnitzer & Jo¨dicke, 2004). Our results indicate that the geographical distribution of threat in crayfish and dragonflies is reliable and could be used to determine the allocation of conservation resources to certain geographical regions with confidence. The effects of geographical scale on the taxonomic selectivity of data deficiency and extinction risk Taxonomic selectivity of extinction risk at the global level may simply be a by-product of geographical selectivity, because clades endemic to certain regions experience different intensities of threatening processes (Russell et al., 1998; Bielby et al., 2006). If this were the case, at a fine scale, each family should display similar levels of threat. On the other hand, the persistence of taxonomic selectivity at smaller geographical scales would suggest that biological differences are at least partially responsible for the observed selectivity. The effect of taxonomic selectivity within geographical regions could not be investigated in detail in crayfish because crayfish families do not often co-occur in biogeographical realms. However, there is evidence for significant differences in extinction risk between genera of Nearctic crayfish living in the same state (Adamowicz & Purvis, 2006). While at the global level, it may be difficult to disentangle the effects of common evolutionary history from the geographical distribution of threatening processes on crayfish extinction risk, at smaller geographical scales species traits seem to be an important factor. This finding has important implications for the creation of predictive models of extinction risk in crayfish, in which both biological traits and geographical factors (and their interaction) are likely to determine extinction risk. The analysis of taxonomic selectivity at sub-global scales in freshwater crabs and dragonflies revealed some complex patterns in information availability and threat prevalence. Data deficiency was generally unevenly distributed among freshwater crab and dragonfly families within realms, the sole exception being dragonflies in Australasia. Freshwater crab and dragonfly families differ in information availability, and poorly known freshwater crab and dragonfly families should be the target of conservation research at sub-global scales.

There was little consistent evidence for taxonomic selectivity of extinction risk at sub-global scales in both taxa, either among biogeographical realms or among scenarios. We therefore can neither accept nor refute the hypothesis that geographical differences in threat intensity are responsible for the non-random pattern in extinction risk at the global level. Our results indicate that when there is a large proportion of DD assessments for any given group, the formulation of a clear understanding of the effects of family-specific attributes on extinction risk is difficult. CONCLUSIONS Freshwater invertebrate conservation faces huge challenges because of the increasing pressures humans are imposing on freshwater systems (Jackson et al., 2001; Malmqvist & Rundle, 2002) and the very limited resources, both in terms of money and scientific effort, allocated to their conservation (Strayer, 2006). For these species, the analysis of global datasets, as opposed to the study of local species and populations, could provide more widely applicable results and recommendations while taking into account heterogeneity among phylogenetic and geographical subsets. In this study, we investigated the robustness of observed macro-ecological patterns of threat in freshwater invertebrates to high levels of data uncertainty. The three taxa included in our analyses all show large differences in ecology, geographical distribution, levels of data uncertainty and threat, and give different perspectives on the selectivity of extinction risk in the freshwater realm. Our study shows that the effect of DD species on the selectivity of extinction risk is not only dependent on the absolute number of DD species in the taxon, but also on the distribution among families and realms of these DD species. Global patterns of taxonomic selectivity and geographical selectivity were generally consistent with one another and robust to different treatments of DD species. However, at sub-global scales, it was not possible to disentangle the effects of common evolutionary descent from those of information availability on extinction risk selectivity. Taxonomic selectivity in amphibians has been shown to be independent of both geographical effects and differences in knowledge of species conservation status (Bielby et al., 2006). However, given the current amounts of data uncertainty, the relative importance of family-specific characteristics and threatening processes in driving extinctions in freshwater invertebrates cannot be determined. In the future, phylogenetically controlled correlative analyses conducted at the species level could highlight which biological traits infer low or high susceptibility to threatening processes. While the understanding of extinction risk in freshwater invertebrates remains compromised by high levels of data deficiency, prioritization of freshwater invertebrates for conservation at the sub-global scale remains a challenge. Given the significant impact of DD species on the understanding of patterns of risk of invertebrates, DD species should be given high


1217

L.M. Bland et al. research priority to determine their true status. Ideally, this should be carried out through field assessments, but the use of contextual information, expert opinion, techniques combining information about collection efforts with the geographical location of specimens (Good et al., 2006) or the outputs of predictive extinction risk modelling may also allow the preliminary re-assignment of a large number of DD species. Our study constitutes a coarse analysis of the factors that determine global extinction risk in invertebrates, and as such observed patterns cannot be easily attributed to processes that occur at a local scale. Additionally, different proportions of DD species from each family or realm may be threatened, rather than none or all, and as a consequence, our analyses are likely to underestimate the effect of DD species on the distribution of threatened species. Our overall findings for each group could also be sensitive to the assessors’ attitude to Red Listing and the number of species yet to be discovered. Any statements made on the reliability of observed trends are therefore limited by the scenarios considered and the best information available to date. However, the analyses presented here encompass a wide range of scenarios, and our conservative approach further highlights the crucial influence of DD species in the selectivity of extinction risk. Broadening the coverage of biodiversity assessments to under-studied taxa and systems is essential to developing a more representative picture of biodiversity. Our study shows that despite recent efforts towards achieving this goal, high level of data deficiency challenge the integration of these assessments into conservation decision-making and supports the need for increased efforts in invertebrate study and conservation.

ACKNOWLEDGEMENTS We thank Nadia Richman and Helen Meredith for valuable discussion, the Esmee Fairbairn Foundation for funding much of the original data collection and the Rufford Foundation (B.C.). We thank Viola Clausnitzer, Neil Cumberlidge and Nadia Richman for the coordination of the Red List assessments of the groups we have examined. We also thank the many contributors to the IUCN Red Listing process.

REFERENCES Adamowicz, S.J. & Purvis, A. (2006) Macroevolution and extinction risk patterns in freshwater crayfish. Freshwater Crayfish, 15, 1–23. Baillie, J.E.M., Collen, B., Amin, R., Akcakaya, H.R., Butchart, S.H.M., Brummitt, N., Meagher, T.R., Ram, M., Hilton-Taylor, C. & Mace, G.M. (2008) Toward monitoring global biodiversity. Conservation Letters, 1, 18–26. Benjamini, Y. & Hochberg, Y. (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society Series B, 57, 289–300.

1218

Bennett, P.M. & Owens, I.P.F. (1997) Variation in extinction risk among birds: chance or evolutionary predisposition? Proceedings of the Royal Society B, 264, 401–408. Bielby, J., Cunningham, A.A. & Purvis, A. (2006) Taxonomic selectivity in amphibians: ignorance, geography or biology? Animal Conservation, 9, 135–143. Brooks, D.R. & McLennan, D.A. (1993) Historical ecology: examining phylogenetic components of community evolution. Species diversity in ecological communities (ed. by M. C. Ricklefs), pp. 267–280. University of Chicago Press, Chicago. Butchart, S.H.M. & Bird, J.P. (2010) Data Deficient birds on the IUCN Red List: what don’t we know and why does it matter? Biological Conservation, 143, 239–247. Butchart, S.H.M., Walpole, M., Collen, B. et al. (2010) Global biodiversity: indicators of recent declines. Science, 328, 1164–1168. Cardillo, M., Purvis, A., Sechrest, W., Gittleman, J.L., Biebly, J. & Mace, G.M. (2004) Human population density and extinction risk in the world’s carnivores. PLoS Biology, 2, 909–914. Cardillo, M., Mace, G.M., Jones, K.E., Biebly, J., BinindaEdmonds, O.R.P., Sechrest, W., Orme, C.D.L. & Purvis, A. (2005) Multiple causes of high extinction risk in large mammal species. Science, 309, 1239–1241. Cardillo, M., Mace, G.M., Gittleman, J.L., Jones, K.E., Biebly, J. & Purvis, A. (2008) The predictability of extinction: biological and external correlates of decline in mammals. Proceedings of the Royal Society B, 275, 1441–1448. Chiew, F.H.S. & McMahon, T.A. (2002) Modelling the impacts of climate change on Australian streamflow. Hydrological Processes, 16, 1235–1245. Clausnitzer, V. & Jo¨dicke, R. (2004) Guardians of the watershed. Global status of dragonflies: critical species, threat and conservation. International Journal of Odonatology, 7, 189–206. Clausnitzer, V., Kalkman, V.J., Ram, M., Collen, B., Bailie, J. M., Bedjanic, M., Darwall, W.R.T., Dijkstra, K.-D.B., Dow, R., Hawking, J., Karube, H., Malikova, E., Paulson, D., Schutte, K., Suhling, F., Villanueva, R.J., von Ellenrieder, N. & Wilson, K. (2009) Odonata enter the biodiversity crisis debate: the first global assessment of an insect group. Biological Conservation, 142, 1864–1869. Collen, B., Ram, M., Zamin, T. & McRae, L. (2008) The tropical biodiversity data gap: addressing disparity in global monitoring. Tropical Conservation Science, 1, 75–88. Collen, B., Ram, M., Dewhurst, N., Clausnitzer, V., Kalkman, V.J., Cumberlidge, N. & Bailie, J.E.M. (2009) Broadening the coverage of biodiversity assessments. Wildlife in a changing world. An analysis of the 2008 IUCN Red List of Threatened Species (ed. by J.C. Vie, C. Hilton-Taylor and S.N. Stuart), pp. 67–75. IUCN, Gland, Switzerland. Cooper, N., Bielby, J., Thomas, G.H. & Purvis, A. (2008) Macroecology and extinction risk correlates of frogs. Global Ecology and Biogeography, 17, 211–221. Cumberlidge, N., Ng, P.K.L., Yeo, D.C.J., Magalha˜es, C., Campos, M.R., Alvarez, F., Naruse, T., Daniels, S.R., Esser,


Extinction risk selectivity in freshwater invertebrates L.J., Attipoe, F.Y.K., Clotilde-Ba, F.-L., Darwall, W., McIvor, A., Baillie, J.E.M., Collen, B. & Ram, M. (2009) Freshwater crabs and the biodiversity crisis: importance, threats, status, and conservation challenges. Biological Conservation, 142, 1665–1673. Davies, R.G., Orme, C.D.L., Olson, V., Thomas, G.H., Ross, S.G., Ding, T.-S., Rasmussen, P.C., Stattersfield, A.J., Bennett, P., Blackburn, T.M., Owens, I. & Gaston, K.J. (2006) Human impacts and the global distribution of extinction risk. Proceedings of the Royal Society B, 273, 2127–2133. Diniz-Filho, J.A.F., De Marco, P. Jr & Hawkins, B.A. (2010) Defying the curse of ignorance: perspectives in insect macroecology and conservation biogeography. Insect Conservation and Diversity, 3, 172–179. Good, T.C., Zjhra, M.L. & Kremen, C. (2006) Addressing Data Deficiency in classifying extinction risk: a case study of a radiation of Bignoniaceae from Madagascar. Conservation Biology, 20, 1099–1110. Hilton-Taylor, C., Pollock, C.M., Chanson, J.S., Butchart, S. H.M., Oldfield, T.E.E. & Katariya, V. (2009). State of the world’s species. Wildlife in a changing world. An analysis of the 2008 IUCN Red List of Threatened Species (ed. by J.-C. Vie´, C. Hilton-Taylor and S.N. Stuart), pp. 15–41. IUCN, Gland, Switzerland. Hughes, L. (2003) Climate change and Australia: trends, projections and impacts. Austral Ecology, 28, 423–443. IUCN (2001) IUCN Red List Categories and Criteria: version 3.1. I. Species Survival Commission, Gland, Switzerland and Cambridge, UK. IUCN (2010) Guidelines for using the IUCN Red List Categories and Criteria. Version 8.1 (August 2010). Available at: http://www.iucnredlist.org/documents/RedListGuidelines. pdf (accessed17 January 2010) Jackson, R.B., Carpenter, S.R., Dahm, C.N., McKnight, D.M., Naiman, R.J., Postel, S.L. & Running, S.W. (2001) Water in a changing world. Ecological Applications, 11, 1027–1045. Kalkman, V., Clausnitzer, V., Dijkstra, K.-D., Orr, A., Paulson, D. & van Tol, J. (2008) Global diversity of dragonflies (Odonata) in freshwater. Hydrobiologia, 595, 351–363. Kerr, J.T. & Currie, D.J. (1995) Effects of human activity on global extinction risk. Conservation Biology, 9, 1528–1538. Leader-Williams, N., Adams, W.M. & Smith, R.J. (2010) Deciding what to save: trade-offs in conservation. Trade-offs in conservation: deciding what to save (conservation science and practice) (ed. by N. Leader-Williams, W.M. Adams and R.J. Smith), p. 432. Wiley-Blackwell, Chichester, UK. Lee, T.M. & Jetz, W. (2010) Unravelling the structure of species extinction risk for predictive conservation science. Proceedings of the Royal Society B, 278, 1329–1338. Malmqvist, B. & Rundle, S. (2002) Threats to the running water ecosystems of the world. Environmental Conservation, 29, 134–153. May, R.M., Lawton, J.H. & Stork, N.E. (1995) Assessing extinction rates. Extinction rates (ed. by J.H. Lawton and R.M. May), pp. 1–24. Oxford University Press, Oxford, UK.

McKinney, M.L. (1997) Extinction vulnerability and selectivity: combining ecological and paleontological views. Annual Review of Ecology and Systematics, 28, 495–516. Millenium Ecosystem Assessment (2005) Ecosystems and human well being: biodiversity synthesis. W.R. Institute, Washington, D.C. Nee, S. & May, R.M. (1997) Extinction and the loss of evolutionary history. Science, 278, 692–694. Olson, D.M., Dinerstein, E., Wikramanayake, E.D., Burgess, N.D., Powell, G.V.N., Underwood, E.C., D‘Amico, J.A., Itoua, I., Strand, H.E., Morrison, J.C., Loucks, C.J., Allnutt, T.F., Riketts, T.H., Kura, Y., Lamoreux, J.F., Wettengel, W. W., Hedao, P. & Kassem, K.R. (2001) Terrestrial ecoregions of the world: a new map of life on Earth. BioScience, 51, 933–938. Owens, I. & Bennett, P.M. (2000) Ecological basis of extinction risk in birds: habitat loss versus human persecution and introduced predators. Proceedings of the National Academy of Sciences USA, 97, 12144–12148. Purvis, A., Agapow, P.-M., Gittleman, J.L. & Mace, G.M. (2000a) Nonrandom extinction and the loss of evolutionary history. Science, 288, 328–330. Purvis, A., Gittleman, J.L., Cowlishaw, G. & Mace, G.M. (2000b) Predicting extinction risk in declining species. Proceedings of the Royal Society B, 267, 1947–1952. Purvis, A., Cardillo, M., Grenyer, R. & Collen, B. (2005) Correlates of extinction risk: phylogeny, biology, threat and scale. Phylogeny and conservation (ed. by A. Purvis, J.L. Gittleman and T.M. Brooks). Cambridge University Press, Cambridge, UK. R Development Core Team (2010) R 2.12.0: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Revenga, C., Campbell, I., Abell, R., de Villiers, P. & Bryer, M. (2005) Prospects for monitoring freshwater ecosystems towards the 2010 targets. Philosophical Transactions of the Royal Society B (Biological Sciences), 360, 397–413. Russell, G.J., Brooks, T.M., McKinney, M.M. & Anderson, C.G. (1998) Present and future taxonomic selectivity in bird and mammal extinctions. Conservation Biology, 12, 1365–1376. Samways, M. & Bo¨hm, M. (2010) Invertebrata. Are vertebrates representative of animal biodiversity as a whole? Evolution lost: status and trends of the world’s vertebrates (ed. by J.E.M. Bailie, J. Griffiths, S.T. Turvey, J. Loh and B. Collen), pp. 55 –61. Zoological Society of London, London, UK. Sodhi, N.S., Bickford, D., Diesmos, A.C., Lee, T.M., Koh, L. P., Brook, B.W., Sekercioglu, C.H. & Bradshaw, C.J.A. (2008) Measuring the meltdown: drivers of global amphibian extinction and decline. PLoS ONE, 3, e1636. Strayer, D.L. (2006) Challenges for freshwater invertebrate conservation. Journal of the North American Benthological Society, 25, 271–287. Strayer, D.L. & Dudgeon, D. (2010) Freshwater biodiversity conservation: recent progress and future challenges. Journal of the North American Benthological Society, 29, 344– 358.


1219

L.M. Bland et al. Stuart, S.N., Chanson, J.S., Cox, N.A., Young, B.E., Rodrigues, A.S.L., Fischman, D.L. & Waller, R.W. (2004) Status and trends of amphibian declines and extinctions worldwide. Science, 306, 1783–1786. Vorosmarty, C.J., McIntyre, P.B., Gessner, M.O., Dudgeon, D., Prusevich, A., Green, P., Glidden, S., Bunn, S.E., Sullivan, C.A., Liermann, C.R. & Davies, P.M. (2010) Global threats to human water security and river biodiversity. Nature, 467, 555–561. SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article: Appendix S1 Global taxonomic selectivity of data deficiency and extinction risk in dragonflies.

delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. BIOSKETCH Lucie Bland is a PhD student at the Institute of Zoology, Zoological Society of London and Imperial College London. Her research programme focuses on the effects of data uncertainty and availability on current conservation approaches, as well as the macro-ecology and macro-evolution of freshwater invertebrates. Author contributions: L.M.B. designed and performed all the analyses and wrote the manuscript. J.B., B.C. and C.D.L.O. designed the analyses and wrote the manuscript. Editor: Anthony Ricciardi

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online

1220
