Technical Comments

Response to Comment on “Status and Trends of Amphibian Declines and Extinctions Worldwide”

Science  23 Sep 2005:
Vol. 309, Issue 5743, pp. 1999c
DOI: 10.1126/science.1113265

Using information on Brazilian species, Pimenta et al. assert that we overestimated the number of threatened amphibians. This claim, based on a misunderstanding of the IUCN–The World Conservation Union Red List criteria and a strongly evidentiary attitude to listing species, almost certainly seriously underestimates the number of threatened amphibians in Brazil.

Much of the argument of Pimenta et al. (1) is based on their belief that IUCN–The World Conservation Union recommends adjusting thresholds for distributional area depending on the taxon in question. This is a misunderstanding: IUCN recommends adjusting the scale at which species are mapped to the taxon in question, not using different thresholds for distribution size (2). Thus, the threshold of 20,000 km2 in criterion B1 (for the Vulnerable category), for example, is the same whether assessing an orchid or a frog (2).

IUCN, with extensive input from many scientists, including the leader of our research (S.N.S), developed the criteria and thresholds as global standards between 1989 and 2001 and tested them on a wide variety of taxa, including plants, marine species, invertebrates, fishes, reptiles, and amphibians (3). Allowing assessors to adjust the thresholds independently would create a chaotic listing process in which two scientists evaluating similar species would rarely choose the same values. The distributional area criterion is not a “straitjacket,” as Pimenta et al. suggest, but a reflection of the intrinsic risks to extirpation that populations occurring in limited geographical areas face as a result of a variety of threats. IUCN developed the Red List criteria so that at least one, but by no means all, of the criteria apply to every species (excluding microorganisms), encompassing the wide diversity of life-history strategies exhibited by life on Earth (2, 3). One would not expect to use the same criteria to assess species as different from each other as a tiger and a leaf-litter frog (4).

That many Brazilian amphibian species are poorly known is not in dispute, and Pimenta et al. demonstrate this well. They assert that Brazilian amphibians are too poorly known for the detailed assessment of conservation status that we carried out (5, 6). However, the Red List categories and criteria represent a coarse, rather than a detailed, measurement of conservation status (7). Data showing species to be more widespread than previously known frequently do not result in a change in the Red List category, because the categories represent broad bands of extinction risk. Therefore, movements of species between Red List categories usually reflect major changes in conservation status (8, 9).

The observed discrepancies between the lists can be accounted for in part by different applications of the Red List criteria (Pimenta et al. admit to using them only tentatively) and by differences in the underlying data (10). However, we believe that many of the discrepancies are due to different attitudes to uncertainty and risk, ranging from evidentiary to precautionary (11, 12). IUCN's guidelines on dealing with uncertainty (13, 14) state that assessors should adopt a “precautionary but realistic” attitude. The Global Amphibian Assessment (GAA) Brazil workshop results represent a strongly evidentiary viewpoint (15). If the approach to identifying threatened species is too evidentiary, the list can fail in its basic purpose of flagging species in urgent need of conservation attention before it is too late. For Brazilian amphibians, this strongly evidentiary attitude leads to a higher percentage of data-deficient (DD) (28%) and nonthreatened (69%) species and a lower percentage of threatened species (3.3%), compared with DD (22%), nonthreatened (63%), and threatened (15%) with the precautionary but realistic “consistent” categories.

New data (16) indicate that perhaps 19 of 38 species in Brazil in which declines have been recorded could be globally at risk (17). Only 5 of these 19 species were listed as threatened in the GAA Brazil workshop, compared with 12 as a result of the consistency check. There is growing evidence (1823) that such declines can rapidly lead to extinction. Under a strongly evidentiary approach, the extinction of species in genera such as Cycloramphus [declines reported or suspected in at least seven species (16)] could occur before any conservation alarm sounds.

We agree with Pimenta et al. that exaggerating the level of threat compromises the credibility of lists. However, minimizing threat levels, especially if extinctions take place, also compromises credibility. Scientists might exaggerate threats to obtain more funding for their research or minimize threats to avoid restrictions on scientific collecting. Both approaches lead to a distortion of the Red List process and support the need for consistency checks.

Pimenta et al. also criticize our concept of rapidly declining species. The use of the IUCN Red List to identify such species is explained by Butchart et al. (8, 24). Of 435 rapidly declining species globally, we identified only 13 from Brazil. The inclusion of two species mentioned by Pimenta et al. (Bokermannohyla claresignata and Scinax heyeri) as rapidly declining was based on the GAA data but could be incorrect in view of the information provided by Pimenta et al. (25). However, Eterovick et al.'s new data (16), with 19 species possibly meeting our definition of rapidly declining (6, 17), indicate that we might have underestimated the number of rapidly declining species in Brazil.

Pimenta et al. raise several important issues with which we agree wholeheartedly: The tendency of governments to restrict scientific collecting of listed species is a disincentive for research on those species that need it most (26, 27); DD species are very high priorities for research funding, especially because many of them are likely to be threatened, and conservation priorities should not be determined using threatened species lists alone (28). We believe that greater harmonization between threatened amphibian lists in Brazil is urgently needed and will most easily be achieved by following the approach suggested by Keith et al. (29, 30). In the longer term, funding and implementing an enhanced program to describe and document Brazil's remarkable amphibian fauna would greatly reduce the level of data uncertainty.

Supporting Online Material

www.sciencemag.org/cgi/content/full/309/5743/1999c/DC1

SOM Text

Fig. S1

Tables S1 and S2

References and Notes

References and Notes

Related Content

Navigate This Article