The Race to Name Earth's Species

Abstract

[Extract] In their Review Can we name Earth's species before they go extinct? (25 January, p. 413), M. J. Costello et al. answer in the affirmative. I am not so sure. Costello et al. could be underestimating global biodiversity by concluding that Earth likely sustains only 2 to 8 million species of eukaryotes. Studies they review suggest that there might be 6.1 to 7.8 million species of terrestrial arthropods alone, as well as 0.3 to 1.0 million marine species and 0.3 to 0.4 million species of plants (1). Even if one ignores the relatively small number of vertebrate species, many other taxa—including nematodes and other worm phyla, fungi, and other terrestrial and freshwater invertebrates—could be impressively species-rich. Beyond this, there are myriad microscopic eukaryotes (2). I am also less sanguine than are Costello et al. about the world's undescribed biodiversity. Such species likely include many local endemics (3), small-sized or non-descript species, species that live in poorly explored places such as the rainforest canopy or deep oceans, morphologically cryptic species, and those that simply are very good at hiding (1). Even with improving genetic and other technologies, some of these taxa may be extremely difficult to find or catalog. Moreover, because the missing species are almost certainly a nonrandom component of all biodiversity, our incomplete view of life on Earth likely suffers from various selective biases. Finally, even with growing taxonomic expertise in developing nations, it is far from certain that sufficient effort will be focused on the world's imperiled biodiversity hotspots (4), which are likely to house the bulk of undescribed species (3, 5).