What would it take to describe the global diversity of parasites?

Abstract

How many parasites are there on Earth? Here, we use helminth parasites to highlight how little is known about parasite diversity, and how insufficient our current approach will be to describe the full scope of life on Earth. Using the largest database of host-parasite associations and one of the world's largest parasite collections, we estimate a global total of roughly 100 000-350 000 species of helminth endoparasites of vertebrates, of which 85-95% are unknown to science. The parasites of amphibians and reptiles remain the most poorly described, but the majority of undescribed species are probably parasites of birds and bony fish. Missing species are disproportionately likely to be smaller parasites of smaller hosts in undersampled countries. At current rates, it would take centuries to comprehensively sample, collect and name vertebrate helminths. While some have suggested that macroecology can work around existing data limitations, we argue that patterns described from a small, biased sample of diversity aren't necessarily reliable, especially as host-parasite networks are increasingly altered by global change. In the spirit of moonshots like the Human Genome Project and the Global Virome Project, we consider the idea of a Global Parasite Project: a global effort to transform parasitology and inventory parasite diversity at an unprecedented pace.

Keywords: biodiversity estimation; museum collections; parasites; systematics; tapeworms.

Figure 1.

Rates of helminth (a) descriptions (from NHM data) and (b) collections (from the US National Parasite Collection (USNPC)). Solid lines (blue) indicate cumulative totals, and dashed lines (red) give a breakpoint regression with a single breakpoint (1912 for the NHM data, 1903 for the USNPC data). Although the current trend appears to be levelling off, it is unlikely this indicates a saturating process (as comparably illustrated by the drop in sampling during the Second World War, 1940–1945). (Online version in colour.)

Figure 2.

We found evidence of weak but highly significant declines over time in (a) parasite adult body length (smooth term p = 0.0003) and (b) host body size across known host associations (smooth term p < 0.0001). This confirms a mild description bias for larger parasites in larger hosts. (Online version in colour.)

Figure 3.

The type species (the first described in a genus) has a statistically significantly higher average host range than those that follow. Parasites described earlier typically have a higher degree of generalism (greater number of recorded hosts), especially prior to the 1840s; specimens collected after roughly the 1870s also apparently tend towards more host-specific species than those from older collections. (Curves are generalized additive models fit assuming a negative binominal distribution, with dashed lines for the 95% confidence bounds.) (Online version in colour.)

Figure 4.

The distribution of (a) maximum possible helminth richness in mammals, (b) the number of known helminth parasites of mammals as recorded by country in the NHM data and (c) the maximum percentage of undocumented helminth fauna by country. (Online version in colour.)