A not-quite-so inordinate fondness for beetles

Abstract

When asked what insights his study of biology had given him into the nature of a Creator, J.B.S. Haldane replied with his now-legendary quip about beetles. Theological questions aside, the fact that Coleoptera are among the most speciose of orders has important ramifications for a deceptively basic question in conservation biology: how many species are there on earth?Although recent informed guesses span orders of magnitude that are a lesson in scientific humility (from around 1.8 million species currently named, up to perhaps 100 million in total), the answer to this question is crucial for quantifying biodiversity. Of the diverse chains of reasoning available for tackling the how-many-species problem, more than a few rest on the number of tropical beetle species there might be. As pointed out by Terry Erwin in his initial attempt to provide an estimate, the number of tropical beetles itself might rest on other unknown quantities, such as the number of tree species in a typical forest, the degree to which beetles found on those trees are specialized to them, the degree to which canopy beetles are representative of beetles as a whole, and so on.A research note by Odegaard et al.1xThe magnitude of local host specificity for phytophagous insects and its implications for estimates of global species richness. Odegaard, F. et al. Conserv. Biol. 2000; 14: 1182–1186CrossRef | Scopus (43)See all References1 provides data that are designed specifically to address the question of specialization in tropical beetles. Over the course of a year, they hand-collected 35 479 adult individuals from 1167 phytophagous beetle species. Their sample covered 24 tree and 26 liana species from an 8000-m2 area in a dry forest in the Parque Natural Metropolitano, Panama, which they accessed with a canopy tower crane. This produced 2561 observations of host–beetle associations. Using nonparametric subsampling of these data, they calculated ∼30% effective specialization on the 50 plant species examined. However, effective specialization decays as host species richness increases; a sample including just one plant species will necessarily find 100% specialization, whereas sampling two host species will find somewhat less than 100%, because some beetle species will be found on both hosts. Therefore, Odegaard et al. use a parametric model to extrapolate beyond the spatial scale of their study, estimating that, in a typical tropical dry forest with 300 higher plant species, effective specialization would hover around 10%. Thus, host specialization appears to be closer to half of the 20% that was previously thought.Because this information brings down estimates of the number of tropical beetle species, it also brings down estimates of total biodiversity; the number of species on earth might be closer to 5–15 million, rather than the 30 million or more that some have estimated. However, the work of Odegaard et al. only addresses one of the links in the chain of reasoning that leads to larger estimates of global species richness. Other questions, such as the ratio of canopy to ground-dwelling arthropod richness or the proportion of all canopy arthropods that are beetles, remain largely unanswered.