Can Morphology Predict the Conservation Status of Iguanian Lizards?

The integrity of regional and local biological diversity is under siege as a result of multiple anthropogenic threats. The conversion of habitats, such as rain forests, into agricultural ecosystems, reduces the area available to support species populations. Rising temperatures and altered rainfall patterns lead to additional challenges for species. The ability of conservation biologists to ascertain the threats to a species requires data on changes in distribution, abundance, life history, and ecology. The International Union for the Conservation of Nature (IUCN) uses these data to appraise the extinction risk for a species. However, many species remain data deficient (DD) or unassessed. Here, I use 14 morphological traits related to locomotor function, habitat, and feeding to predict the threat status of over 400 species of lizards in the infraorder Iguania. Morphological traits are an ideal proxy for making inferences about a species' risk of extinction. Patterns of morphological covariation have a known association with habitat use, foraging behavior, and physiological performance across multiple taxa. Results from phylogenetic general linear models revealed that limb lengths as well as head characters predicted extinction risk. In addition, I used an artificial neural network (ANN) technique to generate a classification function based on the morphological traits of species with an assigned IUCN threat status. The network approach identified eight morphological traits as predictors of extinction risk, which included head and limb characters. The best supported model had a classification accuracy of 87.4%. Moreover, the ANN model predicted >18% of DD/not assessed species were at risk of extinction. The predicted assessments were supported by other sources of threat status, for example, Convention on International Trade in Endangered Species appendices. Because of the functional link between morphology, performance, and ecology, an ecomorphological approach may be a useful tool for rapid assessment of DD or poorly known species.