Body mass is rarely recorded in amphibians, and other body measurements (e.g. snout to vent length, SVL) are generally collected instead. However, length measurements, when used as proxies of body mass in comparative analyses, are problematic if different taxa and morphotypes are included. We developed allometric relationships to derive body mass from SVL measurements. We fitted phylogenetic generalized least square models for frogs (Anura) and salamanders (Caudata) and for several families separately. We tested whether allometric relationships differed between species with different habitat preferences and between morphs in salamanders. Models were fitted with SVL-mass measurements for 88 frog and 42 salamander species. We assessed the predictive performance of the models by cross-validation. Overall, the models showed high explained variance and low forecasting errors. Models differed among semi-aquatic, terrestrial and arboreal frogs, and between paedomorphic and non-paedomorphic salamanders. Body mass estimates derived from our models allow for comparability of studies on multiple taxa and can be used for testing theories built upon evolutionary and ecological processes which are directly related to body mass.
Length-mass allometries in amphibians
Santini Luca;
2018
Abstract
Body mass is rarely recorded in amphibians, and other body measurements (e.g. snout to vent length, SVL) are generally collected instead. However, length measurements, when used as proxies of body mass in comparative analyses, are problematic if different taxa and morphotypes are included. We developed allometric relationships to derive body mass from SVL measurements. We fitted phylogenetic generalized least square models for frogs (Anura) and salamanders (Caudata) and for several families separately. We tested whether allometric relationships differed between species with different habitat preferences and between morphs in salamanders. Models were fitted with SVL-mass measurements for 88 frog and 42 salamander species. We assessed the predictive performance of the models by cross-validation. Overall, the models showed high explained variance and low forecasting errors. Models differed among semi-aquatic, terrestrial and arboreal frogs, and between paedomorphic and non-paedomorphic salamanders. Body mass estimates derived from our models allow for comparability of studies on multiple taxa and can be used for testing theories built upon evolutionary and ecological processes which are directly related to body mass.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.