Going Underground: What the Natural History Traits of Cave Users Can Tell Us about Cave Use Propensity
Amphibians and reptiles use caves in different lithologies and biomes. However, the ecological reasons for such behavior have so far not been investigated in a broad scale. We surveyed 945 Brazilian caves in three biomes (Caatinga, Cerrado, and the Amazon), from which 269 had frogs and reptiles. We classified species of amphibians (34) and reptiles (19) according to natural history traits including habit, activity, diet composition, foraging strategy, and developmental mode (anurans). We used IUCN species distribution maps to predict what species occur in the vicinities of the caves. We then tested whether specific life history traits were distributed differently between species using and not using caves. We also tested seasonality and occurrence of bodies of water inside caves used by anurans, as well as the area occupied by man-made building and infrastructure around caves as explanatory factors for cave use intensity. Anurans with direct development were more likely to use caves than those with a tadpole stage. Terrestrial anurans were more likely to use caves than arboreal and semiarboreal anurans. No fossorial, semifossorial, or semiaquatic species used caves. Among reptiles, sit-and-wait foragers were more likely to use caves than active foragers. Semiaquatic and terrestrial reptiles were the most likely to use caves. Anurans were most likely to use caves with perennial bodies of water and during the dry season. Our results indicate that species of the herpetofauna are likely to search for resources in the cave habitat as indicated by the relationship between specific life history traits and cave use.Abstract
Distribution of caves with records of amphibians and/or reptiles in three biomes (Caatinga, Cerrado, and the Amazon) in Brazil.
Incidence of amphibian and reptile species in caves according to the absence or presence of intermittent and perennial bodies of water in rainy and dry seasons. Bars correspond to standard errors.
Topology of the amphibian species included in the study for use in the Bayesian generalized multivariate nonlinear multilevel models. Species relationships were compiled from Lobo (1995), Faivovich et al. (2005), Nascimento et al. (2005), Grant et al. (2006), Pyron and Wiens (2011), de Sá et al. (2014), and Taucce et al. (2018).
Topology of reptiles included in the study for use in the Bayesian generalized multivariate nonlinear multilevel models. Species relationships were compiled from Bertolotto (2006), Pyron et al. (2013), and Hastings et al. (2016).
Contributor Notes
