Skip to main content
Dryad

Upper thermal limits predict herpetofauna responses to forest edge and cover

Cite this dataset

Watling, James; Veselka, Andrew J.; Aponte-Gutiérrez, Andrés; Medina-Báez, Osmary (2023). Upper thermal limits predict herpetofauna responses to forest edge and cover [Dataset]. Dryad. https://doi.org/10.5061/dryad.dv41ns231

Abstract

Amphibians and reptiles are sensitive to changes in the thermal environment, which varies considerably in human-modified landscapes. Although it is known that thermal traits of species influence their distribution in modified landscapes, how herpetofauna respond specifically to shifts in ambient temperature along forest edges remains unclear. This may be because most studies focus on local-scale metrics of edge exposure, which only account for a single edge or habitat patch. We predicted that accounting for the combined effect of multiple habitat edges in a landscape would best explain herpetofaunal response to thermally-mediated edge effects. We (1) surveyed herpetofauna at two lowland, fragmented forest sites in central Colombia, (2) measured the critical thermal maximum (CTmax) of the species sampled, (3) measured their edge exposure at both local and landscape scales, and (4) created a thermal profile of the landscape itself. We found that species with low CTmax occurred both further from forest edges and in areas of denser vegetation, but were unaffected by the landscape-scale configuration of habitat edges. Variation in the thermal landscape was driven primarily by changes in vegetation density. Our results suggest that amphibians and reptiles with low CTmax are limited by both canopy gaps and proximity to edge, making them especially vulnerable to human modification of tropical forest.

Methods

Animal data (included here) were collected in the field and lab. Landscape data were derived from the vegetation density map (inlcuded here) either directly (tree cover) or using the BioFrag edge response software, available on GitHub.

Usage notes

Summary data may be viewed in Excel and R, whereas spatial data may be viewed in R or GIS software such as ArcMap or QGIS.