Does land use affect perceptual range? Evidence from two marsupials of the Atlantic Forest

Perceptual range is one of the main determinants of dispersal success in fragmented landscapes, which are composed of scattered remnants of original habitat in a matrix of variable composition. How perceptual range varies according to land use is essential information to estimate the functional connectivity of landscapes. We determined for the first time how different types of land use affect the perceptual range of a species, using as model organisms two neotropical marsupials endemic to the Atlantic Forest in Brazil (Philander frenatus and Didelphis aurita). We released and tracked the movements of 196 individuals in three types of land use commonly found in fragmented landscapes: manioc plantation, mowed pasture and abandoned pasture. We also determined how orientation to the nearest forest fragment is affected by distance to the fragment, wind speed, body mass and sex using a model selection approach. The type of land use affected the perceptual ranges of both marsupials. The estimated perceptual ranges for P. frenatus and D. aurita were 100 and 200 m in the mowed pasture, respectively, 50 and <30 m in the abandoned pasture and 30 and 50 m in the plantation. The orientation of both species decreased with increasing distance to the fragment, but for D. aurita orientation also increased with the wind speed and body mass. These results agree with previous studies depicting a general pattern of increased perceptual range with lower vegetation obstruction in the matrix and larger body mass and wind speed, depending on the use of visual versus olfactory cues by animals. Our findings allow more realistic estimates of functional connectivity in fragmented landscapes based on basic information on the biology of each species and the type of matrix.