Individual-based movement models reveals sex-biased effects of landscape fragmentation on animal movement

Animal movement and behavior may respond to habitat modification or fragmentation in non trivial ways, thereby strongly conditioning the fate of populations. This study aims to understand movement patterns of non-dispersal animals in both natural and altered landscapes, using the endangered terrestrial tortoise Testudo graeca as example. We used individual-based simulation models representing competing hypotheses on tortoise movement. Model parameterization and selection was based on radiotracking data and an inverse approach that is able to deal with observation uncertainty, individual variability, and process stochasticity. We find that land use intensification had a strong impact on the movement and behavior of non-dispersing individuals of T. graeca. In natural landscapes, males and females showed a similar movement and behavior profile with a strong home behavior component, and little individual variability. However, in altered landscapes, movement and behavior greatly varied among individuals, particularly in females, and males and females showed different movement patterns. Females showed a wide range of movement patterns, from strong home behavior to an unbounded movement. Our study shows that population or movement models that assume single behavioral states for animals inhabiting different landscape structures can be strongly misleading and, furthermore, that the impact of landscape modification on movement and behavioral patterns can be strongly sex-biased. Flexible, individual-based movement models coupled with inverse parameterization and model selection approaches proved useful in understanding the mechanisms controlling animal movement patterns.