Does population-level risk from movement in human-altered landscapes depend on the historic landscape structure?

Background/Question/Methods

Human land use is widely considered to be a major driver of biodiversity loss; however, species responses to land use are variable. It has been hypothesized that differences in species responses to land use may be due, at least in part, to differences in their movement characteristics. Theoretical and empirical research suggests the balance between the costs and benefits of movement vary with the landscape structure, which in turn may influence the evolution of ‘optimal’ movement characteristics that maximize fitness within a given landscape. Therefore we hypothesized that species responses to land use depend on the historic landscape structure, and the movement characteristics that evolved in response to that structure. To investigate this hypothesis, we developed an individual-based simulation model to simulate the evolution of four characteristics (movement probability, probability of boundary avoidance, movement speed, and path shape in the non-habitat or ‘matrix’) in populations in landscapes which varied in habitat amount (high or low habitat), habitat stability (temporally stable, ephemeral, or disturbed), and matrix quality (high or low quality). We then simulated human land use, and measured changes in population size over time and the extinction probability (i.e. proportion of trials where the population went extinct within 100 years).

Results/Conclusions

As expected, the evolved set of movement characteristics varied with the landscape structure. When subjected to human land use, the extinction probability depended on the landscape structure; temporally stable landscapes (with high or low habitat) had zero probability of extinction while populations in temporally variable landscapes were at risk, with greater extinction probabilities in disturbed than in ephemeral landscapes. For disturbed and ephemeral landscapes, there were greater extinction probabilities in landscapes with low quality matrix. Human land use caused population declines in all landscapes but, surprisingly, we found that differences in the rates of decline in landscapes with different structures did not correspond to the measured extinction probabilities; population declines were greatest in stable landscapes and smallest in disturbed landscapes and, for a given combination of habitat amount and stability, population declines were greater in landscapes with high quality matrix. These results suggest that we may be able to predict the relative extinction risk for species based on knowledge of their historic landscape structure. Furthermore, these results suggest that population size or abundance trends may not be an appropriate indicator of the extinction probability. Future work will focus on evaluating the validity of our predictions using empirical data.