Connectivity and the sensitivity of species to forest loss and fragmentation: multi-species, multi-landscape analysis for birds of the Atlantic rainforest

PE´ER, G.; ZURITA, G.A.; HANSBAUER, M.M.; BANKS-LEITE, C.; HENLE, K.; MARTENSEN, A.; METZGER, J.P.; FRANK, K.

Campos do Jordão, Brazil

Conferencia; 1st Latin American Conference of the International Association of Landscape Ecology (IALE); 2009

Association of Landscape Ecology (IALE)

Introduction: Functional connectivity is the outcome of intricate interactions between the structure of the landscape and animal responses to this structure. Due to the complexity of such animal-landscape interactions, tools are lacking for predicting functional connectivity for multiple species and/or across landscapes and to assess how connectivity affects the sensitivity of species to habitat loss and fragmentation. We therefore developed a spatially explicit, individual-based model (IBM) of connectivity which takes into account that species differ from each other not only in habitat requirements but also in behaviour. One of the model's novel attributes is in separating everyday movements from rare dispersal events, and random movements from directed ones (e.g. gap-crossing between forest fragments). Methods: We used the model to investigate connectivity for birds of the Atlantic forest, using three approaches: first, we modelled connectivity for three focal bird species in Brazil, on the basis of detailed radio telemetry data. Second, we examined connectivity for multiple species within three focal landscapes in the Sao Paulo Plateau, Brazil. Here, we assessed how connectivity is affected by adding corridors. Finally, we explored connectivity across multiple species and multiple landscapes, using empirical data and multiple maps from Argentina and Paraguay. The multiple-species analysis was aided by clustering species into functional types, considering both the habitat requirement of species and their movement behaviour. Results: Analyzing the results for the three species, we found that the model can successfully reproduce the link between biological traits and the sensitivity of species to fragmentation. Particularly, we show that the combination of edge avoidance and territorial behaviour substantially enhance the sensitivity of species to fragmentation. Results for the three landscapes, as well as for the multiple landscapes, indicated that the combination of spatial factors, species traits and behavioural factors interact to determine the spatial pattern of connectivity, species’ sensitivity to fragmentation, the potential efficacy of corridors to enhance connectivity, and even the structure of communities across landscapes. Conclusions: Our work improves connectivity theory along several lines. First, we provide the framework to address connectivity for multiple species in multiple landscapes. Second, we provide the means to link functional connectivity with the sensitivity of species to fragmentation. Third, the model allows comparing different connectivity indices and assessing their predictive powers in different contexts (landscapes or species). And finally, the model allows analysing functional connectivity and determining its effect on species’ persistence in fragmented landscapes, on the basis of empirical data but also, in its absence, using a more theoretical approach. Thus, the model and our approach can potentially serve as a decision-support tool for various conservation projects undertaken in fragmented landscapes.