Incorporating movement behavior into conservation prioritization in fragmented landscapes: An example of western hoolock gibbons in Garo Hills, India

Connectivity is increasingly of conservation concern due to ongoing habitat fragmentation, land-use dynamics and climate change. Connectivity patterns result from interactions between dispersers and the environment. However, connectivity assessments often ignore responses of dispersers to matrix types or characterize matrix resistance using habitat-use, rather than movement, data. We compare conservation rankings for connectivity of forest fragments in Garo Hills, India, where matrix resistance was quantified based on (a) distance among fragments, (b) habitat use, and (c) movement constraints for the arboreal western hoolock gibbon Hoolock hoolock. We first quantified matrix resistance based on gibbon movement, in terms of gap-crossing behavior, as a function of canopy gap and tree height, which was estimated using focal scans on seven gibbon groups. Second, we estimated matrix resistance based on gibbon habitat use of major land-use types in the study landscape. We then compared rankings of forest fragments using patch connectivity indices from network analyses using these quantified matrix resistances and Euclidean distance among sites. We found that matrix resistances based on movement data suggested greater resistance of plantations than did habitat-use data. Conservation rankings based on movement data were uncorrelated with those based on Euclidean distance. Rankings derived frommovement data were correlated with those obtained from location data; however, there were several discrepancies between rankings, which were explained by landscape heterogeneity in the neighborhood of fragments. Incorporating movement behavior into connectivity assessments can improve our understanding of dispersal and provide a mechanistic basis for conservation prioritization in heterogeneous