Matrix mediates avian movements in tropical forested landscapes: Inference from experimental translocations

Maintaining animal movement in fragmented landscapes depends upon the levels of connectivity among habitat patches, which in turn may depend upon the landscape matrix. Little is known about how the matrix affects dispersal abilities, in part because few experimental tests have been conducted. We experimentally translocated 142 migratory American Redstarts (Setophaga ruticilla) and resident Jamaican Todies (Todus todus) 0.6-4 km from their territories across landscapes fragmented by pen-urban development and bauxite mining and continuous forest. Redstarts returned more rapidly and with greater success than todies across all landscapes, with 95% of redstarts returning in an average of 2.5 days versus 60% of todies in >20 days. Return success was best predicted by translocation distance for redstarts and by sex for todies, with a trend of fewer birds returning when released in bauxite landscapes. Return time was strongly affected by matrix type, with both species returning more rapidly in a forested relative to a bauxite matrix and intermediately in a pen-urban matrix. These findings provide strong experimental evidence that land cover surrounding forested habitat influences species mobility. (C) 2010 Elsevier Ltd. All rights reserved.