Experimental forest fragmentation alters Amazonian mixed-species flocks

Habitat fragmentation has been associated with myriad negative effects for forest-dependent birds in the Neotropics. However, the vast majority of negative effects have been inferred from comparisons of pre-existing fragments with separate control sites. Such comparisons confound area loss with isolation and ignore effects of patchy distributions and local habitat heterogeneity. To directly test the effects of fragmentation on Amazonian mixed-species flocks-a complex and diverse species interaction network-we observed birds before and after re-isolation of three 10-ha fragments at the Biological Dynamics of Forest Fragments Project in central Amazonia. Following initial isolation in the 1980s, these fragments have been surrounded by a matrix of developing second growth that was cut again in 2013-2014 (re-isolation). Simultaneously, we also followed three control flocks in primary forest that bordered tall secondary forest. We quantified species richness and attendance, home range size, proportional use of edge and second growth, and space use for fragment and control flocks before and after re-isolation. Following re-isolation, one flock disappeared entirely and half of the obligate flock-followers either vanished or decreased attendance rates. Home ranges of fragment flocks shrunk, and movements were confined by newly created hard edges. These results provide direct experimental evidence that isolation leads to the deterioration and collapse of flocks in forest fragments, affecting both direct metrics and emergent properties of a complex social network. This study also provides retrospective insight into the value of adjacent second growth habitat as immigration corridors for birds in mixed-species flocks.