Urban land use does not limit weaver bird movements between wetlands in Cape Town, South Africa

Urbanisation often has profound impacts on ecological processes. Management of these impacts is central to urban conservation efforts. We used data from 6591 individually ringed weaver birds from 42 ringing locations to investigate the influence that the urban matrix, as well as patch isolation, size and quality, had on weaver bird movement between ringing locations in Cape Town, a growing city within a global biodiversity hotspot. Distance-based linear models revealed that proximity to other sites was the dominant predictor of weaver movement while the site variables (wetland size and bird abundance) had a limited and inconclusive influence. Once the variation explained by the proximity and site variables had been accounted for, the composition of the surrounding urban matrix, the length of the least cost path between wetlands, and the presence of rivers as potential movement corridors (measured at three spatial scales) all had little influence on weaver movement. Analysis of the weaver bird movement-wetland network using social network analysis showed that the network is simple, clustered, and non-random, with relatively high vulnerability to node loss and some indication of preferential attachment (i.e., increased use of more used sites). Since proximity (site isolation) is the dominant influence on weaver movements, and the network is already sparse, further wetland loss is likely to reduce population viability. Our results match the predictions of classical theory and suggest that patch management will matter more for wetland passerines than matrix management.