The pattern of distribution and diversity of avifauna over an urbanizing tropical landscape

An investigation was done to determine the occurrence and composition of avian fauna community in the urbanizing city of Nairobi, Kenya. We conducted bird counts in sample sites randomly distributed over the Nairobi landscape within a two-year period. The proportion of seven different land cover types derived from within a 500 m radius of classified satellite image described the habitat condition of each sample site. Multivariate analysis of the site and species data indicated that savannah vegetation, forest and agriculture land cover types were the main environmental gradients that differentiated the sample sites. Four clusters of sample sites occurred on the ordination plane according to canonical correspondence analysis (CCA). From 50 families of birds observed, species related to bush and scrub habitats occurred at a rate of about 31 %, followed by grassland species at 20 % and forest species at about 16 % rate. Out of CCA, five functional groups of birds were distinguishable. The proportion of sites occupied by birds per functional group and mean count of individual birds declined significantly, p?<?0.0001, as groups of birds changed from being urban-related to savannah and then to forest-related species. One cluster comprised birds marked as having conservation concern status and related mainly to savannah vegetation of scrub and grasslands. In this landscape context, features such as native savannah vegetation and woodland accompanied by a process of controlled land use, could greatly mitigate negative impacts of ecosystem degradation on the sensitive tropical urban bird biodiversity.