The hypothesis that more plant species with vertebrate-dispersed fruits occur on fertile soils because there is a greater probability of fruit removal from the parent plant was tested at 16 sites around Sydney, Australia. Removal rates from artificial fruit spikes were two and a half times greater on fertile than infertile soil sites, although this was not quite statistically significant. High variability in removal rate between sites was evident irrespective of fertility. Most removal occurred during the day indicating that birds were important consumers, rather than nocturnal mammals. Bird abundance and diversity did not differ between soil types. More frugivorous species were found in plant communities growing on fertile soil. Two models could explain the patterns observed. Firstly, plants with vertebrate-dispersed fruits could be favoured on fertile soils because of a high abundance of frugivorous birds accomplishing seed dispersal. Alternatively, plants with vertebrate-dispersed fruits could be favoured on fertile soil sites for some other reason and frugivorous birds could be attracted to these areas of abundant food. The correlation between soil fertility and the percentage of vertebrate-dispersed fruits was stronger than the correlation between soil fertility and removal rates and suggests that the second model is more likely to be true. Frugivorous birds are unlikely to be responsible for the high percentage of species with vertebrate-dispersed fruits in fertile soil environments. �� 1992 Springer-Verlag.