The importance of realistic dispersal models in conservation planning: application of a novel modelling platform to evaluate management scenarios in an Afrotropical biodiversity hotspot
I compared understory bird community structure and composition among primary, slightly disturbed, and moderately disturbed forest in the East Usambara Mountains in Tanzania. Comparisons were conducted at two spatiotemporal scales: short term, conducted in 1999 and 2000, in which treatments (disturbance levels) were replicated, and long term, conducted from 1989-2004 along the same disturbance gradient of a subset of the short-term sites. I used capture-recapture models to assess the probability of detection and estimate species richness and population growth rates. The probability of detection of species did not vary significantly among disturbance levels but did vary significantly among species. Over the short- and long-term surveys, estimated species richness did not vary significantly among disturbance levels. Temporal variability in estimated species richness and the relative abundance of guilds did vary significantly among disturbance levels yet was contingent on survey length. The coefficient of variation in species richness over the short-term survey was <5% across all disturbance levels, whereas over the long-term survey it was 35% in slightly disturbed forest, 11% in moderately disturbed forest, and 0% in primary forest. In the short-term survey, zero of seven feeding guilds varied significantly in relative abundance among disturbance levels, whereas over the long-term survey four of seven (57%) feeding guilds did so. Terrestrial insectivores were most adversely affected by forest disturbance. Population growth rates (lamda) between 1989 and 2004 for the more common species of terrestrial insectivores did not vary significantly among disturbance levels and for these species were significantly < or = 1.00, indicating the recovery time for terrestrial insectivores in slightly and moderately disturbed forest is very long. These results illustrate the importance of long-term studies in assessing ecological impacts of forest disturbance and the importance of protecting primary forest in the Eastern Arc Mountains.