The ecology of the Madagascar buzzard (Buteo brachypterus) was studied to determine the potential for using the species’ ecological requirements as a criterion for reserve design and management. The study was conducted on the Masoala Peninsula of northeastern Madagascar over two breeding seasons from August to December 1992 and from September 1992 to January 1993. Nest-site characteristics, breeding biology, and food habits were examined and compared between buzzards nesting in primary forest and slash-and-burn clearings. Seven nest sites in primary forest, six nest sites in slash-and-burn clearings, and 21 random sites were chosen for habitat analysis. Fourteen habitat variables were recorded at each site, and the density of nine different vegetation classes was determined from 20 m arm’s width belt transects at each site.
The buzzards selected prominent nest trees on west- or north-facing slopes. Nest sites were significantly farther than random sites from other nest sites, had a lower forest canopy height, and had a lower density of understory trees. Observations of the buzzard’s breeding biology were made at eight nests in 1991 and six nests in 1992. Data were taken on breeding density, nest dispersion, breeding season, nesting success, productivity, nest re-occupancy, and breeding behavior. Breeding densities of 0.79 an d0.60 nests per km2 were determined for 1991 and 1992, respectively. The mean minimum inter-nest distance was 1.0 + 0.2 km(SD) in 1991 and 1.1+ 0.5 km in 1992. Nest dispersion was uniform in 1991 but did not significantly differ from random in 1992.
The breeding season lasted from September to January. Productivity for the two years combined was 0.71 young fledged per occupied nest. Four out of eight nesting attempts fledged at least one young in 1991 and six out of six in 1992. Nest attendance was greater and prey deliveries were fewer at nests in slash-and-burn clearings than at nests in primary forest throughout the nesting season, although the differences were not statistically significant.
A total of 223 prey deliveries to the nest were identified, including 122 at nests in primary forest and 101 at nests in modified forest. The prey types brought in were 3 (3.1%) rats, 74 (33.2%) Birds, 35 (15.7% chameleons, 65 (29.1%) lizards, 19 (8.5%) snakes, 14 (6.3%) frogs and 9 (4.0%) arthropods. Prey type was not found to be statistically dependent on the habitat type of the nest. The buzzards in this study had broad nesting and dietary requirements, and the species is, therefore, not recommended as an ecological indicator in rain forest habitat.