Search

Large avian frugivores are important in ecosystem function, but are seriously threatened across the tropics. To conserve them we must understand their habitat needs and the effects of improved forest management on individual species and the community as a whole. We recorded the presence/absence of 18 parrot, pigeon and hornbill species along nearly 500 km of transects at 24 sites in Luzon, Philippines, and used logistic GLMMs to identify bird-habitat associations based on topographical, forest structure and floristic data taken at 1227 habitat plots. We then searched for more complex relationships and thresholds in species responses along forest quality/restoration gradients using GAMMs. Frugivore species richness was highest in forest with large-girthed trees, although some small-scale agricultural disturbance was tolerated or even favoured. Importantly, richness was highest in forests on flat ground, areas which are usually the first to be converted to agriculture. Individual species were positively associated with large trees but responses to floristic gradients were more variable. Very few species had complex relationships with forest quality; for the great majority, the probability of occurrence increased linearly along the forest quality/restoration gradient. While the precise benefits in terms of seed dispersal, and costs of management, at different points along the quality/restoration gradient are likely to be themselves complex, avian frugivores benefit proportionately from step improvements right along the gradient. Thus, any actions to improve forest quality on Luzon, from reforesting the most degraded lands to preventing degradation of relatively healthy forests, are likely to benefit frugivores.

The lack of scientific baseline information hinders appropriate design and management of protected areas. To illustrate the value of science to management, we consider five scenarios for the 202.0 km² Puerto Princesa Subterranean River National Park, Philippines: (1) closure to human activities, (2) and (3) two levels of increase in unplanned human activities, (4) creation of a forest corridor and (5) additional allocation of land for permanent or shifting agriculture. We then use habitat-specific bird density estimates to simulate the net effect of each scenario on 18 focal bird populations. Closure has significant benefits-populations of five species are predicted to increase by >50 % and nine by >25 %, but two secondary forest flycatchers, including the endemic and 'Vulnerable' Palawan flycatcher, decline dramatically, while the creation of a 4.0 km² forest corridor yields average increases across species of 2 ± 4 % (SD). In contrast, heavier unplanned park usage produces declines in all but a few species, while the negative effects of an extra 2.0 km² of shifting cultivation are 3-5 times higher than for a similar area of permanent agriculture and affect species whose densities are highest in primary habitats. Relatively small changes within the park, especially those associated with agricultural expansion, has serious predicted implications for local bird populations. Our models do not take into account the full complexities of bird ecology at a site, but they do provide park managers with an evidence base from which to make better decisions relating to biodiversity conservation obligations which their parks are intended to meet.