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.