To make tropical forests more resilient to climate change, we need a coordinated effort to refocus conservation tools at regional and international levels.

The tools include expansion of protected areas, control of fires, and application of REDD policy ('reduced emissions from deforestation and forest degradation'). The latter is intended to protect forest carbon but lacks explicit mechanisms for increasing forest resilience.

These conservation instruments should be focused on two goals. First, they should be deployed to increase the large-scale connectivity of tropical forests, especially across latitudinal and elevational gradients, to facilitate range shifts by tropical species in response to future climate change. Second, they should be coordinated to reduce or halt agricultural expansion in areas of rapid deforestation, especially when such areas are also susceptible to drying, as in the Amazon's arc of deforestation.

The biodiversity benefits of REDD projects and new protected areas would be augmented by strategically locating them to protect connectivity between major ecozones — for example, between the Amazon lowlands and the uplands of the Brazilian and Guiana shields — or to span large-scale moisture gradients, as in the Central Amazonian Conservation Corridor.

In tropical Asia, the strategic use of REDD projects could help to link existing protected areas into large-scale conservation networks in central Borneo, in the forests along the border of Thailand and Myanmar and in the Annamite Mountains of Laos and Vietnam.

Similarly, a more strategic approach is needed for fire-control education and regulation. Although they will always be important in southeast Asia, we can predict when these measures will be crucial: after El Niño events, for example, which can dry out Indonesia's and Malaysia's tropical forests and increase the risk of huge fires.