Anticipated climate and land-cover changes reveal refuge areas for Borneo's orang-utans

Habitat loss and climate change pose a double jeopardy for many threatened taxa, making the identification of optimal

habitat for the future a conservation priority. Using a case study of the endangered Bornean orang-utan, we identify

environmental refuges by integrating bioclimatic models with projected deforestation and oil-palm agriculture

suitability from the 1950s to 2080s. We coupled a maximum entropy algorithm with information on habitat needs to

predict suitable habitat for the present day and 1950s. We then projected to the 2020s, 2050s and 2080s in models

incorporating only land-cover change, climate change or both processes combined. For future climate, we incorporated

projections from four model and emission scenario combinations. For future land cover, we developed spatial

deforestation predictions from 10 years of satellite data. Refuges were delineated as suitable forested habitats identified

by all models that were also unsuitable for oil palm – a major threat to tropical biodiversity. Our analyses indicate

that in 2010 up to 260 000 km2 of Borneo was suitable habitat within the core orang-utan range; an 18–24%

reduction since the 1950s. Land-cover models predicted further decline of 15–30% by the 2080s. Although habitat

extent under future climate conditions varied among projections, there was majority consensus, particularly in northeastern

and western regions. Across projections habitat loss due to climate change alone averaged 63% by 2080, but

74% when also considering land-cover change. Refuge areas amounted to 2000–42 000 km2 depending on thresholds

used, with 900–17 000 km2 outside the current species range. We demonstrate that efforts to halt deforestation could

mediate some orang-utan habitat loss, but further decline of the most suitable areas is to be expected given projected

changes to climate. Protected refuge areas could therefore become increasingly important for ongoing translocation

efforts. We present an approach to help identify such areas for highly threatened species given environmental

changes expected this century.