Patterns of twenty‐first century forest loss across a global network of important sites for biodiversity

The contribution that remotely sensed satellite imagery could make to biodiversity monitoring is widely recognized, but remains largely unfulfilled at regional and global scales. Here, we use a publicly accessible global dataset to quantify forest loss during 2000–2012 across the world's 7279 Important Bird and Biodiversity Areas (IBAs) that support forest bird species, and examine correlates of forest loss. Forest covered around 2 912 000 km2 in these IBAs in 2000, but had decreased by around 73 000 km2 by 2012, a 2.52% loss across all IBAs. The median loss within IBAs was 0.87% during this period, although high losses on some IBAs skewed the mean to 2.59%. Rates of forest loss in IBAs were highest in South America and South East Asia. The rate of forest loss increased notably between 2003 and 2007. In addition to the effect of year, the best fit model of forest loss contained coverage of IBAs by protected areas, altitude, slope, GDP and human population density. The parameter estimates of a best fit model indicated forest loss decrease with increased protected area coverage, increased IBA altitude, increased IBA steepness and increased IBA human population. Forest loss decreased with per capita national GDP. The model explained only a small amount of variation in forest loss and as such is inappropriate to make predictions about where future loss will occur. We provide open access code to undertake online analysis of future forest cover change data, aiding the remote monitoring of forest cover at any suite of priority sites.