Researchers have used remote sensing data to map out the functional diversity of forests in the Peruvian Andes and Amazon basin, a technique that has revealed hotspots for conservation. The high-resolution views of tropical forest traits they provide may help inform conservation-based decision-making that has otherwise been difficult to date, hampered in part by the sparseness of ground data on biological diversity. Recently, advances in airborne laser-guided imaging spectroscopy have offered a means to gather sufficient data on tree diversity and ecosystem types remotely. This approach can shed light on forest traits such as photosynthesis rates and foliar phosphorus and calcium levels, for example, which are indicative of species composition. Here, Gregory Asner and colleagues combined such spectroscopy data with computational machine learning to generate diversity maps for a large portion of the forests of the Peruvian Andes and Amazon basin. In total they identified 36 forest functional classes that could be further subcategorized into six forest functional groups (FFGs). Next they used government land allocation data to determine the degree to which each group is affected by petroleum oil exploration and logging. Between 32 and 46% of each mapped FFG is currently protected, of which about two-thirds are under government control and one-third is on indigenous lands, the tiemans report. They also find that 6.4 million hectares of lower foothill forests and high Andean forests - mountain forests containing distinct functional compositions relative to forests in the western Amazon - are highly threatened by unofficial uses, thus offering important opportunities for protection. This work is highlighted in a Perspective by Valerie Kapos.