Mapping the worlds dry tropical biomes

Dry tropical biomes cover 40 – 50% of the tropical regions of the world, but are both highly threatened and poorly mapped. This is challenging because savanna and dry forest are often both structurally overlapping and heterogenous. Biomes are units occupying a large geographical area, at the global scale, that have distinct ecosystem functioning. Past biome mapping has been based upon climate data and expert opinion. However, climate is not suitable for defining biomes existing in the same climatic space, and work based on opinion is inherently subjective. This work makes progress towards an ecologically meaningful map of the global dry tropics, which has many applications, for example in mapping threats facing these ecosystems, accurately mapping land use/cover change, and improving understanding of global relationships between different types of dry tropical vegetation.Here we use vegetation structural, phenological and disturbance parameters, determined using remote sensing, to map dry tropical vegetation, and understand various aspects of their ecology. By working from a base of measuring vegetation rather than a pre-conceived idea of dry tropical biomes, we are able to establish a quantitative and objective approach to biome mapping.The work comprised three sections, throughout we scaled up geographical area to answer specific research questions. First, vegetation in North-East Brazil was mapped, establishing the method in a highly heterogeneous region, and understanding the potential utility of this method for aligning with and operationalizing the new IUCN Global Ecosystem Typology. Eight vegetation structural groups were found in the region, matched to ‘Ecosystem Function Groups’ as identified by the GET. Secondly, we mapped the distribution of vegetation types in Southern Africa, and compared these structural vegetation types to groups based upon floristic information established using the SEOSAW network of plot data. We found that vegetation structural groups do not directly align to floristic groups, and that dry forest in the region may be more extensive than previously suggested. Thirdly, we produced a global map of the dry tropics, with a particular focus on understanding cross-continental differences and similarities between identified groups.