Diverging land-use projections cause large variability in their impacts on ecosystems

Team members Anita Bayer, Almut Arneth and Peter Anthoni published a study in Earth System Dynamics (https://esd.copernicus.org/articles/12/327/2021/) demonstrating the large variability in ecosystem service indicators caused by diverging future land-use scenarios. We evaluated 16 projections of future land use until 2040 that reflected different assumptions on socio-economic demands and modeling protocols. The LULC projections were input to a state of the art dynamic global vegetation model, to simulate their effect on selected ecosystem service indicators related to ecosystem productivity and carbon sequestration potential, agricultural production and the water cycle. We found that although a common trend for agricultural expansion exists across the scenarios, where and how particular land-use changes are realized differs widely across land-use models and scenarios and is very much triggered by model-specific considerations and implementations. As a result, some scenarios show questionable and possibly unrealistic features in their LULC allocations and a number of examples for these is given in the paper. Across the diverging LULC projections we identified positive global trends of net primary productivity (+10.2% ± 1.4%), vegetation carbon (+9.2% ± 4.1%), crop production (+31.2% ± 12.2%) and water runoff (+9.3% ± 1.7%), and a negative trend of soil and litter carbon stocks (-0.5% ± 0.4%). The variability in ecosystem service indicators across scenarios was especially high for vegetation carbon stocks and crop production and regionally it was highest in tropical forest regions. Our results emphasize that information on future changes in ecosystem functioning and the related ecosystem service indicators should be seen in light of the variability originating from diverging projections of LULC. This is necessary to allow for adequate policy support towards sustainable transformations.