What ecologists should know before using land use/cover change projections for biodiversity and ecosystem service assessments

Scenarios of biodiversity and ecosystem services (BES) are key for decision-makers to understand the consequences of future environmental change on BES. Though a major driver of terrestrial biodiversity loss, land use and land cover changes (LUCC) have been largely overlooked in previous BES assessments. But ecologists lack practical guidance for the general use of LUCC projections. We review the practices in use in LUCC-driven BES assessments and summarize the questions ecologists should address before using LUCC projections. LUCC-driven BES scenarios rely on a substantial set of different socioeconomic storylines (> 200 for 166 papers). Studies explore different futures, but generally concentrate on projections obtained from a single LUCC model. The rationale regarding time horizon, spatial resolution, or the set of storylines used is rarely made explicit. This huge heterogeneity and low transparency regarding the what, why, and how of using LUCC projections for the study of BES futures could discourage researchers from engaging in the design of such biodiversity scenarios. Our results call on those using LUCC projections to more systematically report on the choices they make when designing LUCC-based BES scenarios (e.g. time horizon, spatial and thematic resolutions, scope of contrasted futures). Beyond the improvement of reliability, reproducibility, and comparability of these scenarios, this could also greatly benefit others wanting to use the same LUCC projections, and help land use modellers better meet the needs of their intended audiences. The uncertainties in LUCC-driven BES futures should also be explored more comprehensively, including different socioeconomic storylines and different LUCC models, as recommended in studies dealing with climate-driven BES futures.