Joint modelling of environmental and species responses to restoration

Restoration outcomes have unexplained variation. As environmental variables can both affect the restoration outcome and be affected by restoration, studying the interplay between environmental and species responses could reveal reasons behind this variation. Here, we present a methodological framework of jointly modelling environmental and species responses to restoration and demonstrate it by studying the effect of water table (WT) on vegetation in restored forestry-drained boreal peatlands. Restoration was consistently successful in raising WT in the higher but not in the lower productivity peatland types. Jointly modeling WT and vegetation produced the highest accuracy for predicting vegetation response, highlighting the need to monitor relevant environmental factors in restoration research. The study demonstrates how the framework can be used to explain differences in restoration outcomes, even with sparse environmental data. The framework allows predicting restoration outcomes in different scenarios of environmental variables, facilitating the development of restoration methods with predictable outcomes.