Modelling groundwater-dependent vegetation patterns using ensemble learning

摘要

Vegetation patterns arise from the interplay between intraspecific and interspecificbiotic interactions and from different abiotic constraints and interacting drivingforces and distributions. In this study, we constructed an ensemble learning model that,based on spatially distributed environmental variables, could model vegetationpatterns at the local scale. The study site was an alluvial floodplain with markedhydrologic gradients on which different vegetation types developed. The modelwas evaluated on accuracy, and could be concluded to perform well. However, modelaccuracy was remarkably lower for boundary areas between two distinct vegetationtypes. Subsequent application of the model on a spatially independent data setshowed a poor performance that could be linked with the niche concept to concludethat an empirical distribution model, which has been constructed on local observations,is incapable to be applied beyond these boundaries.