A GIS-based plant prediction model for wetland ecosystems

An existing non-spatial model for the prediction of response of wetland plant species on ecological factors has been transformed into a GIS-based prediction model which produces spatial output at the landscape scale. The input, spatial patterns of the ecological factors, were constructed with geostatistical spatial interpolation (kriging). With this GIS-based model the spatial patterns of presence and absence of 78 wetland plant species are predicted for an area with wetlands in the Netherlands of approximately 500 square kilometers. The GIS-based model has been validated, and the estimated uncertainty of the input has been propagated through the model. At the species level the output shows spatially coherent and non-random patterns. The validation is affected by the propagation of input errors through the model. The number of valid predictions declines approximately 10-20% when 95% confidence intervals are used in the validation. This study shows that it is feasible to use a geostatistical interpolation method to construct spatial patterns of ecological factors on a landscape scale and to use these patterns as input for a GIS-based prediction model. The added uncertainty on the input values however, affects the number of valid predictions of the model.