I have been modelling spatial data using function likfit in package geoR. Now that I have the spatial regression models, I would like to calculate the amount of variance explained by both the trend and the spatial parts of the model. Any advice on how to do this would be much appreciated. Cheers, Ailsa