Rejoinder on: Comparing and selecting spatial predictors using local criteria