EMPIRICISM AND THE PREDICTION OF PRIMARY PRODUCTION AT THE MESOSCALE - A SAVANNA EXAMPLE

The role of empiricism in predicting mean annual primary production at the mesoscale is demonstrated through development of a simple model ( RESAP) for application over all land surfaces within a given savanna r egion. The model is necessarily pragmatic and constrained by availabil ity of standardized data on a network scale. The model was applied to a large test area in southern Africa and provides results using a SYMA P routine for representing spatially continuous information. The model discriminates between woody and herbaceous plant components and also assesses the effects of anthropogenic disturbance on plant productivit y. Productivity of natural vegetation was found to be remarkably equit ably distributed across the diverse conditions which suggests various compensatory production mechanisms at the mesoscale. Model output grap hically indicated that intensive agriculture successfully converts tot al productivity into both a more usable and more accessible form at th e apparent cost of long-evolved total production levels.