ANIMAL DISPERSAL PATTERNS - A REASSESSMENT OF SIMPLE MATHEMATICAL-MODELS

Attempts to fit simple mathematical models-of animal movement to obser ved distributions of movements have been extremely successful. However , the extent to which these results were influenced by use of potentia lly distance-weighted sampling methods has not been evaluated. Of the 15 field studies used to test model fits, 13 used methods that sample unevenly over distance. We use a simple Monte-Carlo simulation model t o evaluate the effects of unequal sampling over distance on results ob tained from simple mathematical models of animal movement. For three d ata sets that provide detailed maps of observation sites, the effects of unequal sampling are profound. Prior to sampling, data simulated by our model follow a uniform distribution, but a geometric model adequa tely fits the ''sampled'' simulated data. Following corrections to the field data to mitigate the effects of distance-weighted sampling, in only one of five studies do simple mathematical models adequately fit the observed distribution of movements.