Most data of ecological and evolutionary interest are located in geographic space. This spatial context of data such as patterns of birds' nests, genetic variation patterns, or tree distributions leads to interesting problems in ecological and evolutionary theory, and, at the same time, requires allowance for these spatial patterns in statistical analyses that involve such data. The latter have largely been ignored to date, and Dr. Sokal's work will contribute to a recognition of their importance among ecologists and evolutionists. Dr. Sokal will examine, by a variety of analytical (mathematical) and computer simulation approaches, the validity of a number of tests that have become popular in recent years. The results of this research should greatly facilitate inferences from patterns to processes in a wide variety of problems in population biology. These procedures have considerable practical application since virtually all environmental problems have a spatial dimension. Thus, tests of the effects of pollutants, such as acid rain, need to take the spatial relationships of the data into consideration before evaluation of those results. This is an approach that has so far rarely been followed in either environmental or epidemiological work.
