The use of computer intensive statistical modeling in estimating the variability of marine fouling communities.

Abstract

The variability of the fouling community in Monterey Bay was investigated by suspending 100 mild steel plates in Monterey Harbor. The plates were painted with either a nontoxic control paint or one of three antifouling paints. Following the monthly retrieval of a group of these plates, a census of the fouling organisms was conducted and initial variability estimates determined. These estimates were used as inputs for bootstrap computer simulations of the experiment. The results of the bootstrap simulations were then used to determine an appropriate strategy for sampling the fouling community in Monterey Bay. The results indicate that twenty to thirty plates are required to resolve ambiguities concerning the mean percent cover of a group of plates while many more are required to quantify the variability of the fouling population.