Community analysis and flatfish: diagnostic patterns, processes, and inference

Abstract

Traditional methods of analysis of community ordinate and/or cluster survey data. The patterns extracted are the basis for inferences about habitat preferences of species and interactions among them. Following the logic of those inferences we develop a set of hierarchical simulations to investigate what patterns are diagnostic of the biological processes of habitat selection and interspecific competition (or facilitation). The simulations also include fishing on the simulated populations. The simulations illustrate that preferences of a pair of species along a habitat gradient produce characteristic patterns when species abundances are plotted on axes of abundance of each species. Directed fishing on one of the species eliminates much of the evidence of the pattern. Adding competitive interactions to habitat preferences creates additional diagnostic patterns in the simulated data. Fishing again eliminates much of the evidence of these patterns. Data from multispecies trawl surveys which capture flatfish produce plots that are consistent with the simulation results, although frequent zero catches for one species in a pair produces patterns not included in the simulations. With regard to explaining the distribution and abundance of flatfish in survey data, any evidence of competitive interactions and much of the evidence of differential habitat selection may be removed by the effects of directed fishing. Further simulations with more complex scenarios of fishing, habitat selection, and competition are warranted. However, gains in realism are likely to be offset by losses in clarity.