A statistical hypothesis test for a difference between the spatial distributions of two populations is presented. The test is based upon a generalization of the two—sample Cramer—von Mises test for a difference between two univariate probability distributions. It is designed to be sensitive to differences in the way the populations are distributed across the study area but insensitive to a difference in abundance between the two populations. The procedure is nonparametric and uses the methodology of a randomization test to determine the level of significance of the test statistic.

No abstract

A Before–After Control–Impact (BACI) experiment was conducted to investigate the effects of a commercial bottom trawl on benthic invertebrates in a sandy and previously untrawled area of the eastern Bering Sea. Six pairs of experimental and control corridors were sampled with a research trawl before and after four consecutive tows with the commercial otter trawl. A major storm event occurred during the experiment, and it was possible to differentiate its effect from that of the trawling using the BACI model. Species composition changed very little; Asterias amurensis and Paralithodes camtschaticus comprised over 80% of the total invertebrate biomass (kg ha−1) during each year of the study. In general, the commercial trawl did not significantly affect the biomass of the benthic invertebrate populations. The trawling effect after 4–14 d was statistically significant in three of the 24 taxa that were analysed, which was, as expected, because of nothing more than random variation with α = 0.10. Biomass immediately after the trawling disturbance was lower for 15 of the taxa and higher for the other nine, with a median change of −14.2%. Similarly, the effect of trawling on invertebrate biomass after one year was not statistically significant for any of the taxonomic groups (p ≥ 0.23), indicating no evidence of a delayed response to the commercial-trawl disturbance. Further analysis suggests that storms have an overall greater effect on the benthos than do bottom trawls at this location. Both the numbers of taxa significantly affected by trawling and the storm (3 vs. 12), as well as the median sizes of these effects −14.2% vs. −22.0%), were greater for the storm event. Results from this study are combined with those from a related investigation of chronic trawling effects to propose an adaptive management strategy for the study region, including rotating area closures to mitigate for temporary trawling effects.