Benthic Habitats and the Effects of Fishing

Spatial and Temporal Distributions of Bottom Trawling off Alaska: Consideration of Overlapping Effort When Evaluating the Effects of Fishing on Habitat

Craig S. Rose and Elaina M. Jorgensen

doi: https://doi.org/10.47886/9781888569605.ch120

Abstract. The spatial and temporal distribution of fishing effort is a critical component in assessing the effects of bottom fishing gear on benthic habitats. Fishing effort influences habitat unevenly in space and time, creating areas swept once or several times interspersed with areas unaffected by fishing gear. When initial sweeps by fishing gear damage or remove habitat features, subsequent overlapping sweeps may produce less additional damage or removal effects than comparable fishing would produce on unaffected habitat. Therefore, assuming fishing effort is uniform, with minimal overlap, may produce errors in estimating the area and effects of fishing on habitat. The distribution of bottom trawl effort off Alaska from 1997 to 2001 was used to examine changes in the estimated distribution and intensity of bottom trawl sweeps due to (1) the size of blocks used to accumulate data and (2) the assumptions used to model the distribution of fishing within blocks. The use of larger (20 × 20 km and 60 × 60 km) analytical blocks underestimated the amount of effort that overlapped previous tows by averaging effort from high-intensity locations over larger areas. Even with the smallest block size (5 × 5 km), use of statistical distributions that did not allow variation within blocks of fishing intensity (uniform) or the probability of being swept (Poisson) underestimated overlap relative to a distribution which modeled the contagious nature of fishing effort (negative binomial). Comparison of these models with the observed distribution of the 2001 fishery for Atka mackerel Pleurogrammus monopterygius indicated even more overlap in the fishery than even a highly contagious (N = 2) negative binomial distribution could emulate. Fishing distributions with reduced overlap overestimated habitat reductions when applied to a simple model of fishing effects and also overestimated the total area subject to being swept by gear and the average frequency of sweeping.