Search

Since 1984, annual bottom trawl surveys of the west coast (California–Washington) upper continental slope (WCUCS) have provided information on the abundance, distribution, and biological characteristics of groundfish resources. Slope species of the deep-water complex (DWC) are of particular importance and include Dover sole, Microstomus pacificus; sablefish, Anoplopoma fimbria; shortspine thornyhead, Sebastolobus alascanus; and longspine thornyhead, S. altivelis. In the fall of 1994, we conducted an experimental gear research cruise in lieu of our normal survey because of concerns about the performance of the survey trawl. The experiment was conducted on a soft mud bottom at depths of 460–490 m off the central Oregon coast. Treatments included different combinations of door-bridle rigging, groundgear weight, and scope length. The experimental design was a 2 ´ 2 ´ 2 factorial within a randomized complete-block. Analysis of variance was used to examine the effects of gear modifications on the engineering performance of the trawl (i.e. trawl dimensions, variation in trawl dimensions, and door attitude) and to determine if catch rates in terms of weight and number of DWC species and invertebrates were affected by the gear modifications. Trawl performance was highly variable for the historically used standard trawl configuration. Improvements were observed with the addition of either a 2-bridle door or lighter ground gear. Changes in scope length had relatively little effect on trawl performance. The interaction of door bridle and ground gear weight had the most effect on trawl performance. In spite of the standard trawl’s erratic performance, catch rates of all four DWC species and invertebrates were not significantly different than the 2-bridle/heavy combination, which did the best in terms of engineering performance. The most important factor affecting DWC catch rates was ground gear. Scope length and the type of door bridle had little effect on DWC catch rates. Subsequent revisions to survey gear and towing protocol and their impact on the continuity of the slope survey time series are discussed.