Approaches to Improving Groundfish Survey Abundance Estimates by Controlling the Variability of Survey Gear Geometry and Performance

Abstract

Options for using trawl mensuration data to improve survey abundance estimates are discussed, with examples from Scotian Shelf groundfish surveys. Once the survey performance parameters of individual trawls are known, the next step should be to control the variability of parameters which can be controlled. rather than to standardize the catch using catch/parameter relationships which are themselves highly variable or subject to dubious assumptions. It is possible to interactively change warp length usinq spread sensor information to achieve constant door spread, just as distance travelled during a tow is now controlled by interactively monitoring ship speed with Doppler logs. Abundance indices from Scotian Shelf surveys are subject to significant bias from systematic variations in gear spread, confirming the importance of routine trawl mensuration during standard survey sets, and the need for corrective action.

The effect of improved footrope contact measurements on trawl area swept estimates

“…A major disadvantage of adjusting the catches by area swept is that it inappropriately assumes a linear relationship between the measured parameter (distance fished) and fish catch, and that the measured distance fished is the effective distance involved in the catching process (Koeller, 1991). Because of the high haul-tohaul catch variability inherent in experimental trawling there is little quantitative evidence for such a relationship.…”

The effect of net spread on the capture efficiency of a demersal survey trawl used in the eastern Bering Sea

“…Similarly, for species, such as yellowfin sole, where males are distributed in shallower waters than the females, the depth variation in trawl efficiency will lead to a bias in the estimates of sex ratio (Nichol, 1998). Koeller (1991) argued that it is possible to maintain a constant door spread throughout a survey as well as throughout a tow by systematically varying the amount of warp paid out at each depth. From regressions of warp-depth (scope) ratio on door spread, he determined the scope ratios required to achieve constant door spreads at different depths.…”

“…For species having a vertical distribution extending above the headrope, capture efficiency may decrease with increasing net spread because headrope height may decrease (Koeller, 1991;Rose and Nunnallee, 1998). For species that remain in direct contact with the bottom, capture efficiency may decrease with increasing net spread because the footrope distance from the bottom may increase (Rose and Nunnallee, 1998).…”

“…Published by Elsevier B.V. doi:10.1016/j.fishres.2005.04.007 constant swept area leads to biased abundance estimates (Godø and Engås, 1989;Rose and Walters, 1990;Koeller, 1991). Rose and Walters (1990) showed that net spread can be highly variable among trawl tows and that the variation between tows was primarily related to change in the length of towing cable (scope) required to fish at different depths.…”

“…Godø and Engås (1989) also found that net spread varied with depth but also was influenced by bottom type. After correcting swept area abundance estimates for the variation in net spread, biases in the Svalbard and Barents Sea cod survey ranged between 9 and 37%, depending on the choice of reference net spread (Godø and Engås, 1989); biases in the Bering Sea survey ranged from 7 to 15%, depending on species (Rose and Walters, 1990); and biases in the Scotian Shelf groundfish survey ranged from 10 to 15% for haddock (Koeller, 1991). Because of such variability in net spread, the eastern Bering Sea shelf bottom trawl survey now is based on catch per swept area, where swept area is calculated as the product of net spread and tow length measured during each haul.…”