Resource Assessment & Conservation Engineering (RACE) Division

Groundfish Assessment Program

Sampling Efficiency Estimated for Poly Nor’eastern Trawl Used on Gulf of Alaska and Aleutian Islands Bottom Trawl Surveys

Animal density can be estimated from bottom trawl catch-per-swept-area data provided there is knowledge of the whole-gear sampling efficiency or the proportion of animals that are captured within the area spanned by the trawl doors. One approach to the estimation of whole-gear sampling efficiency is to consider it as a function of the efficiencies of herding, mesh retention, and other components of the trawl catching process, because such components are often more tractable to field experimentation and estimation.

For flatfish, which are unlikely to pass over the trawl headrope or through the mesh and are herded only by the section of the lower bridle that is sufficiently close to the bottom to elicit a behavioral response, a mathematical model of trawl efficiency can be formulated as

where W_n and W_d are the trawl net and door spread, k_n is the net efficiency or proportion of fish retained at the trawl footrope, W_on is the width of the area swept by the bridles and h is the fraction of the flatfish within W_on that are herded into the net path.

Figure 1.  Efficiency of the 83-112 bottom trawl for arrowtooth flounder, flathead sole, rex sole and Dover sole as a function of total body length in cm. The dashed lines about each curve represent the 95% confidence intervals.

Evaluating this model requires data from three distinct experiments. In this study, k_n was estimated from data obtained from a net efficiency experiment which consisted of attaching an auxiliary bag under the trawl to capture fish escaping beneath the footrope, h was estimated from data obtained from a herding experiment which consisted of repeatedly conducting trawl hauls in which W_d was varied by varying the length of the bridles, and W_on was estimated from data obtained on a bridle measurement experiment using bottom contact sensors to measure the off-bottom distance along the lower bridle.

These experiments were directed at four flatfish species (flathead sole (Hippoglossoides elassodon), rex sole (Glyptocephalus zachirus), Dover sole (Microstomus pacificus) and arrowtooth flounder (Atheresthes stomias)) using the Poly Nor’eastern trawl, the standard trawl used by the AFSC on its bottom trawl surveys of the Gulf of Alaska and the Aleutian Islands.

Estimates of the herding coefficient (h) averaged 0.55 for the three sole species (rex sole=0.53; Dover sole=0.58; flathead sole=0.55), all of which were higher than arrowtooth flounder (0.391). Thus, roughly 40%-50% of the flatfish encountering the lower bridle were ultimately herded into the path of the net. Estimates of the net efficiency (k_n) for arrowtooth flounder, flathead sole and rex sole increase with fish length and reach maximum values between 0.85 and 0.95. Estimates of k_n for Dover sole, however, decrease with increasing size both because small fish were not sampled and because this species apparently becomes more adept at escaping under the footrope with increasing size.

Trawl efficiency (E) estimates for arrowtooth flounder, flathead sole, and rex sole increased with increasing fish length and reached maxima of 0.45, 0.42 and 0.43 (Fig. 1), indicating that slightly more than 40% of the largest individuals that passed between the doors of the trawl were ultimately caught. In contrast, the efficiency estimates for Dover sole were considerably lower over the sampled size range, and monotonically decreased with increasing length.

Since bottom trawl surveys conducted by the AFSC calculate swept area in terms of wing spread rather than door spread, as is the convention for most International Council for the Exploration of the Sea (ICES) sponsored surveys, to be useful in stock assessment models the above values must be multiplied by the quotient of the door spread and net spread, which for the Poly Nor’eastern trawl is approximately equal to 3 (47.8 m/16.1 m). Thus maximum efficiency for these species is slightly greater than 1.2.

Details of this study are available in a draft manuscript by Dave Somerton, Peter Munro, and Ken Weinberg titled “Whole-gear efficiency of a benthic survey trawl for flatfish.”

By David Somerton


>>>continued

AFSC Quarterly
Research Reports July-Sept 2005