278 



Abstract — Whole-gear efficiency (the 

 proportion of fish passing between 

 the otter doors of a bottom trawl 

 that are subsequently captured) was 

 estimated from data collected during 

 experiments to measure the herding 

 efficiency of bridles and doors, the 

 capture efficiency of the net, and 

 the length of the bridles sufficiently 

 close to the seafloor to elicit a herd- 

 ing response. The experiments were 

 focused on four species of flatfish: 

 arrowtooth flounder iAtheresthes sto- 

 mias), flathead sole {Hippogtossoides 

 elassodon), rex sole (Glyptocephalus 

 zachirus), and Dover sole (Microsto- 

 mus pacificus). Whole-gear efficiency 

 varied with fish length and reached 

 maximum values between 40% and 

 50% for arrowtooth flounder, flathead 

 sole, and rex sole. For Dover sole, how- 

 ever, whole-gear efficiency declined 

 from a maximum of 33% over the 

 length range sampled. Such efficiency 

 estimates can be used to determine 

 catchability, which, in turn, can be 

 used to improve the accuracy of stock 

 assessment models when the time 

 series of a survey is short. 



Whole-gear efficiency of a 

 benthic survey trawl for flatfish 



David A. Somerton (contact author) 



Peter T. Munro 



Kenneth L. Weinberg 



National Marine Fisheries Service, NOAA 



Alaska Fisheries Science Center 



7600 Sand Point Way NE 



Seattle, Washington 98115 



Email address for D. A. Somerton: davld.somerlon@noaa.gov 



Manuscript submitted 1 August 2006 

 to the Scientific Editor's Office. 



Manuscript approved for publication 

 23 October 2006. 



Fish. Bull. 105:278-291 12007). 



Fish density can be estimated from 

 bottom trawl catch-per-swept-area 

 data if there is knowledge of the 

 whole-gear efficiency (the proportion 

 of fish that are captured within the 

 area spanned by the trawl doors). One 

 approach to the estimation of whole- 

 gear efficiency is to consider it as a 

 function of three separate and under- 

 lying trawling processes: vertical and 

 horizontal herding offish, retention of 

 fish by the net, and escapement offish 

 beneath the trawl footrope, which are 

 often more tractable to field experi- 

 mentation and estimation. Perhaps 

 the earliest example of this approach 

 was the development of a mathemati- 

 cal model by Dickson (1993a) for the 

 efficiency of trawl gear in capturing 

 fish and the application of this model 

 for capturing Atlantic cod (Gadus 

 inorhua) and haddock (Melanogram- 

 mus aeglefmus; Dickson, 1993b) with 

 the use of experimental data on herd- 

 ing (Engas and Godo, 1989a) and on 

 escapement under the footrope (Engas 

 and God0, 1989b). Somerton and 

 Munro (2001) proposed a modifica- 

 tion to Dickson's (1993a) model for 

 application to flatfish to account for 

 the observation that flatfish herding 

 is restricted to the length of the lower 

 bridle that is sufficiently close to the 

 bottom to elicit a behavioral response 

 (Main and Sangster, 1981b). A vari- 

 ant of this model was then applied to 

 seven species of North Pacific flatfish 

 to estimate the efficiency of the cap- 

 ture process that was due to herd- 

 ing (Somerton and Munro, 2001). 

 Although this application was followed 



by experiments to estimate escape- 

 ment under the footrope for some of 

 the same flatfish species (Munro and 

 Somerton, 2002), the flatfish efficiency 

 model was never used to combine the 

 herding and escapement estimates 

 and thereby produce an estimate of 

 whole-gear efficiency. In this article, 

 we again use the flatfish trawl effi- 

 ciency model, extended by developing 

 an estimator for the variance of effi- 

 ciency, and apply the model to new 

 experimental data for four flatfish spe- 

 cies (flathead sole [Hippoglossoides 

 elassodon], rex sole [Glyptocephalus 

 zachirus], Dover sole [Microstomus 

 pacificus], and arrowtooth flounder 

 [Atheresthes stomias]) to estimate 

 whole-gear efficiency for the Poly 

 Nor'eastern trawl, which is used by 

 the Alaska Fisheries Science Center 

 (AFSC) on its bottom trawl surveys of 

 the Gulf of Alaska and the Aleutian 

 Islands region. 



Materials and methods 



The Poly Nor'eastern trawl 



The Poly Nor'eastern trawl, pictured 

 in Figure 1 and detailed in the Appen- 

 dix, has the following basic features: 

 the net has a four-seam design and 

 has a 27.2-m headrope and a 36.5-m 

 footrope equipped with 36-cm diam- 

 eter bobbins to allow operation on 

 moderately rocky terrain. The trawl 

 doors are "V" style measuring 1.8 m 

 by 2.7 m and weighing 816 kg each. 

 Tailchains constructed of two 3-m 



