FISHERY BULLETIN: VOL. 79, NO. 4 



morhua, and the dab, Limanda limanda). Levings 

 (1974) investigated seasonal changes in feeding 

 and particle selection by winter flounder, 

 Pseudopleuronectes americanus, and Moore and 

 Moore (1976) studied various factors influencing 

 the selection of food by the flounder Platichthys 

 flesus. In North Pacific demersal communities, 

 however, even the qualitative aspects of selectivity 

 and the role of selectivity in trophic dynamics have 

 yet to be estimated. 



The objectives of this study are to: 1) describe the 

 food habits of the Dover sole in an area of an active 

 commercial fishery for this species off Oregon, 2) 

 determine if the species is a selective feeder, and 3) 

 determine how feeding habits are related to loca- 

 tion offish capture, size offish, and size and depth 

 of prey in the sediment. 



METHODS AND MATERIALS 



Samples of demersal fishes and benthic inver- 

 tebrates were taken on the central continental 

 shelf off Oregon in locations of high Dover sole 

 abundance (Demory et al. footnote 3; Tyler"*) (Fig- 

 ure 1). Station SG29 (lat. 44°05.0' N, long. 

 124°35.0' W) was located in Heceta Swale, the 

 region east of Heceta Bank. The mean sampling 

 depth was 119 m; the sediment is silty sand 

 (Maloney 1965). Station SGIO (lat. 43°49.3' N, 

 long. 124°50.0' W) was located south of Heceta 

 Bank. The mean sampling depth was 426 m. The 

 sediment is glauconitic sand; however, sediment 

 distribution is patchy in this area (Bertrand 1971). 

 Samples were taken over a limited area and time 

 interval to reduce large-scale spatial and temporal 

 variability (SG29: 41.6 km^ 44 h; SGIO: 34.21 km^, 

 27 h; day and night 20-24 June 1976). 



Benthic infauna was sampled at each station by 

 two box corers: a 0.1 m^ Bouma box corer (Bouma 

 1969) and a modified 0.25 m^ Hessler-USNEL box 

 corer (Hessler and Jumars 1974). The Hessler- 

 USNEL box corer is designed to reduce pressure 

 waves which often blow away small surface inver- 

 tebrates before corer impact. Box cores provide the 

 largest, deepest, and least disturbed sample of 

 consistent surface area when compared with other 

 commonly used sediment samplers (Word^). In 



A. V. Tyler, Department of Fisheries and Wildlife, Oregon 

 State University, Corvallis, OR 97331, pers. commun. June 1976. 

 ■'Word, J. Q. 1977. An evaluation of benthic invertebrate 

 sampling devices for investigating feeding habits of fi.sh. In C. 

 A. Simenstad and S. J. Lipovsky (editors), Proc. 1st Pac. NW. 

 Tech. Workshop. Fish food habits studies, p. 43-55. Washington 

 Sea Grant, Seattle. WSG-WO-77-2. 



<45 min after retrieval of the core, core samples 

 were extruded in 0.1 m^ boxes and sectioned at 1 

 cm intervals for the first 10 cm, 2 cm intervals for 

 the next 10 cm, and 4 cm intervals for the remain- 

 der of the core. These sections were then washed 

 onto a 1 mm aperture sieve screen, and washed 

 invertebrate samples were preserved in 4% unbuf- 

 fered formaldehyde. 



Beam trawl tows to sample Dover sole were 

 made in the immediate area of box core sampling. 

 A beam trawl with an effective trawling width of 

 2.72 m (Carey and Heyomoto 1972) and 3.8 cm 

 stretched mesh lined with 1.3 cm mesh netting was 

 towed at 3 kn for 30 min per haul at SG29 and 20 

 min at SGIO. Fish were preserved in S^c (unbuf- 

 fered) formaldehyde as soon as possible after the 

 trawl was brought aboard. The body cavities offish 

 >12 cm were slit to allow rapid formaldehyde 

 penetration into the coelom. 



A total of four 0.1 m^ box cores, eight 0.25 m^ box 

 cores, and 15 successful beam trawl tows were 

 made at SG29. At SGIO, nine 0.1 m^ box cores and 

 10 successful beam trawl tows were made. 



In the laboratory, invertebrates from box cores 

 were transferred to 70% isopropyl alcohol, sorted 

 into major taxa, and identified to species whenever 

 possible. Dover sole were measured (standard 

 length) and stomachs (from esophagus to constric- 

 tion before pyloric caeca) were removed and trans- 

 ferred to 70% isopropyl alcohol. A total of 202 

 stomachs from SG29 and 63 stomachs from SGIO 

 were processed. Stomach contents were sorted into 

 phyla and identified to species whenever possible. 

 Total lengths of polychaetes, aplacophorans, and 

 scaphopods were measured. Gastropod and 

 pelecypod measurements were made along the 

 longest axis and included shells. Crustaceans 

 were measured from base of rostrum to point of 

 flexure of abdomen. Since ophiuroids occurred as 

 pieces, a single measurement of ophiuroid volume 

 per stomach was made. 



Dry weights of prey items were estimated using 

 conversion factors. Shells, tubes, massive paleal 

 setae (in the case of the polychaete Pectinaria 

 californiensis) , and posterior scutes (in the case of 

 the polychaete Sternaspis fossor) were removed 

 before individual items of known length were 

 dried (36 h, 65° C). Items were weighed using an 

 electrobalance or Mettler^ balance. Regression 

 curves were fitted to sets of length-weight points 



"Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



750 



