128 



Fishery Bulletin 93(1). 1995 



upon request (discards); and 3) survey samples col- 

 lected at sea from September to November 1992 on 

 the National Marine Fisheries Service (NMFS) tri- 

 ennial bottom trawl and slope surveys off Washing- 

 ton State and Vancouver Island, British Columbia 

 (Table 1). All of the commercial trawlers sampled 

 used larger codend mesh than did the trawl survey 

 (11.4 cm and 8.9 cm, respectively) and reported 

 arrowtooth flounder catch from grounds off north- 

 west Washington (Fig. 1). Commercial trawlers com- 

 bine catches from different tows so trawl landings 

 and landed discards were sampled on a per trip ba- 

 sis. No samples were obtained in January or Febru- 

 ary 1992. Up to 100 fish were selected at random for 

 each sample. If fewer than 100 fish were present, all 

 fish were sampled. Each fish was measured to the 

 nearest centimeter fork length (FL), sexed, and as- 

 signed a maturity stage based on the macroscopic 

 appearance of the gonads (Table 2). A gonosomatic 

 index (GSI), calculated as ovary weight divided by 

 somatic weight (body weight with ovaries removed 

 and stomach empty), was used to track seasonal 

 changes in ovarian condition. In each sample, the 

 first 10 females encountered at each macroscopic 

 maturity stage were sampled for GSI. Whole ovaries 

 were removed and weighed to the nearest gram, and 



Figure 1 



Reported arrowtooth flounder, Atheresthes stomias, catch 

 area for sampled commercial bottom trawl trips, and loca- 

 tions of survey tows sampled for maturity, with fathom 

 contours (1 fm=1.83 m). 



somatic weight was recorded to the nearest gram. 

 Average monthly GSI's were calculated by maturity 

 stage. 



Fish at any maturity stage other than "immature" 

 were defined as mature. Length at maturity was es- 

 timated for each sex by calculating the fraction ma- 

 ture in each 1-cm interval and by fitting a logistic 

 model 



n- l 



1 + e 



a+bL) 



where P L = fraction mature at length L, and a and b 

 are constants (Gunderson et al., 1980). The predicted 

 size at which 50% of the fish are mature is 



^50% 



-a 

 ~b~ 



Confidence intervals for L 50% were calculated by us- 

 ing the delta method (Seber, 1982). To see if time of 

 collection affected estimates of L 50% , logistic equa- 

 tions were fit to the female length-maturity data that 

 were grouped into the following periods for 1991 and 

 1992 separately: September-December, January- 

 April, and May-August, corresponding to the 

 prespawning, spawning, and postspawning seasons 

 as determined below. 



Ovarian tissue samples were collected from five 

 tows on board the commercial trawler FV Larkin in 

 late December 1991, during normal trawl operations 

 off northwest Washington, to compare microscopic 

 and macroscopic staging and provide additional in- 

 formation on arrowtooth flounder reproduction. Im- 

 mediately after the net was retrieved, arrowtooth 

 flounder were sorted by the crew into discard and 

 market categories and were held alive until they 

 could be sampled. Discard and market categories 

 were maintained for maturity sampling but combined 

 for histology. Length, sex, and maturity stage were 

 recorded, and for the first 10 females encountered at 

 each maturity stage an approximately 3-mm section 

 of ovarian tissue was cut from the anterior third of 

 the blind-side (left) ovary, placed in a tissue cassette, 

 and fixed in Davidson's fixative (Mahoney, 1973). 

 Tissue samples cut from the anterior third of the 

 ovary were assumed to be representative of the 

 arrowtooth flounder ovary as a whole. Weights for 

 GSI were not taken owing to rough weather. 



After fixation, tissues were embedded in paraffin, 

 sectioned at 6 /im, and stained with hematoxylin and 

 eosin. Tissue slides were examined under a compound 

 microscope and measurements were taken with a 

 calibrated ocular micrometer. A classification scheme 

 of five stages of oocyte development, two stages of 

 atresia, and a final stage of postovulatory follicles 

 was adopted (Table 3), with terminology proposed by 



