360 



Fishery Bulletin 94(2), 1996 



termine whether the unusually high percentage of 

 females in the arrowtooth flounder population, seem- 

 ingly an odd occurrence, was also typical of the other 

 species from the same area caught in the same sur- 

 vey. Additional comparisons could not be made from 

 data gathered in the Aleutian Islands because of the 

 limited number of sex determinations of other flat- 

 fish species. 



The unusually high percentage of females in the 

 arrowtooth flounder population prompted us to ana- 

 lyze the data in different ways to look for trends in 

 percentages of females. The percentage of females 

 was calculated from individual trawl hauls in which 

 the sex was determined from a minimum of 10 fish. 

 Percent-female values for arrowtooth flounder were 

 examined geographically to determine whether there 

 were regions where male arrowtooth flounder were 

 abundant. Percent-female values were also regressed 

 against bottom temperature and depth. By search- 

 ing for trends in percent-female values, we hoped to 

 identify areas where high proportions of arrowtooth 

 flounder males may have occurred but were not 

 sampled or were underrepresented in our survey. 



For fish from the eastern Bering Sea slope, sagittal 

 otoliths were collected from up to three fish per sex per 

 centimeter interval for ageing and were stored in a glyc- 

 erol-thymol solution. Because previous methods for 

 preparing the otoliths for ageing did not clearly reveal 

 the annular rings, researchers at the AFSC developed 

 a new technique ( Scott 4 ). Otoliths were cut along their 

 dorsal-ventral axis through the nucleus and evenly 

 heated in a muffle furnace at 270°C for 5 to 10 min. 

 If the annular rings were not sufficiently darkened, 

 the otoliths were burned in an alcohol flame. When 

 cool, oil was applied to the otolith surface and the 

 cut surface read with a binocular dissecting scope. 



Von Bertalanffy growth curves were calculated for 

 the age and length data by using nonlinear regres- 

 sion to provide estimates of L , K, and / . To deter- 

 mine whether growth differed significantly between 

 sexes or between species, the residual sum of squares 

 (RSS) from a combined model was compared by us- 

 ing an approximate F-test ( Draper and Smith, 1981 ): 



showed that this approximation of the test statistic, 

 which tests for differences between all growth curve 

 parameters simultaneously, can be useful for test- 

 ing von Bertalanffy growth parameters. The test sta- 

 tistic was compared against values from the F -dis- 

 tribution at the «=0.05 level of significance. The von 

 Bertalanffy parameters were then used to provide 

 estimates of length at age for the range of ages found 

 for each species and sex group. 



Results 



Atheresthes spp. were caught in 492 of the 779 trawl 

 hauls. Neither species was caught in the shallowest 

 trawl hauls (<25 m), whereas catch rates for both 

 species were low in the deepest trawl hauls (700 m 

 to 800 m). Arrowtooth flounder were abundant from 

 76 m to 450 m, with mean CPUE ranging from 3.3 to 

 13.6 kg/ha ( Fig. 1 ). Kamchatka flounder were abun- 

 dant from 226 m through 500 m, with mean CPUE 

 ranging from 3.8 to 12.8 kg/ha. Arrowtooth flounder 

 catch rates showed a bimodal distribution in rela- 

 tion to depth, with high catch rates in both shallow 

 (76 m to 225 m) and deep (226 m to 450 m) trawl 

 hauls. Kamchatka flounder catch rates were weakly 

 bimodal, with only low catch rates in shallow waters 

 and all the highest catch rates in deeper waters. 



A comparison of the percent of the total Atheresthes 

 CPUE by species for each depth interval shows that 

 arrowtooth flounder were more abundant in the shal- 

 lower depth intervals (<250 m) (Fig. 2). There was a 

 zone of transition between the two species from 251 

 m to 425 m. Finally, Kamchatka flounder were more 

 abundant than arrowtooth flounder from 426 m to 

 800 m. For trawl hauls in which both species were 

 caught, a logistic regression relating percentage of 

 arrowtooth flounder of the total Atheresthes CPUE 

 against depth (Fig. 3) showed a significant, negative 

 relationship (df=277,F=98.8,P<0.001,« 2 =0.26). The 

 model was 



3.47-0.009(Deptfi) = ln(p/l-p), 



[ RSS 



TOTAL 



l RSS, +RSS,))/1 



(RSS l +RSS 2 )/N~6 



where N is equal to the number of otoliths aged, and 

 1 and 2 are either male and female arrowtooth or 

 male and female Kamchatka flounder. Kimura ( 1990) 



4 Scott, K. 1994. Alaska Fisheries Science Center. Resource 

 Ecology and Fisheries Management Division, 7600 Sand Point 

 Waj NE, Seattle, WA98115 Personal commun. 



where p = Arrowtooth flounder CPXJE/Atheresthes 

 CPUE. 



Data on the water temperature distribution of 

 Atheresthes spp. were recorded from 252 of the 492 

 trawl hauls in which Atheresthes spp. were caught. 

 Athcresthes spp. were caught only in temperatures 

 between 0.3" and 5.9"C (Fig. 4), whereas the survey 

 temperatures ranged from -1.5" to 11.7°C. Arrow- 

 tooth flounder were abundant at a wide range of bot- 

 tom temperatures, but they were most abundant at 

 temperatures of 2.1" to 4.6"C. While the total range 



