Nichol: Effects of geography and bathymetry on growth and maturity of Pleuronectes asper 
495 
Figure 1 
Map of the survey stations where yellowfin sole were sampled by the eastern Bering 
Sea crab-groundfish bottom trawl survey of the Alaska Fisheries Science Center from 
1982 to 1994. Northwest (NW) and southeast (SE) areas are delineated by the strata 
boundary line extending from north of Kuskokwim Bay, southeast through the Pribilof 
Islands. 
research surveys are not rou- 
tinely carried out. Thus, cur- 
rent survey biomass esti- 
mates underestimate popula- 
tion abundance of the spe- 
cies; exclusion of juveniles in 
shallow water, as well as sexu- 
ally mature yellowfin sole 
that inhabit these nearshore 
spawning areas during the 
survey period (June-August), 
may introduce a sampling bias 
into estimates of growth and 
estimates of size at matura- 
tion. To determine the extent 
of this potential problem, I 
considered the effects of bot- 
tom depth on fish size. 
Materials and 
methods 
Survey area 
Resource assessment surveys 
were conducted in the eastern 
Bering Sea (Bakkala, 1993; 
Wakabayashi et al., 1985), 
from June to mid-August. Standard survey stations 
were based on a 20 by 20 nautical mile grid that cov- 
ered the area from inner Bristol Bay west to the con- 
tinental slope edge and from the Alaska Peninsula 
north to approximately latitude 61°N (Fig. 1). Bot- 
tom trawl tows of approximately 1.5 nautical miles 
and of 30-min duration were made at each station. 
Surveys began in inner Bristol Bay and generally 
followed north- and south-directed transects, pro- 
ceeding westward with each finished transect. 
Data 
Yellowfin sole otoliths were collected from 3,891 
males and 5,209 females during AFSC surveys, 1982- 
94 (Table 1). Fish were measured to the nearest cen- 
timeter total length (TL) and sagittal otoliths were 
removed and stored in 50% ethanol for subsequent 
age determination. Ages were determined by using 
the break-and-burn technique (Chilton and Beamish, 
1982). 
Female maturity data were collected during 1992- 
94 surveys (Table 2). Maturity codes were based on 
macroscopic gonadal appearance (Nichol, 1995). For 
the purpose of this study, codes were simplified to 
either mature or immature. Females were consid- 
ered mature if ovaries contained yolked or hydrated 
oocytes, or were recently spent. Immature ovaries 
contained no visible oocytes. Bata collected during 
1993 (Table 2) were best for examining spatial dif- 
ferences in female length at maturity because 
samples were collected at nearly all survey stations 
within the 50-m contour line, and maturity code as- 
signments were verified by histological examination 
(Nichol, 1995). 
Analysis 
The total survey area was subdivided into northwest 
and southeast areas (Wakabayashi, 1989; Bakkala, 
1993) extending from north of Kuskokwim Bay south- 
east through the Pribilof Islands (Fig. 1). 
Length at age Factorial analysis of variance 
(ANOVA) models (Reish et al., 1985) was constructed 
independently for males and females to examine the 
variance in length at each age due to geographic area 
(northwest or southeast) and to bottom depth. For 
ANOVA, bottom depths were grouped into three lev- 
els (<30 m, 30-49 m, and >50 m; Table 1). A model, 
which included interaction terms and which was 
pooled across years, was analyzed for both males and 
