Tons of Fishing
Web Directories And resources!

Survey design

Triennial bottom trawl surveys were conducted during the summer months (May-Sep) of 1984, 1987, 1990, 1993, and 1996. These surveys have consistently been conducted in a stratified random sampling pattern (Martin and Clausen, 1995; Munro and Hoff, 1995; Stark and Clausen, 1995; Martin, 1997). The GOA is divided into 49 strata according to water depth and area boundaries defined by the North Pacific Fisheries Management Council (./rr/figures/ ). Allocation of sampling effort within each stratum was based on coefficients of variation, mean CPUE, and sampling densities for all fish species from data collected on previous triennial surveys (see Cochran, 1977). Sampling density within a stratum depended on the anticipated fish density of that particular stratum (Martin and Clausen, 1995; Stark and Clausen, 1995; Martin, 1997). Stations were prioritized within each stratum so that in the event of complications, sampling density would remain controlled. The total survey area was reduced by 7% after 1987 when stations deeper than 500 m were eliminated (Stark and Clausen, 1995). In the interest of consistency, only data from stations less than 500 m in bottom depth were used in our analyses. Station density was spread over a wide range; thus the survey data were most suitable for analyses of geographic distribution and community composition on a large scale (Fig. 2).

[FIGURE 2 OMITTED]

For each survey a Nor'eastern trawl (NT) with a inch codend liner was used. The liner helped retain juvenile pollock (Brown, 1986). The nylon and polyethylene NTs had net dimensions of m wide by m high and m by m, respectively. In 1984 and 1987 a Japanese bottom trawl (JBT) was also used. The horizontal opening of the JBT ranged from 19 to 30 m, the vertical opening from to m (Brown, 1986). CPUE from the JBT were corrected (using Tables 28 and 31 in Munro and Hoff, 1995) to account for differences between the catch-ability of the NTs and the JBT.

Data analyses and statistical considerations

Analyses were conducted on total pollock catch as well as on four size groups of pollock: <150 mm (age-0 or young-of-the-year); 150-230 mm (age-1 juveniles); 230-330 mm (age-2 juveniles); and >330 mm (adults). The length-at-age categories were based on a histogram of pollock lengths from a historical database of all NMFS research surveys conducted during the summer. The catchability of age-0 pollock may not have been as high as that for the other juvenile age classes and thus supports the interpretation of all juvenile data as index values (Bailey and Spring, 1992; Guttormsen and Wilson, 1999). Comparable numbers of length measurements were taken every triennial year but measurements were not taken at every station. Thus, analyses of different age groups were based on subsets of the data. The abundance of fish in a specified age category was influenced by year-class strength (Table 1; Megrey et al., 1996; Hollowed et al. (4)). Strong year classes were present in 1984 (age 0), 1990 (age 2), and 1996 (age 2).

Marine surveys typically yield data sets that are highly variable and contain a substantial proportion of zero catches, particularly when the data set is broken down into age groups (Stefansson, 1996). Thus, in some cases the data were analyzed in terms of presence and absence of pollock (hereafter termed "occurrence"). Because nonzero data often follow a lognormal distribution (Pennington, 1996), the lognormal transformations (log(CPUE+1)) of nonzero density data were used during our analyses.

Changes in the depth distribution of pollock were analyzed by 100-m bottom depth intervals, corresponding to the depth stratification used in the triennial surveys. Changes in geographic distribution of CPUE were evaluated by GOA area boundaries, which again corresponded to the stratification design used during the triennial surveys. In our study, three areas were taken into consideration: "Shumagin" (159[degrees]-170[degrees]W, area 610), "Chirikof" (154[degrees]-159[degrees]W, area 620), and "Kodiak" (147[degrees]-154[degrees]W, area 630).

The distribution of pollock in relation to distance from land was evaluated in 20 nautical mile (nmi) increments by using GIS software (ESRI, Inc., 1996). However, the 20-nmi increments crossed the triennial survey strata boundaries, which may have resulted in the disruption of the stratified random sampling scheme. By addressing pollock occurrence or density at each station, instead of biomass, we avoided the need for stratifying the data after collection. Pollock usually reside at depths less than 300 m within the shelf and slope regions of the GOA (NPFMC (5)). Thus, the relatively narrow shelf region would limit the offshore distribution of pollock.

Initial exploratory analyses involved three-dimensional contingency tables to assess the relationship between the occurrence of pollock and year, bottom depth, geographic region, and distance from land. In cases where the three-dimensional analyses rejected the null hypothesis of independence, more detailed partial contingency tables were constructed for the occurrence of pollock against each individual variable. Other studies suggested that bottom depth and geographic location explain most of the variability in fish distribution data (. Overholtz and Tyler, 1985; Jay, 1996).

Single-factor ANOVAs were performed on the nonzero CPUE data to examine the sources of variation in the data. The In(CPUE+1) of the nonzero data was used as the dependent variable among years. A separate ANOVA was calculated for each category of bottom depth, geographic region, and distance from land.

Examination of the population density data revealed that for most age groups, the number of stations with a low density (<1000 fish/[.2]) of pollock increased whereas the number of high-density stations remained relatively stable. This finding suggested that the distribution of suitable habitat for the demersal fraction of the population might have been expanding. Therefore, we examined the distribution of low-density concentrations of pollock separately in relation to the three physical characteristics across the survey years. Although ocean bottom temperature (OBT) was not measured at every survey station, available data were used to help interpret our results. Mean bottom temperature was averaged over bottom depth intervals for each year to provide a general overview (Fig. 3). Additional information on OBT has been summarized by S. Hare (http://www. .edu/ staff/hare/html/papers/OBT/) from a variety of data sources and averaged over five-year periods.

[FIGURE 3 OMITTED]

Interspecies associations were examined with Bray-Curtis clustering techniques (Boesch, 1977; Walters and McPhail, 1982). Such classification techniques are useful in generating hypotheses about community structure, which may then be used to aid management actions (Cormack, 1971). Because the surveys used in the cluster analyses were conducted during the summer months, the results could not be extrapolated to other seasons. Triennial survey data were log-transformed and clustered by using the group average fusion strategy. For the species clusters, the top thirty species by weight were chosen in addition to the four age groups of pollock. After careful evaluation of all the cluster dendrograms, a common dissimilarity level (dissimilarity coefficient (h)=27) was chosen for all years as the level at which the most clearly defined clusters occurred throughout the triennial survey years.

Diversity indices were calculated for all five survey years. The top ten species by number of fish were combined for all years to make one list of species for which diversity was analyzed. Simpson's diversity index was calculated both in terms of richness and evenness (Simpson, 1949; Tokeshi, 1993). Richness was interpreted to mean "effective number of species," whereas evenness was understood to be the "distribution of numbers of fish amongst those species." Confidence intervals were calculated by jack-knifing the diversity index (Magurran, 1988).

Results

All-pollack

Three-dimensional contingency tables indicated whether occurrence of pollock was mutually independent of all combinations of bottom depth, geographic region, distance from land, and year. The three-dimensional contingency tables for all three null hypotheses of mutual independence resulted in rejection of the null hypothesis (P<). The null hypotheses considered for the three-dimensional and partial contingency tables are listed in Table 2. A summary of the distribution of stations within categories of the three physical characteristics is given in Table 3. All the hypotheses of partial independence on combinations of two of the physical characteristics could be rejected (P< ). When each physical characteristic was tested separately, the only hypothesis that could not be rejected ([alpha]=) was the independence of pollock occurrence against geographic region (P=, Table 2). Graphs of pollock occurrence versus each of the physical characteristics revealed distinctive patterns (Fig. 4). The graphs indicated that pollock are most frequently observed in the 100-200 m bottom depth interval and the 0-20 nmi distance category (Fig. 4, A and B). However, this interpretation may be an artifact of the percentage of stations sampled within each category (Table 3). For all years, the greatest number of stations were sampled in the 100-200 m bottom-depth category and in the 0-20 nmi distance-from-land category.

[FIGURE 4 OMITTED]

The results were standardized by taking the proportion of stations within each category that had a positive occurrence of pollock (Fig. 5). These graphs showed that for each bottom depth category, the proportion of positive occurrences had increased over the years. The proportion of stations where pollock were present was consistently high in the 0-20 nmi category (Fig. 5). Farther from land, the proportion of positive stations increased between 1984 and 1996. In all three geographic regions, there was an increase in the proportion of stations where pollock were present.

[FIGURE 5 OMITTED]

Table 4 lists the number of hauls with pollock occurrence and summarizes the CPUE values for the nonzero data. Results from the single-factor ANOVAs calculated for the nonzero density data are summarized in Table 5. Significant differences were observed among most years, except for the Chirikof region and the distances from land of 0-20 and 40-60 nmi.

Adult pollock

The initial three-dimensional contingency tables of mutual independence between adult pollock occurrence and the three physical characteristics proved significant. Subsequent contingency analyses revealed significant results for all partial independence tests except for the two-dimensional test for independence between occurrence of adult pollock and geographic region (P=).

Examination of all the positive data showed that the mean of the log-transformed CPUE decreased and the variance increased over time for bottom depths 100-200 m in each geographic area (Table 6). The mean CPUE also decreased over time for stations <40 nmi from land but there was no trend in the changes in variance (Table 6). The ANOVA confirmed that changes in the distribution of pollock from 1984 to 1996 were significant () among years at all depth intervals, all geographic areas, and at a distance from land <60 nmi (Table 7). Thus, although the number of stations where pollock could be encountered increased (. the low-density distribution data), the mean adult pollock density decreased and became more variable between 1984 and 1996 (Fig. 5, Table 6).

Changes in the distribution of adult pollock were further examined by sorting the data into seven CPUE bins and by making histograms of the frequency of stations within each bin (Fig. 6). The histogram revealed that over the years, the occurrence of stations with zero adult pollock CPUE decreased, whereas stations with low-density (CPUE <1000 fish/[.2]) concentrations of pollock increased. Higher density stations did not exhibit any particular trend.

[FIGURE 6 OMITTED]

To better understand observed changes in the distribution of low-density concentrations of adult pollock, all the stations were characterized by bottom depth, geographic location, and distance from land (Fig. 7). The number of stations that had low-density concentrations of adult pollock increased in all habitat categories over the years. The proportion of stations within each bottom depth bin with low-density concentrations of adult pollock more than doubled from 1990 to 1996. Comparable dramatic increases were observed with regard to distance from land and geographic region.

[FIGURE 7 OMITTED]

Juvenile pollock

Contingency table analysis for juvenile pollock occurrence resulted in significant (P<) results across all ages for tests of mutual independence among all the physical characteristics. Two-dimensional contingency analyses also resulted in significant results. It was difficult to discern any chronological trends in the mean and variance of all the positive data for juvenile pollock. Minor fluctuations in juvenile pollock density may have been related to interannual differences in year-class strength (Table 1).

Histograms similar to those for adult pollock were made for all three juvenile age groups of pollock, where juvenile pollock density was parsed into bins and plotted (Fig. 8). The same trend seen in the adults was also seen with the juvenile pollock. As the years progressed, there was an increase in the proportion of stations that had low-density (< 1000 fish/[.2]) concentrations of all juvenile age groups of pollock. For age-0 and age-1 pollock, there was about a tenfold difference in the proportion of stations with low-density concentrations after 1987 (Fig. 8). Age-2 pollock also exhibited this trend but not to the same degree as age-0 and age-1 pollock.

[FIGURE 8 OMITTED]

Stations with low-density concentrations of juvenile pollock were examined with respect to bottom depth, geographic area, and distance from land (Fig. 9). For all ages of juvenile pollock there was a marked difference between 1984-87 and the 1990s. A greater proportion of stations with age-0 pollock was observed at bottom depths <200 m than at deeper depths, especially in 1996. Age-0 pollock occurrence increased in the Chirikof and Shumagin regions before also increasing in the Kodiak region in 1996. Age-0 pollock also occurred more frequently at distances <60 nmi from land.

[FIGURE 9 OMITTED]

In 1990, the proportion of stations with low-density concentrations of age-1 pollock increased the most at bottom depths >200 m, before the same increasing trend was also observed at shallower bottom depths in 1993 and 1996

There was a general increase over time of low-density concentrations of age-2 pollock with respect to bottom depth and geographic region (Fig. 9). The increase in the proportion of low-density concentrations of age-2 pollock was the greatest 20-60 nmi from land until 1996, when the proportion of stations with low-density concentrations increased the most >60 nmi from land.

Community composition

Results of clustering by species by using the triennial bottom trawl data are presented in Table 8. The top 30 species or species types (. adult and juvenile pollock) by weight over all survey years were chosen. Adult pollock were consistently associated over the years most closely with flathead sole (Table 8). The cluster containing adult pollock usually included 3-6 other species of importance, either commercially or in terms of abundance. These included arrowtooth flounder (Atheresthes stomias (Jordan and Gilbert, 1880)), Pacific halibut (Hippoglossus stenolepis Schmidt, 1904), Pacific cod (Gadus macrocephalus Tilesius, 1810), sablefish (Anoplopoma fimbria (Pallas, 1814)), Dover sole (Microstomus pacificus (Lockington, 1879)), and rex sole (Glyptocephalus zachirus Locking, ton, 1879). This small cluster was usually isolated from the larger main cluster (. high dissimilarity between the two clusters). Adult pollock were also clustered separately from the juvenile pollock age groups. In all years except 1987, age-1 and age-2 pollock were clustered with each other, whereas age-0 pollock were clustered separately. In 1987, age-0 and age-1 pollock clustered together, although not as closely associated as ages 1 and 2 in the other years. Until 1996, age-0 pollock clustered with Pacific sleeper shark (Somniosus pacificus Bigelow and Schroeder, 1944), Aleutian skate (Bathyraja aleutica (Gilbert, 1896)), or silvergray rockfish (Sebastes brevispinis (Bean, 1884)). In the larger cluster that contained all the juvenile pollock age groups, it was common to find Pacific herring (Clupea pallasii (Cuvier and Valenciennes, 1847)) in all years.

Cluster analysis by station of the triennial data resulted in clusters that fell into clean zoogeographic groups that followed depth contours. Maps of clusters identified seven groups of species that could be tracked in the GOA throughout most years (Table 9). Groups 1, 2, and 5 were present only in 4 out of the 5 triennial survey years. In 1984, groups 2 and 5 were absent whereas in 1990, group 1 was absent. The species within each cluster were listed in order of dominance. Although many of the clusters had the same main species in common, as listed in Table 9, differences in dominance distinguished the various clusters. Adult pollock were most dominant in groups 1, 3, and 6, which were either made up of nearshore or deep shelf stations (Table 9). Mean values of several environmental variables in these station groups in which pollock were found indicate that bottom depth and temperature varied (Table 9). Shallow water stations (groups 1 and 2) were grouped into warm ([degrees]C) and average ([degrees]C) bottom temperature clusters (Table 9). As expected, deeper water stations tended to be characterized by cooler temperatures. Juvenile (ages 0-2) pollock were found mostly in the nearshore stations.

Diversity

Separate diversity indices were calculated for habitat between 0-100 m bottom depth and 100-200 m bottom depth. In the 0-100 m habitat, there was a slight decrease in richness from 9-10 in the 1980s to 7-9 effective number of species in the 1990s (Fig. 10A). Calculations of evenness for the same data indicated that there had been little change in the evenness component of species diversity in the 0-100 m habitat (Fig. 10C). Histograms of the proportion (in terms of number of fish) contributed by each of the top species indicated that the decrease in richness might partially be explained by an increase in the predominance of pollock and a decrease in the presence of rockfish species (Fig. 11, A-E).

[FIGURES 10-11 OMITTED]

Richness in the 100-200 m category increased from 5 to 10 effective numbers of species (Fig. 10B). The predominance of pollock and arrowtooth flounder in the 1980s had shifted to include eulachon (Thaleichthys pacificus (Girard, 1858)) and Pacific ocean perch (Sebastes alutus (Gilbert, 1890); Fig. 11, F-J). Evenness was relatively consistent throughout the survey years (Fig. 10D).

Discussion

A major result of our study was that for all age groups of pollock, the number of stations where pollock occurred at low densities (<1000 fish/[.2]) increased during 1984-96 while the mean density decreased. This pattern suggests that the pollock population was increasing its range, a characteristic often seen in growing populations that is consistent with the "basin hypothesis" (MacCall, 1990). MacCall hypothesized that at low-densities, marine fish will occupy habitats that are optimal for survival. As populations grow, however, some portions of the populations will expand into locations of less suitable habitat quality.

Our analyses of the data indicated that the demersal portion of the pollock population was stable but that the overall (pelagic and demersal) pollock population declined during the study (1984-96, Fig. 1, B and C). Year-class strength was only sporadically strong and therefore would not account for the sustained increased positive occurrence by station during the early- to mid-1990s. Under MacCall's hypothesis, the range of the population should have been stable or contracting. Thus, patterns emerging from our data are not entirely consistent with MacCall's model.

Analysis of CPUE of all positive tows revealed that there was a decrease in mean CPUE over the years from 1984 to 1996. The data showed an increase in low-density concentrations of adult pollock stations across all categories but in particular at bottom depths 200-300 m and in the Chirikof and Shumagin regions. We conclude that adult pollock had expanded into deeper water in the 1990s but that the expansion had resulted in a decrease of adult pollock in high-density stations, combined with decreased mean density of adult pollock throughout the region. We hypothesize that the expansion of pollock was due to an increase in the suitability of habitats for adult pollock, possibly caused by a spread in the distribution of pollock forage, during a period of stable or decreasing pollock population trends. If predators of pollock rely on high-density patches, the decreased mean density of pollock may have negative ramifications for the successful foraging of top predators. We refer to this as the "forage density hypothesis"; . habitat suitability has changed, perhaps on a local scale, such that there has been an expansion of the overall population distribution. However, the expansion, combined with decreasing population numbers, had caused the density of pollock patches to decrease below a threshold at which top predators can successfully forage. The underlying assumption here is that the predators need patches of high prey density rather than a uniform distribution of average or low prey density.

Analysis of trends in the distribution and abundance of juvenile age groups is more difficult because data for these age groups are strongly influenced by interannual variations in year-class strength. However, spatial trends in distribution were consistent with those seen for adult pollock. For all age groups of juvenile pollock there was an increase in the proportion of stations where low densities occurred, and a decrease in the mean CPUE during the time period examined.

The spatial expansion by a species should be reconciled with more detailed analyses of the characteristics of the area it occupied to determine whether changes in habitat suitability have occurred. Adult pollock are not expanding into areas with physical characteristics previously not associated with pollock. Cluster analysis revealed that adult pollock are associated with average bottom temperatures between [degrees]C and [degrees]C. The bottom temperature data from field surveys revealed a consistent range over time of 5-6[degrees]C (Dorn et al. (6)). It was difficult to discern a trend from either our bottom temperature or the OBT data. Temperatures may have been slightly warmer in the 1980s than in the 1990s. Temperature changes, together with the shoaling of the mixed layer depth (Polovina et al., 1995; Shima, 1996), may have caused a redistribution of prey, possibly contributing to changes in pollock distribution.

Evidence of the importance of external forcing on the spatial range of fish species has been noted in other ecosystems. Movement of fish in response to environmental change was recorded in the Barents Sea when Atlantic cod (Gadus morhua Linnaeus, 1758) shifted westward as a consequence of cooler waters (entering from the east) across the region from 1977-81 (Loeng, 1989). Primary shifts in distribution, in response to temperature changes, were seen in younger age classes of fish. Because there were some species (. haddock, Melanogrammus aeglefinus (Linnaeus, 1758)) that did not respond to changes in temperature, it may be that the movement of the fish may be dependent on the sensitivity of fish prey to the temperature shifts (Shevelev et al., 1987). Cod (like pollock) feed mostly on planktonic organisms that may respond rapidly to temperature changes whereas haddock feed mostly on benthic species (Shevelev et al., 1987).

External forcing may also include anthropogenic factors. Domestication (.) of the pollock fishery occurred during the period investigated (Megrey, 1989) with the result that fishing operations shifted from at-sea processors to a reliance on shoreside processing. The shift moved fishing operations closer to shore after the mid-1980s (Fritz (7)). Studies in other regions have found that groundfish species (Pacific whiting and haddock) disperse as a result of trawling. However, within minutes of a vessel's passage, the fish usually revert back to their original distribution (Ona and Godoo, 1990; Nunnallee (8)). It should be noted that this behavior is the result of disturbance by a single vessel. It is possible that the activity of multiple vessels that make up commercial fisheries could cause long-term redistribution of pollock. Studies are currently underway to examine the potential effect of commercial fisheries on the distribution of pollock (Hollowed et al. (9)).

Predator-prey interactions also influence the distribution of marine fish. Adult pollock are capable of exhibiting a strong top-down influence on juvenile abundance through cannibalism (Livingston, 1991). However, in the GOA the incidence of cannibalism is minimal (Yang, 1993), perhaps because, as shown in our study, the ranges of adult and juvenile pollock do not overlap. Spatial distributions showed that the bulk of both adult and juvenile pollock occurred in the Kodiak and Chirikof regions in the GOA. Mean density for both juveniles and adults was high within 20 nmi from land. However, age-0 (and often age-1) pollock were in greatest density in shallow waters <20 nmi from land, whereas adult pollock tended to occupy deeper waters. Age-2 pollock, which are beyond the size range that adult pollock target as food in the Bering Sea, tend to have a distribution similar to that of adults. The separation between juveniles and adults could be a seasonal effect due to the summer sampling season of the surveys because juvenile pollock are known to migrate ontogenetically (Brodeur and Wilson, 1996). However, the pattern of spatial separation between age-0 and age-1 pollock and adult pollock may indicate avoidance of adults by juvenile pollock.

An examination of changes in density of juvenile pollock in relation to geographic location allowed evaluation of the prevailing conceptual model for pollock ontogeny in the Gulf of Alaska (Kendall et al., 1996). The majority of spawning for the GOA stock occurs in the Shelikof Strait region (Kendall and Picquelle, 1989; Hinckley et al., 1991; Kendall et al., 1996). Survey and ocean modeling studies show that larvae and early juveniles are advected by the Alaska Current towards the Shumagin Islands (Hinckley et al., 1991; Hermann et al., 1996; Hinckley, 1999). Our results are partially consistent with the conceptual model for pollock ontogeny: the area encompassing Shelikof Strait and west to the Shumagin Islands (. the Kodiak and Chirikof regions) was important for all juvenile age groups. However, the disproportionate increase of pollock in the Kodiak region in the 1990s, with two strong year classes (1988 and 1994), supports the suggestion that substantial spawning may also occur outside of Shelikof Strait (Smith et al., 1984; Brodeur and Wilson, 1996).

The species composition of the cluster in which pollock was included changed little over the course of the survey years. Adult pollock were usually located together with commercially important flatfish, arrowtooth flounder, and Pacific cod. This finding is consistent with other findings from the eastern Bering Sea where pollock often were associated with snow and Tanner crabs (Chionoecetes spp.), Pacific cod, flathead sole, Greenland turbot (Reinhardtius hippoglossoides (Gill, 1861)), and yellowfin sole (Limanda aspera (Pallas, 1814)) from 1978 to 1981 (Walters and McPhail, 1982).

Depth contours served as demarcations of station clusters in our study as well as in others conducted in the eastern Bering Sea and the eastern Pacific Ocean along the western coast of the United States (Gabriel and Tyler, 1980; Walters and McPhail, 1982; Jay, 1996). The consistency in how the stations were grouped together is common throughout all these studies. Walleye pollock in the eastern Bering Sea also dominated clusters of stations in the central shelf or the outer shelf. Other species commonly found together with pollock in these groups were yellowfin sole and Pacific cod.

A hypothesis relating pollock distribution to sea lion foraging behavior

Because pollock predominate the Steller sea lion diet (Merrick and Calkins, 1996), the temporal and spatial changes in pollock abundance and distribution could affect Steller sea lions. The Steller sea lion population underwent precipitous declines in the 1980s and declines continued in the 1990s but not as steep (Sease et al., 1999; NMFS (10)). The rate of decline continues to be high towards the center of the Gulf of Alaska (150-158[degrees]W), whereas the western Gulf region (158-162[degrees]W) has stabilized (. the rate of decline has decreased).

Survey results suggest that the "forage density hypothesis" may apply to Steller sea lions in the GOA. There is evidence indicating that Steller sea lions target dense aggregations of prey (Sinclair and Zeppelin, in press), so that the lower mean density of pollock may reduce the number of successful Steller sea lion foraging trips in the central GOA. This effect may have prevented the sea lion population from recovering from the precipitous decline that occurred in the 1980s. However, it is important to keep in mind that pollock are only one component of the Steller sea lion diet and that all pollock survey data are collected during the summer.

Table 1
Summary of year-class strength of juvenile walleye pollock in bottom
trawl survey years.

 1984 1987 1990 1993 1996

Age 0 Strong Weak Weak Weak Weak
Age 1 Weak Weak Average Weak Weak/Average
Age 2 Weak Strong/Average Strong Weak Strong

Table 2
Contingency table analysis of walleye pollock occurrence.

Type of table P-value

Three-dimensional tables of mutual independence
 Null hypothesis
 Pollock occurrence, year, and bottom
 depth are mutually independent P< 
 Pollock occurrence, year, and distance
 from land are mutually independent P< 
 Pollock occurrence, bottom depth, and
 region are mutually independent P< 
 Pollock occurrence, region, and distance
 from land are mutually independent P< 

Three-dimensional tables of partial independence
 Null hypothesis
 Pollock occurrence is independent of year
 and bottom depth P< 
 Pollock occurrence is independent of year
 and distance from land P< 
 Pollock occurrence is independent of
 bottom depth and region P< 
 Pollock occurrence is independent of
 region and distance from land P< 

Two-dimensional tables of partial independence
 Null hypothesis
 Pollock occurrence is independent of
 year P< 
 Pollock occurrence is independent of
 bottom depth P< 
 Pollock occurrence is independent of
 region P=
 Pollock occurrence is independent of
 distance from land P< 

Table 3
Summary of the percentage of stations within each bin for the
three types of physical parameters discussed in the text. The
number in parentheses indicates the total number of stations.

 Percentage of stations
 in category by year

Physical category 1984 (749) 1987 (648) 1990 (534)

Bottom depth (m)
 0-100 
 100-200 
 200-300 
 300-500 

Geographic location
 Shumagin 
 Chirikof 
 Kodiak 

Distance from land category
 (nmi)
 0-20 
 20-40 
 40-60 
 60+ 

 Percentage of stations
 in category by year

Physical category 1993 (616) 1996 (613)

Bottom depth (m)
 0-100 
 100-200 
 200-300 
 300-500 

Geographic location
 Shumagin 
 Chirikof 
 Kodiak 

Distance from land category

 (nmi)
 0-20 
 20-40 
 40-60 
 60+ 

Table 4
Summary information of walleye pollock occurrence and CPUE by
triennial survey year. CPUE values are given in number of
fish/[.2] and summarize only the nonzero positive tows.

 Number of hauls CPUE

 Pollock
Year Total present Minimum Maximum Mean SD

1984 672 375 
1987 603 412 
1990 506 386 
1993 583 454 
1996 593 497 

Table 5
Single-factor ANOVAs of ln(CPUE+1) of positive walleye
pollock catches among years by bottom depth category,
geographic region, and distance from land. SS = sum of
squares; MS = mean square.

 SS df MS F P

Bottom depth (m)
 0-100 4 
 100-200 4 <
 200-300 4 <
 300-500 4 

Geographic area
 Shumagin 4 <
 Chirikof 4 
 Kodiak 4 <

Distance offshore (nmi)
 0-20 4 
 20-40 4 
 40-60 4 
 60+ 4 <

Table 6
Mean and variance estimates of ln(CPUE+1) for 1984-96 for nonzero
adult pollock data by physical characteristic. The number of
observations is given in parentheses.

 Mean

Physical characteristic 1984 1987 1990

Bottom depth (m)
 0-100 (24) (37) (23)
 100-200 (121) (110) (158)
 200-300 (52) (24) (60)
 300-500 (1) 0 (1) (1)

Geographic area
 Shumagin (63) (37) (33)
 Chirikof (57) (49) (52)
 Kodiak (78) (82) (134)

Distance from land (nmi)
 0-20 (141) (140) (161)
 20-40 (42) (19) (57)
 40-60 (13) (10) (16)
 60+ (2) (2) (8)

 Mean Variance

Physical characteristic 1993 1996 1984 1987

Bottom depth (m)
 0-100 (74) (108) 
 100-200 (204) (192) 
 200-300 (64) (82) 
 300-500 (12) (30) 0 0

Geographic area
 Shumagin (101) (121) 
 Chirikof (83) (125) 
 Kodiak (138) (138) 

Distance from land (nmi)
 0-20 (236) (269) 
 20-40 (96) (86) 
 40-60 (20) (40) 
 60+ (10) (7) 

 Variance

Physical characteristic 1990 1993 1996

Bottom depth (m)
 0-100 
 100-200 
 200-300 
 300-500 0 

Geographic area
 Shumagin 
 Chirikof 
 Kodiak 

Distance from land (nmi)
 0-20 
 20-40 
 40-60 
 60+ 

Table 7
Single-factor ANOVAs of ln(CPUE+1) of positive adult
walleye pollock catches among years by bottom depth
category, geographic region, and distance from land.
SS = sum of squares; MS = mean square.

 SS df MS F P

Bottom depth (m)
 0-100 4 <
 100-200 4 <
 200-300 4 <
 300-500 4 <

Geographic area
 Shumagin 4 <
 Chirikof 4 <
 Kodiak 4 <

Distance (nmi)
 0-20 4 <
 20-40 4 <
 40-60 4 <
 60+ 4 

Table 8
Species associated with walleye pollock according to Bray-Curtis
cluster analyses of triennial survey data. The species
are listed in order of association.

Adults Age 0 Ages 1 and 2

Flathead sole Pacific sleeper shark Atka mackerel
Sablefish silvergray rockfish eulachon
Dover sole lingcod Pacific herring
Rex sole longnose skate bigmouth sculpin
Arrowtooth sharpchin rockfish lingcod
 flounder
Pacific halibut Pacific herring chinook salmon
Pacific cod

Table 9
Description of consistent station groups found as a result of
cluster analyses of the triennial bottom trawl survey data. The
group designation "flatfish" includes all flatfish except
arrowtooth flounder. Mean values are given for the environmental
variables. Standard deviations are given in parentheses.

 Bottom depth SST
Group no. Location (m) ([degrees]C)

1 Nearshore 1 () ()
2 Nearshore 2 () ()
3 Shallow shelf () ()
4 Deep shelf () ()
5 Inner slope () ()
6 Middle slope () ()
7 Outer slope () ()

 Temperature at
 depth
Group no. ([degrees]C) Main species

1 () Age 2 and Adult pollock, Flatfish
2 () Adult pollock, Flatfish, Pacific cod
3 () Arrowtooth flounder, Adult pollock,
 Pacific halibut, Pacific cod
4 () Arrowtooth flounder, Flatfish,
 Pacific cod, Adult pollock
5 () Northern rockfish, Pacific ocean perch,
 Arrowtooth flounder
6 () Adult pollock, Pacific ocean perch,
 Arrowtooth flounder, Sablefish
7 () Giant grenadier, Sablefish, Rockfish

Acknowledgments

We would like to thank Eric Brown, Lowell Fritz, Beth Sinclair, Matt Wilson, Warren Wooster, and several anonymous reviewers for comments on earlier drafts of the manuscript. An earlier version of this manuscript appeared as part of a . dissertation (Shima, 1996) for the University of Washington, Seattle. This study was made possible through the support of the Resource Ecology and Fisheries Management Division of the Alaska Fisheries Science Center, National Marine Fisheries Service and the Washington Cooperative Fish and Wildlife Research Unit, Biological Resources Division, . Geological Survey.

(1) Mundell, G. 1999. Personal commun. Alaska Fisheries Science Center, NMFS/NOAA, 7600 Sand Point Way NE, Seattle, WA 98115.

(2) Wespestad, V. 1995. Walleye pollock. In Plan team for groundfish fisheries of the Bering Sea/ Aleutian Islands stock assessment and fishery evaluation report for the ground fish resources of the Bering Sea/Aleutian Islands region as projected for 1996, 35 p. North Pacific Fishery Management Council, . Box 103136, Anchorage, AK 99510.

(3) Guttormsen, M. A., and C. D. Wilson. 1999. Echo integration-trawl survey results for walleye pollock in the Gulf of Alaska during 1998. In Stock assessment and fishery evaluation report for the 1998 Gulf of Alaska groundfish fishery, 25 p. North Pacific Fishery Management Council, 605 W 4th Avenue, Suite 306, Anchorage, Alaska 99501.

(4) Hollowed, A. B., E. Brown, B. Megrey, and C. Wilson. 1996. Walleye pollock. In Stock assessment and fishery evaluation report for the 1997 Guff of Alaska groundfish fishery, 64 p. North Pacific Fishery Management Council, 605 W 4th , Suite 306, Anchorage, AK 99501.

(5) NPFMC (North Pacific Fishery Management Council). 1998. Essential fish habitat assessment report for the groundfish resources of the Gulf of Alaska region. NPFMC, 605 West 4th Avenue, Suite 306, Anchorage, AK 99501, 117 p.

(6) Dorn, M. W., A. B. Hollowed, E. Brown, B. Megrey, C. Wilson, and J. Blackburn. 1999. Walleye pollock. In Stock assessment and fishery evaluation report for the 2000 Gulf of Alaska groundfish fishery, 65 p. North Pacific Fishery Management Council, . Box 103136, Anchorage, Alaska 99510.

(7) Fritz, L. W. 1993. Trawl locations of walleye pollock and Atka mackerel fisheries in the Bering Sea, Aleutian Islands and Gulf of Alaska from 1977-92. . Dep. Commer., AFSC Processed Report 93-08, 162 p.

(8) Nunnallee, E. P. 1991. An investigation of the avoidance reactions of Pacific whiting (Merluccius productus) to demersal and midwater trawl gear. ICES . 1991/B:5, 16 p. [Available from Alaska Fisheries Science Center, NMFS/NOAA, 7600 Sand Point Way NE, Seattle, WA 98115.]

(9) Hollowed, A. B., C. Wilson, M. Shima, and P. Walline. 2001. Study to determine the effect of commercial fishing on walleye pollock distribution and abundance. In Steller sea lion investigations, 2000 03. S. Fadely, ed.), p. 89-104. . Dep. of Commer., NOAA, NMFS, AFSC Processed Report 2001-05.

(10) NMFS (National Marine Fisheries Service). 1995. Status review of the United States Steller sea lion, Eumetopias jubatus, population. National Marine Mammal Laboratory, Alaska Fisheries Science Center, 7600 Sand Point Way NE, Seattle, WA 98115, 61 p.

Literature cited

Bailey, K. M., and S. M. Spring. 1992. Comparison of larval, age-0juvenile and age-2 recruit abundance indices of walleye pollock, Theragra chalcogramma, in the western Gulf of Alaska. ICES J. Mar. Sci. 49:297-304.

Boesch, D. F. 1977. Application of numerical classification in ecological investigations of water pollution. Va. Inst. Mar. Sci., Spec. Sci. Rep. 77, 114 p.

Brodeur, R. D., and M. T. Wilson. 1996. A review of the distribution, ecology, and population dynamics of age-0 walleye pollock in the Gulf of Alaska. Fish. Oceanogr. 5 (suppl. 1):148-166.

Brown, E. S. 1986. Preliminary results of the 1984 .-Japan cooperative bottom trawl survey of the Central and Western Gulf of Alaska. In Condition of groundfish resources of the Gulf of Alaska region as assessed in 1985 (R. L. Major, ed.), p. 259-296. . Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-106.

Cochran, W. G. 1977. Sampling techniques. Wiley Series in Probability and Mathematical Statistics-Applied. John Wiley & Sons. New York, NY, 428 p.

Cormack, R. M. 1971. A review of classification . Statist. Soc., series A, 134(3):321-367.

ESRI, Inc. (Environmental Systems Research Institute). 1996. ArcView, version . ESRI, Inc., Redlands, CA.

Gabriel, W. L., and A. V. Tyler. 1980. Preliminary analysis of Pacific coast demersal fish assemblages. Mar. Fish. Rev. 42:83-88.

Hermann, A. J., S. Hinckley, . Megrey, and . Stabeno. 1996. Interannual variability of the early life history of walleye pollock near Shelikof Strait as inferred from a spatially explicit individual-based model. Fish. Oceanogr. 5(suppl. 1):39-57.

Hinckley, S. 1999. Biophysical mechanisms underlying the recruitment of walleye pollock (Theragra chalcogramma). . diss., School of Fisheries, Univ. Washington, seattle, WA, 259 p.

Hinckley, S., K. M. Bailey, S. J. Picquelle, J. D. Schumacher, and P. J. Stabeno. 1991. Transport, distribution, and abundance of larval and juvenile walleye pollock (Theragra chalcogramma) in the western Gulf of Alaska. Can. J. Fish. Aquat. Sci. 48:91-98.

Jay, . 1996. Distribution of bottom-trawl fish assemblages over the continental shelf and upper slope of the . west coast, 1977-1992. Can. J. Fish. Aquat. Sci. 53:1203-1225.

Kendall, A W., Jr., and S. J. Picquelle. 1989. Egg and larval distributions of walleye pollock Theragra chalcogramma in Shelikof Strait, Gulf of Alaska. Fish. Bull. 88:133-154.

Kendall, A. W., Jr., J. D. Schumacher, and S. Kim. 1996. Walleye pollock recruitment in Shelikof Strait: applied fisheries oceanography. Fish. Oceanogr. 5 (suppl. 1):4-18.

Livingston, P. A. 1991. Walleye pollock. In Ground fish food habits and predation on commercially important prey species in the eastern Bering Sea from 1984 to 1986 (P. A. Livingston, ed.), p. 9-30. . Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-207.

Loeng, H. 1989. The influence of temperature on some fish population parameters in the Barents Sea. J. Northwest. Atl. Fish. Sci. 9:103-113.

MacCall, A. 1990. Dynamic geography of marine fish populations. Books in recruitment fishery oceanography, vol. 2. Univ. Washington Sea Grant Program, Seattle, WA, 153 p.

Magurran, A. E. 1988. Ecological diversity and its measurement. Princeton Univ. Press, Princeton, NJ, 179 p.

Martin, M. H. 1997. Data report: 1996 Gulf of Alaska bottom trawl survey. . Dep. Commer., NOAA Tech. Memo. NMFS-AFSC-82, 235 p.

Martin, M. H., and D. M. Clausen. 1995. Data report: 1993 Gulf of Alaska bottom trawl survey. . Dep. Commer., NOAA Tech. Memo. NMFS-AFSC-59, 217p.

Megrey, B. A. 1989. Exploitation of walleye pollock resources in the Gulf of Alaska, 1964-1988: portrait of a fishery in transition. In Proceedings of the international symposium on the biology and management of walleye pollock, p. 679-900. Alaska Sea Grant Rep. 89-1, Univ. Alaska, Fairbanks, AK.

Megrey, B .A., A. B. Hollowed, S. R. Hare, S. A. Macklin, and P. J. Stabeno. 1996. Contributions of FOCI research to forecasts of year-class strength of walleye pollock in Shelikof Strait, Alaska. Fish. Oceanogr. vol. 5. (suppl. 1), p. 189-203.

Merrick, R. L., and D. G. Calkins. 1996. Importance of juvenile walleye pollock in the diet of Guff of Alaska Steller sea lions. In Ecology of juvenile wall-eye pollock, Theragra chalcogramma, (R. D. Brodeur, R A. Livingston, T. R. Loughlin, and A. B. Hollowed, eds.), p. 153-166. . Dep. Commer., NOAA Tech. Rep. NMFS 126.

Munro, P. T., and R. Z. Hoff. 1995. Two demersal trawl surveys in the Gulf of Alaska: implications of survey design and methods. . Dep. Commer., NOAA Tech. Memo. NMFS-AFSC-50, 139 p.

Ona, E., and O, R. Godo. 1990. Fish reaction to trawling noise: the significance for trawl sampling. Rapp. . Reuns. Cons. Int. Explor. Mer 189:159-166.

Overholtz, W. J., and A. V. Tyler. 1985. Long-term responses of the demersal fish assemblages of Georges Bank. Fish. Bull. 83:507-520.

Pennington, M. 1996. Estimating the mean and variance from highly skewed marine data. Fish. Bull. 94:498-505.

Pitcher, K. W. 1981. Prey of the Steller sea lion, Eumetopias jubatus, in the Gulf of Alaska. Fish. Bull. 79:467-472.

Polovina, J. J., G. T. Mitchum, and G. T. Evans. 1995. Decadal and basin-scale variation in mixed layer depth and the impact on biological production in the Central and North Pacific, 1960-88. Deep-sea Res. 42:1701-1716.

Schumacher, J. D., and A. W. Kendall Jr. 1991. Some interactions between young walleye pollock and their environment in the western Gulf of Alaska. Calif. Coop. Oceanic Fish. Invest. Rep. 32:22-40.

Sease, J. L., J. M. Strick, R. L. Merrick, and J. P. Lewis. 1999. Aerial and land-based surveys of Steller sea lions (Eumetopias jubatus) in Alaska, June and July 1996. . Dep. Commer., NOAA Tech. Memo. NMFS-AFSC-99, 43 p.

Shevelev, M. S., V V. Tereschenko, and N. A. Yaragina. 1987. Distribution and behaviour of demersal fishes in the Barents and Norwegian Seas, and the factors influencing them. In The effect of oceanographic conditions on distribution and population dynamics of commercial fish stocks in the Barents Sea: proceedings of the third Soviet-Norwegian symposium, Murmansk, 26-28 May 1986 (H. Loeng, ed.), p. 181-190. Institute of Marine Research, Bergen.

Shima, M. 1996. A study of the interaction between walleye pollock and Steller sea lions in the Gulf of Alaska. . diss., School of Fisheries, Univ. Washington, Seattle, WA, 197 p.

Shima, M., A. B. Hollowed, and G. R. VanBlaricom. 2000. Response of pinniped populations to directed harvest, climate variability and commercial fishery activity: a comparative analysis. Rev. Fish. Sci. 8:89-124.

Simpson, E. H. 1949. Measurement of diversity. Nature 163:688.

Sinclair, E. H., and T. K. Zeppelin. In press. Seasonal and spatial differences in diet in the western stock of Steller sea lions (Eumetopias jubatus). J. Mammol.

Smith, G. B., G. E. Walters, P. A. Raymore Jr., and W. A. Hirschberger. 1984. Studies of the distribution and abundance of juvenile groundfish in the northwestern Gulf of Alaska, 1980-82: Part I: Three-year comparison. . Dep. Commer., NOAA Tech. Memo. NMFS F/NWC-59, 100 p.

Stark, J. W., and D. M. Clausen. 1995. Data report: 1990 Gulf of Alaska bottom trawl survey. . Dep. Commer., NOAA Tech. Memo. NMFS-AFSC-49, 221 p.

Stefansson, G. 1996. Analysis of groundfish survey abundance data: combining the GLM and delta approaches. ICES J. Mar. Sci. 53:577-588.

Steinberg, E. K., and P. Kareiva. 1997. Challenges and opportunities for empirical evaluation of "spatial theory." In Spatial ecology: the role of space in population dynamics and interspecific interactions (D. Tilman and P. Kareiva, eds.), p. 318-332. Princeton Univ. Press, Princeton, NJ.

Tokeshi, M. 1993. Species abundance patterns and community structure. Advances in Ecological Research 24:111-186.

Walters, G. E., and M. J. McPhail. 1982. An atlas of demersal fish and invertebrate community structure in the eastern Bering Sea: Part 1, 1978-81. . Dep. Commer. NOAA Tech. Memo. NMFS F/NWC-35, 122 p.

Wyllie-Echeverria, T., and W. S. Wooster. 1998. Year to year variations in Bering Sea ice cover and some consequences for fish distributions. Fish. Oceanogr. 7:159-170.

Yang, M-S. 1993. Food habits of the commercially important groundfishes in the Gulf of Alaska in 1990. . Dep. Commer., NOAA Tech. Memo. NMFS-AFSC-22, 150 p.

Manuscript accepted 21 September 2001.

Michiyo Shima
School of Aquatic and Fishery Sciences
University of Washington
1122 Boat St. NE
Seattle, Washington 98105
E-mail address: .edu

Anne Babcock Hollowed
Alaska Fisheries Science Center
7600 Sand Point Way NE
Bin C15700
Seattle, Washington 98115

Glenn R. VanBlaricom
Washington Cooperative Fish and Wildlife Research Unit
School of Aquatic and Fishery Sciences
University of Washington
Box 355020
Seattle, Washington 98195