© 2006 International Council for the Exploration of the Sea
Interannual changes in distribution of age-0 walleye pollock near the Pribilof Islands, Alaska, with reference to the prediction of pollock year-class strength
School of Aquatic and Fishery Sciences, University of Washington Seattle, WA 98195, USA
*Correspondence to A. G. Winter: Present address: Fishery Industrial Technology Center, University of Alaska, Kodiak, AK 99615, USA; tel: +1 907 486 1509; fax: +1 907 486 1540. e-mail: ffagw{at}uaf.edu.
Walleye pollock (Theragra chalcogramma) is a key species and a major commercial fishery target in the eastern Bering Sea (EBS). Recruitment of juveniles to the adult stock is largely determined by first-year survival in favourable nursery areas, and the waters around the Pribilof Islands are an important nursery area. Based on a six-year acoustic survey programme (September 1994–1999) to investigate abundance and distribution variations of age-0 pollock, we developed a set of predictive indices relating the age-0 Pribilof population to the EBS pollock stock at recruitment (age-3). EBS year-class strength at recruitment correlates with age-0 density, the spatial relationship between juveniles and euphausiids, and the degree of centricity of the age-0 distributions around the Pribilof Islands, i.e. neither too far inshore nor offshore. Seabird numbers and density ratios of predatory groundfish also influenced age-0 pollock distribution, but did not provide consistent discrimination among year classes. We hypothesize that age-0 pollock must transition from a copepod diet to a euphausiid diet by September to maximize their survival potential and to estab ish a strong adult cohort.
Keywords: acoustic surveys, Pribilof Islands, walleye pollock, year-class strength
Received 29 September 2005; accepted 17 March 2006.
 |
Introduction |
|---|
 Top Introduction MethodsResultsDiscussionSummaryReferences |
|---|
Predicting juvenile recruitment is considered key to managing marine fish stocks that have variable abundance (Russell and Yonge, 1936; Bradford, 1992). In the eastern Bering Sea, Alaska, juvenile stages have been the focus of studies on walleye pollock (Theragra chalcogramma) year-class strength (Brodeur et al., 1996; Macklin et al., 2002). Walleye pollock is a key species in the Bering Sea ecosystem (Macklin et al., 2002) and is the mainstay of major commercial fisheries (National Research Council, 1996; Ianelli, 2005). Variability of eastern Bering walleye pollock year classes (cf. Table 1.14 of Ianelli et al., 2003) is compatible with a lognormal distribution. Hennemuth et al. (1980) first proposed this statistical distribution for marine fish stocks because of the common characteristic of having mostly years of low-to-average abundance interspersed with occasional very strong ones.
Efforts to develop a year-class strength index for walleye pollock (hereafter referred to as pollock) have focused on the waters surrounding the Pribilof Islands, near the edge of the eastern Bering Sea (EBS) shelf (Swartzman et al., 2002, 2005; Winter et al., 2005). The Pribilof area has been noted as a centre of abundance of age-0 juveniles (Nishimura et al., 1996; Traynor and Smith, 1996) and characterized as a pollock nursery (Brodeur, 1998; Macklin et al., 2002; Swartzman et al., 2002). Swartzman et al. (2005) synthesized the results of multiple surveys and estimated that in some years, as much as half the EBS stock of age-0 pollock may come from the Pribilof Islands. Schumacher and Macklin (2004) cited evidence suggesting that the age-0 pollock population around the Pribilof Islands can provide a useful index of year-class strength.
Research cruises by the National Oceanic and Atmospheric Administration (NOAA) surveyed the Pribilof area each year in September from 1994 through 1999, with emphasis on measuring the distribution of age-0 pollock and associated biological and oceanographic data. Through summer and autumn, the Pribilof area is structured by a hydrographic front that separates the tidally mixed nearshore zone from stratified offshore waters (Stabeno et al., 1999). Zooplankton aggregate at this front (Coyle and Cooney, 1993) and they attract consumers including seabirds and juvenile and adult pollock (Kinder et al., 1983; Decker and Hunt, 1996; Springer et al., 1996; Brodeur et al., 2000). The limited foraging ranges of age-0 pollock suggest that close proximity to prey is an important determinant of feeding success, especially in times of limited prey abundance (Swartzman et al., 1999a). Predation pressure can force juvenile pollock away from their food (Swartzman et al., 2002), but stratification of the water, where present, may keep juveniles segregated from cannibalistic adults (Francis and Bailey, 1983; Bailey, 1989; Swartzman et al., 2002).
The 1994–1999 NOAA Pribilof cruises collected primarily two-frequency acoustic data which, through validated algorithms, are scalable as biomass indices of fish and zooplankton (Swartzman et al., 2002). We utilized this database to investigate the potential of Pribilof age-0 pollock as an early predictor for year-class strength by analysing the acoustically derived biomass indices in the context of (i) the distribution of age-0 pollock in relation to the distribution of euphausiids, a major prey item (Brodeur et al., 2000; Schabetsberger et al., 2000, 2003; Ciannelli et al., 2004); (ii) the distribution of age-0 pollock in relation to the distribution of groundfish and seabird predators; and (iii) the distribution of age-0 pollock in relation to frontal regions and relative north–south (latitudinal) distance from the Pribilof Islands. Interannual variations in these acoustic density and distribution relationships were then compared with the subsequent abundance of EBS pollock year classes at recruitment (age-3). Specifically, we tested the null hypothesis that age-0 population abundance alone in the Pribilof area predicts year-class strength in the EBS. The alternative hypotheses were that one or more of the three relationships investigated (age-0 pollock distribution vs. euphausiid prey, predators, distance) are important in predicting year-class strength.
|
Methods |
|---|
|
TopIntroductionMethodsResultsDiscussionSummaryReferences |
|---|
Surveys
The Pribilof surveys were carried out each year between 5 September and 22 September (exact dates varied by year), and run along four transect lines radiating north and south of St Paul and St George Islands (Figure 1). In addition to acoustic sampling, CTD casts were used to derive temperature profiles, bird counts at the sea surface were recorded, and trawl data were collected. Transect lines were perpendicular to the frontal gradient, and each transect consisted of distinct habitat regions on the basis of its hydrographic or bathymetric structure (Swartzman et al., 1999a, 2002). Transect A was divided into a tidally mixed nearshore region, a front region that is partially stratified, and an offshore, fully stratified region. The nearshore boundary of the front region was defined at the first CTD station, showing a well mixed (or weakly stratified) vertical profile. The offshore boundary of the front was defined at the location of the stratified region, where the depth range of the thermocline increased by a factor of two over its average width farther offshore (Stabeno et al., 1999). By these criteria, the front extended 13.5–28.5 km from shore in 1994, 7–33 km from shore in 1995, 6–18 km from shore in 1996, 6–30 km from shore in 1997 (Ciannelli et al., 2004), and 8–26 km from shore in 1998 and 1999 (Swartzman et al., 2002). On transects B, C, and D, the tidally mixed nearshore zones were absent or <1 km wide, and therefore not considered a separate habitat. Instead, transects B and C were each divided into a shallow and a deep region by the 70 m isobath. Transect D was divided into a slope region (<130 m) and a basin region (>130 m; Swartzman et al., 2002).
|
Generally, each transect was passed over at least once during daylight and once at night per survey year (see Table 1 of Swartzman et al., 2002, for a summary of transect passes). Exceptions were transects C and D in 1994 and B in 1999, which were run by day only. Some transect passes included twilight, and to quantify the proportion of each pass' runtime in daylight we calculated the sum of multiplying 1, 0.5, 0, times the fraction of extent by day, twilight, and night (available from interpolating the sunrise, sunset, and twilight time points of the corresponding date and coordinates on the US Naval Observatory website: aa.usno.navy.mil).
Echo-integration
Acoustic data were collected by the survey ship at 38 and 120 kHz using a hull-mounted SIMRAD EK-500 split-beam echosounder. In 1994 only, the 38 and 120 kHz frequencies were recorded separately on two survey ships operating in tandem. The echo data were integrated to units (pixels) of approximately 9 m horizontal x 0.5 m vertical resolution (1.0 m vertical resolution in 1998), and processed according to the image-analysis methods described in Swartzman et al. (1999a, 2002). Briefly, pixel echo-integration was classified as fish if it was within the threshold range –53 dB to –40 dB backscattering strength at 38 kHz. Echo-integration was classified as zooplankton if it was within the threshold range –62 dB to –45 dB at 120 kHz; plus 
5 dB higher at 120 kHz than at 38 kHz. Acoustic evidence to distinguish fish from zooplankton by 120 – 38 kHz differencing is discussed in Miyashita et al. (1997) and Kang et al. (2002). For the 2x lower pixel resolution in 1998, the 120 – 38 kHz differencing spread was rescaled to 5 dB x (0.5 m/1 m)0.5 = 3.54 dB, according to the calibration algorithm in Swartzman (2004). Pixels retained by the respective thresholds were morphologically filtered (Haralick and Shapiro, 1992) to delineate contiguous shoals of fish or zooplankton, while scattered or isolated pixels were eliminated. Figure 2a shows an example of a transect with fish backscatter in blue and zooplankton backscatter in red. Shading of the pixels in Figure 2a is proportional to backscatter strength.
|
The area backscattering coefficient of a pixel, sa (m2 m–2) (where backscattering strength, dB = 10 log10(sa); MacLennan et al., 2002), is linearly proportional to the sum of individual acoustic targets (i.e. fish or zooplankton) within the perimeter of that pixel (Foote, 1983). By extension,
sa of contiguous pixels making up a shoal, normalized for the horizontal and vertical resolutions of the pixels, represents an acoustic index of biomass within that shoal. Acoustic fish shoals in the Pribilof surveys were assumed to be juvenile pollock, because juvenile pollock comprised >90% of fish catch in trawls taken along transects (Brodeur et al., 2002; Schabetsberger et al., 2003). Acoustic zooplankton shoals were assumed to be euphausiids. Target-strength models (Stanton et al., 1993) together with survey catch data indicate that euphausiids of 15–25 mm body length (Thysanoessa spp.; Schabetsberger et al., 2000, 2003) would dominate the acoustic backscatter in the zooplankton threshold range, as described above (Swartzman et al., 2002). Data analyses of these acoustic indices assumed that the algorithms correctly partition age-0 pollock and euphausiid components from the backscatter integration. These assumptions were included in the variance components of the analyses, and the same algorithms were applied to each year's data.
Acoustic target strength
Individual target-strength values (TS) were collected by the echosounder system in conjunction with echo-integration. TS values can be converted to estimates of fish length through standard equations in the fisheries acoustics literature (Love, 1971; Foote and Traynor, 1988; Traynor, 1996). Strong TS can thus be taken to represent fish of a size likely to be predators of age-0 pollock (as shown in Figure 2b for one acoustic transect). Around the Pribilof Islands, predators are predominantly adult and sub-adult pollock, followed by Pacific cod (Gadus macrocephalus) (Lang et al., 2003), which present a similar acoustic profile as inferred from studies of the morphometrically analogous (Schultz and Welander, 1935) congener Gadus morhua (Foote, 1987; Rose and Porter, 1996). As pollock and Pacific cod are acoustically indistinguishable at this scale of analysis, we refer to them together as groundfish predators. We used Foote and Traynor's (1988) standard equation of mean acoustic target strength at 38 kHz as a function of pollock fork length (L, in cm):
|
|
|
|
Groundfish lengths derived from these equations were divided into four categories corresponding to relationships between nominal year class and length cited in Dwyer et al. (1987) for pollock ages 1, 2, 3, and 4+ (Table 1). TS values were upper-limited to the equivalent of 100 cm, based on Hart's (1973) length maximum for Pacific cod (slightly bigger than Hart's reported length maximum of pollock: 91 cm). TS values corresponding to age-0 pollock lengths were based on extrapolations of mean backscattering cross-section 
bs (TS = 10 log10(bs)) from the Kirchhoff ray-mode model (J. Horne, University of Washington, pers. comm.). Dates of age-0 pollock surveys varied interannually, so a length range for age-0 pollock was set for each year separately based on Methot trawl samples. As measured target strength of any fish varies with its tilt angle (Blaxter and Batty, 1990; MacLennan et al., 1990), TS thresholds must be considered an approximate index of true size categories. However, observations by Horne (2003, cf. his Figure 5) showed that pollock predominantly maintain a horizontal aspect without bias towards positive or negative inclination or lateral roll. We therefore considered the categorization sufficiently precise for qualitatively discriminating size groups.
|
Interannual comparisons of average pollock density
To compare Pribilof Islands densities of age-0 pollock with the eventual size of the adult cohort, we set up a simple ratio table of the NOAA stock assessment estimates for abundance at recruitment (age-3) and the average acoustic age-0 densities by transect, in each year, expressed as a fraction of the 1996 age-3 abundance and age-0 density, respectively (Table 2). Average age-0 pollock densities (m–3) were estimated from the survey echo-integrations by dividing area backscatter sa (m2 m–2) of pollock pixels by pixel depth (m) and mean backscattering cross-section
bs (m2) of age-0 pollock, i.e. number m–3 = sa depth–1 bs–1. As Pribilof surveys were not all taken over the same days of the year, age-0 densities for this table were standardized to 22 September (the latest date of any transect) by forward-calculating from the date of each survey an instantaneous mortality rate of 0.055 day–1 (the average instantaneous mortality rate from Swartzman et al., 2005). The 1996 eastern Bering pollock year class was by far the most abundant of the study years 1994–1999, as evidenced by pre-recruit surveys (Swartzman et al., 2002) and subsequent stock assessments (Ianelli et al., 2003). Using 1996 as a benchmark, the relative success of the other year classes could be gauged. As body size influences the viability of juvenile fish (Sogard, 1997), we estimated length distributions of the age-0 pollock from anchovy trawls taken during the surveys. Lengths were also standardized to 22 September, using an average growth estimate of 0.5 mm day–1 (Swartzman et al., 2005). Significance of length difference among years was tested by nested ANOVA (anchovy trawls nested within years) and Tukey's method for pairwise a posteriori comparisons. Ciannelli et al. (2002b) found 80 mm body length to represent a threshold beyond which energy content (kJ g–1) of age-0 pollock no longer increases with size. Therefore, we calculated the proportion of age-0 pollock 80 mm by the 22 September standardization for each survey.
|
Spatial distributions of age-0 pollock density
Spatial distributions of age-0 pollock acoustic densities1 were analysed as a function of (i) the spatial distributions of euphausiid acoustic densities, (ii) the ratio of potential groundfish predators relative to age-0 pollock, and (iii) the distribution of seabirds on the water. Details of these three effects are described below. Analyses were carried out per individual transect pass, using non-parametric regression generalized additive models (GAM; Hastie and Tibshirani, 1990) with spline smoothing (s) at 4 degrees of freedom (1 d.f. would imply a linear fit). For sampling units, each transect pass was subdivided at a spatial scale of 5-m depth x 1000-m distance bins (cf. Figure 2c). In GAM, covariates are assumed to affect the dependent variable through unspecified (not necessarily linear) additive functions or smooths (denoted by s(covariate)), and data can come from any distribution in the exponential family (Swartzman et al., 1992). Here, we assumed the age-0 pollock densities to come from the normal distribution that was verified by Kolmogorov–Smirnov goodness-of-fit tests.
For each transect pass we retained the best fitting GAM model among possible combinations of covariates s(euphausiid density), s(predator/prey ratio), s(seabird effect), and si(pairwise interactions). Covariates were ranked by their pseudo-R2 values (1 minus the ratio of model deviance over null deviance; Swartzman et al., 1992) and added to the model by forward selection. The second-highest pseudo-R2 covariate was added to the highest pseudo-R2 covariate if it improved the model as evaluated by an approximate F-test (Hastie and Tibshirani, 1990):
|
|
, then the next covariate was added and the F-test repeated between the 2- and 3-covariate models. If 
, the calculation was performed between the highest and third-highest covariate (the third highest might add significance even though the second highest did not, by being more orthogonal to the highest). This was repeated until each covariate was either included or rejected. As bird effects at different depths in the same (1000 m wide) columns were autocorrelated, degrees of freedom of any model including birds were adjusted before the F-test, to allow only one degree of freedom per column per bird effect. GAM plots for an example transect are shown in Figure 3a.
|
Euphausiid density
As for age-0 pollock, euphausiid densities were estimated by dividing echo-integration backscatter of pixels identified as euphausiids by pixel depth and backscattering cross-section (
bs). A mean euphausiid bs value of 10–7.78 m2 (Coyle and Pinchuk, 2002) was used across all surveys. In addition to the whole-transect analyses, relationships between age-0 pollock and (only) euphausiid densities were analysed by habitat region within transect passes. Echo-integration data were in this case averaged into 250-m horizontal (full water column) bins (Figure 2d), as utilized by Swartzman et al. (1999a, b). Density calculations were normalized for bottom depth but restricted to the upper 130 m of the water column (only transect D, basin goes deeper anyway) to avoid biasing the analysis by including depths the juvenile pollock do not habitually enter. GAM results for the habitat regions of one example transect are plotted in Figure 3b, and the directional pseudo-R2 values (R2 x the +/– slope of the GAM) for all habitat regions in all surveys are summarized in Figure 4.
|
Groundfish predators
Direct groundfish predator information from trawls would be optimal for this study. However, the Pribilof surveys took few, if any, trawls suitable for capturing larger, predator-sized fish (i.e. trawls taken with an Aleutian wing trawl or a bottom trawl), and publication of such data is otherwise sparse for this area. Lang et al. (2000) analysed arrowtooth flounder stomachs in 1995, and adult pollock stomachs in 1996, on Pribilof transect A (Figure 1) only. Ciannelli et al. (2002a, cf. their Table 2) summarized bottom trawl catches as combined densities (kg ha–1) of arrowtooth flounder and adult pollock between 1994 and 1997.
To obtain more complete coverage of groundfish predator data, we utilized the distributions of acoustic TS values corresponding to size–age categories that potentially prey on age-0 pollock. The occurrence of TS of a given strength on a transect is not directly quantifiable as an index of the number of large fish present, because the echosounder only identifies TS to the extent that they are demarcated from the overall backscatter. Fish in dense shoals will not be identified individually. But over a given transect interval, the approximation can be made that demarcation of TS is independent of the strength of the TS, e.g. small fish are equally likely as large fish to show up as singles. Gauthier and Rose (2001) noted that at frequencies used in fisheries acoustics (10–200 kHz), fish with swimbladders can be considered point scatterers, and are unlikely to significantly influence reverberation volume as a function of size differences. Under the equally likely assumption, a ratio of large over small TS on a transect interval thus estimates the density ratio of predators to which juvenile fish are exposed (i.e. as an index of predators per prey). Behaviourally, TS detection may favour large fish, which maintain greater separation distance from other fish (MacLennan and Menz, 1996). However, size-specific behaviour would be consistent across the extent of transects. The distribution of TS ratio values can therefore be utilized as a relative measure of spatial relationship between groundfish predators and prey when comparing multiple observations within a single transect.
In this study, spatially explicit predator/prey ratios of age-2+ over age-0 TS were calculated at the resolution of the 5-m depth x 1000-m distance bins (Figure 2c). Age-1 TS were not included in the analysis. Age-1 pollock prey on age-0 pollock to some extent (Dwyer et al., 1987), but have a much lower quantitative impact (Lang et al., 2000). Also, distributions of age-1 pollock are likely to be determined by their own efforts to avoid the larger adults. Before computing this TS predator/prey ratio (hereafter PPR) in each bin, numbers of TS were smoothed to eliminate zero counts of age-0 pollock in a bin, which would cause infinite ratios. Smoothing was calculated as the average TS count of neighbouring bins (and self) weighted by the inverse distance of separation between bins. This PPR method has been validated by correlation with trawl samples in several eastern Bering Sea surveys (Winter, 2005). Target-strength data were available for all four transects in 1995, 1996, 1997, and 1999, and for transect A in 1994.
Seabirds
Seabird counts had been recorded by species along several daylight transects during each survey (Logerwell et al., 1998; Swartzman and Hunt, 2000). To model seabirds' potential influence on age-0 pollock distribution through the water column, we took into account species' different feeding strategies: kittiwakes are surface feeders (Decker et al., 1995), fulmars feed at or near the surface (Hunt et al., 1982), shearwaters dive to at least 40 m (Lovvorn et al., 2001), auklets to 35–40 m (Bedard, 1969), puffins to 
50 m (Burger and Simpson, 1986), and murres to 210 m (Croll et al., 1992). Depth effects of the feeding strategies were calculated as the inverse linear function:
|
|
For example, the presence of a murre would be considered to influence age-0 pollock at 40-m depth 81% as much as it influences age-0 pollock at the surface: 1 – (40/210) = 0.81. The depth effect was then multiplied by the number of birds (of that species) superjacent, and summed for all birds on a given transect interval.
Latitude distribution
As transects run primarily north–south (Figure 1), a metric of along-transect age-0 pollock distribution was recorded as the latitudinal quantiles of acoustic density. On each transect pass we calculated the latitudes representing the cumulative 10%, 20%, ..., etc. of depth-normalized age-0 pollock acoustic biomass. For inter-year comparability, these calculations were restricted to the latitude range common to all passes of each transect. Average latitude quantiles per year and transect are plotted in Figure 5.
|
|
Results |
|---|
|
TopIntroductionMethodsResultsDiscussionSummaryReferences |
|---|
Interannual comparisons of average pollock density
Inspection of Table 2 shows that the 1994 year class "lost ground" relative to 1996. At the age-0 stage in the Pribilof Islands nursery, 1994 age-0 fish had 93% of the average acoustically derived density of the 1996 benchmark, but by the recruitment age of 3, that proportion had fallen to 27%. Transect D appears to follow different trends from the other transects, being the only transect with lower age-0 pollock densities in 1994 than in 1996. All other year classes gained against 1996 between ages 0 and 3, to varying degrees. Except for 1994, year classes also tended to conserve their rankings from age-0 to age-3, i.e. 1996 > 1999 > 1995 > 1998 > 1997. Age-0 pollock lengths standardized to 22 September were smallest in 1994 and had the lowest proportion
80 mm (Table 3). On the actual survey dates, 1994 lengths were intermediate between the other years (Table 3), so the 1994 in situ spatial relationships to predators and prey cannot be considered influenced by smaller length of the fish.
|
Euphausiid density
Spatial relationships between age-0 pollock and euphausiids were predominantly positive, at both the scale of whole transects (5 x 1000 m bins) and of habitat regions within transects (250 m x full water column bins). Explanatory powers of the relationships (R2) were higher for the habitat region analyses than for whole-transect analyses (Tables 4–7, Figures 3 and 4). That may indicate in part that full water column bins are more suitable for measuring the relationship than depth x distance strata bins. We conjecture that, as age-0 pollock and euphausiids undertake diel migrations (Swartzman et al., 2002), vertical separation on the shallow eastern Bering shelf may not be an important factor. Analysis of directional R2 values by habitat regions showed that 1994 had the smallest percentage of spatially positive relationships, followed by 1996, 1995 and 1999, 1997, and 1998 (one-way ANOVA, p < 0.001; Figure 4). The outcome suggests that despite 1994 recording the second highest concentration of age-0 pollock in September (Table 2), the fish were relatively unsuccessful at tracking their zooplankton prey.
|
|
|
|
Groundfish predators
Groundfish PPR main effects were included in whole-transect GAM models on 41 of 63 transect passes (Tables 4–7). Explanatory power contributed by the PPR was almost always less than that of euphausiid density (A3 95, the example transect pass shown in Figures 2 and 3, is exceptional in this regard), and PPR as a sole covariate was never the best model. Statistically significant PPR trends were mostly negative, by a proportion of 2:1 or more on all transects (Tables 4–7). As predators and prey are both mobile, a negative trend is interpreted as age-0 pollock evading high concentrations of larger groundfish (i.e. it would be less plausibly a case of all age-0 pollock having been eaten at such locations, because predators and prey would then redistribute themselves). Occurrence of significantly negative vs. positive or non-significant (n.s.) PPR trends did not show significant differentiation with respect to average values of age-0 pollock, euphausiid, bird, PPR, or daylight parameters (Tables 4–7; logistic regression; Venables and Ripley, 1997). Interactions between PPR and euphausiid density were significant on 33 transect passes. Most of these interactions tended to show a positive effect on age-0 pollock density, with high values of both covariates (Tables 4–7, Figure 3a).
Seabirds
Seabird main effects contributed to the best GAM model on 18 of the 35 transect passes for which bird counts were recorded: 10 positive, 6 negative, and 2 bimodal GAM trends. Seabird interaction effects were mostly not significant (Tables 4–7). Among all transect passes, seabird trends showed no significant relationship with the age-0 pollock, euphausiid, bird, PPR, or daylight parameters (Tables 4–7; logistic regression; Venables and Ripley, 1997).
Latitude distribution
Average latitudinal distributions of age-0 pollock varied among years on the order of 0.04° (4400 m). On the four transects, 1994 always had either the most northerly or most southerly centroid (50% quantile; Figure 5). For each transect, the absolute difference between each year's centroid and the mean of all years (i.e. the centroid latitude anomaly) was calculated. These anomalies are plotted in Figure 6 against the age-3 EBS abundance fractions taken from Table 2. The highly significant (p < 0.005, r2 = 0.42) second-order regression in Figure 6 indicates an inverse relationship between centroid latitude anomaly and recruitment-age year-class strength.
|
|
Discussion |
|---|
|
TopIntroductionMethodsResultsDiscussionSummaryReferences |
|---|
Age-0 juvenile abundance from nursery and spawning surveys predicts year-class strength for gadoid populations, including cod (G. morhua) along the Norwegian Skagerrak (Tveite, 1984), in the Barents Sea (Helle et al., 2000), and off Iceland (Begg and Marteeinsdottir, 2002), and cod and haddock (Melanogrammus aeglefinus) off Nova Scotia (Campana et al., 1989). A predictive correlation from age-0 to age-3 was not detected in Newfoundland coastal cod stocks (Ings et al., 1997), but these authors suggested that recruitment from other nursery areas away from the coast may have concealed a pre-recruit signal. In the eastern Bering Sea, the Pribilof Islands nursery is a primary source of the pollock stock (Swartzman et al., 2005). Initial comparisons between Pribilof age-0 and EBS age-3 indices demonstrated a strong link for some years (Schumacher and Macklin, 2004; Table 2). As in many studies of marine recruitment, an absence of linkage in all years motivates investigation.
A question in particular is why 1994 fared so poorly (Table 2). One difference between 1994 and later years was found in the acoustically determined spatial proximity relationships of age-0 pollock and euphausiid density distributions. Spatial proximity between gadoids and their prey has been inferred as an indicator of feeding success for Pacific hake Merluccius productus (Mackas et al., 1997; Swartzman, 2001), Cape hake M. capensis and M. paradoxus (Drapeau et al., 2004), Atlantic cod (Rose and Leggett, 1990), and previously for juvenile (Swartzman et al., 1999a, b) and large (Miyashita et al., 2004) pollock. The acoustic density relationships among age-0 pollock and euphausiids in this study (Table 4) are consistent with consumption data reported by Ciannelli et al. (2004; cf. their Figure 3): 3.5% euphausiid in the age-0 pollock diet in 1994, 15% in 1995, 28% in 1996, and 78% in 1997. Age-0 pollock around the Pribilof Islands in 1994 hatched later and were smaller by September than in the subsequent five years (Brodeur et al., 2002; Table 3), and euphausiid selection is known to increase with size (Merati and Brodeur, 1996; Schabetsberger et al., 2000). Nevertheless, the 48.4 mm mean standard length in 1994 (Table 3) is well within the size range of age-0 pollock that consumed significant proportions of euphausiids in 1996 and 1997 (Winter et al., 2005; cf. their Figure 6).
The corroboration between consumption data (Ciannelli et al., 2004) and spatial relationships observed during the acoustic surveys suggests that in 1994, age-0 pollock largely omitted euphausiids from their diet as a function of decreased encounter opportunities. Figures 5 and 6 present a potentially relevant scenario to this effect: age-0 pollock in 1994 occurred on average either farther offshore (transects A and B, St Paul Island) or closer inshore (transects C and D, St George Island) than in any other year. However, neither cause nor effect for this pattern is obvious. Oceanographic studies (Schumacher and Stabeno, 1998; Kowalik and Stabeno, 1999) have revealed a clockwise flow circulation around the Pribilof Islands which, together with the dominant tidal motion, would not have the net result of transporting juvenile fish inshore to St George and simultaneously away from St Paul. Ciannelli et al. (2002a) hypothesized that in a cold year such as 1995, groundfish predators foray into shallower water, pushing age-0 pollock closer inshore. According to our analysis, 1995 did have consistently negative PPR GAMs (Tables 4–7), but was unexceptional with respect to the inshore distribution of age-0 pollock. 1994 and 1997 averaged the highest and second highest latitude centroid deviations, respectively (Figure 6), but had the weakest and strongest positive spatial relationships between age-0 pollock and euphausiid distributions. There is therefore a strong statistical correlation between latitudinal deviation and year-class strength (Figure 6), but an absence of connection to any one proximate parameter. Brodeur et al. (2002) concluded that juvenile pollock did not benefit more from some habitats than others around the Pribilof Islands, and observed that "In a physically dynamic system such as the Pribilof Islands, age-0 pollock may need to continuously search for optimal conditions of high prey availability and low predation pressure."
We propose therefore that the advantage of distributions associated with a mid-range centroid may generally be that these are strategically centralized, in the vicinity of the Pribilof Islands nursery, for the juvenile fish to search optimally both shoreward and offshore in the face of varying environmental challenges. In 1994, all 10% density quantiles were monotonically biased either northward (transects A and D) or southward (transects B and C). The other five years were more uneven with some northward- and some southward-biased quantiles on at least half the transects (Figure 5). This pattern is further evidence that in 1994, movement or transport of the age-0 pollock was relatively restricted across the entire latitude extent of transects. Such restriction perhaps supports the Nursery Size Hypothesis described by Abella et al. (2005) for European hake (M. merluccius), whereby recruitment of cohorts correlates positively with surface area of the nursery grounds.
Precedent studies have examined large-scale distribution patterns of juvenile pollock in the eastern Bering Sea and postulated that recruitment success varies with the extent of temperature-mediated spatial overlap between juvenile and adult pollock (Francis and Bailey, 1983; Wyllie-Echeverria and Wooster, 1998; Wespestad et al., 2000). Overlap increases in cold years, when reduced current transport keeps juveniles on the outer shelf, resulting in increased cannibalism (Wespestad et al., 2000). Lang et al. (2003) reported that from 1993 through 1996, the highest annual level of predation on age-0 pollock (including cannibalism) took place in 1994. However, the data of Lang et al. (2003, their Figure 8) do not verify that cannibalism alone would explain the weak year class of 1994 compared with 1995 and 1996 (Table 2). The PPR data available from 1994 indicate significant negative correlation between groundfish predators and age-0 pollock, but not more so than in other years (Table 4). Among all surveys (1994–1999), PPR relationships showed no overall covariance trend with age-0 pollock, euphausiid, or TS density.
Results of this study provide indications that groundfish and seabird predators influenced distributions of age-0 pollock on local scales around the Pribilof Islands, but without incurring significant differentiation among years. Potential seabird influences may have been mitigated by the timing of the surveys: by September, seabirds are less bound to their nesting colonies and can forage farther from the Pribilof Islands than earlier in the season (G. Hunt, University of California, pers. comm.). Winter (2005) found also that frontal zone boundaries and thermoclines showed effects on individual transects, without overall relationship to pollock distribution or year-class strength. It is plausible that all these factors exercised some degree of indirect control on the prevalent spatial correlations between age-0 pollock and euphausiids. Vertical migration of euphausiids is commonly reported to result from predation pressure (Zaret and Suffern, 1976; Iwasa, 1982; DeRobertis et al., 2000). Euphausiid response to thermoclines appears to be species-specific and based on physiological characteristics (Brinton, 1967). Brinton (1967) and Youngbluth (1976), working in the California Current system, found one species each of Euphausia and Thysanoessa that did not migrate above the thermocline at night, in contrast to all other species of the same genera they encountered in their samples. Brinton (1967) further observed that for those species which did migrate, the thermocline would become the lower limit of the species' night-time range. Swartzman et al. (2002) suggested that euphausiids near the Pribilof Islands can locate the thermocline using density cues. Seabirds as well as adult groundfish feed heavily on euphausiids (Bedard, 1969; Lang et al., 2003), and are therefore competitors to as well as predators on age-0 pollock. In fact, age-0 pollock may benefit from a refuge effect of predators selecting euphausiids. As Leggett and DeBlois (1994) remarked for the parallel question of larval fish survival, an "either/or approach" to food vs. predation as the primary determinant of recruitment variability is likely to be inadequate. Given this caveat, the analyses of this study do show spatial correlation between age-0 pollock and euphausiid density distributions to be a potentially useful indicator for year-class strength. The statistically significant GAM relationships often had relatively low R2 values (Tables 4–7, Figure 4), consistent with findings that a large (in some years majority) proportion of age-0 pollock by September do not feed primarily on euphausiids (Brodeur et al., 2000; Schabetsberger et al., 2003; Ciannelli et al., 2004; Winter et al., 2005). We hypothesize that those age-0 pollock which do feed on euphausiids carry the strength of the cohort.
Euphausiids have the highest caloric density of age-0 pollock's usual zooplankton prey (Davis et al., 1998), and Sogard and Olla (2000) determined that greater energy storage enhances age-0 pollock survival under simulated overwintering conditions. Smith et al. (1986) observed that captive juvenile pollock fed every other day, rather than daily, tended to compensate by eating larger meals. If age-0 pollock overwinter by metabolic adaptation and occasional feeding (Ciannelli et al., 2002b), then survival should further be enhanced by access to large prey. Therefore, it is plausible that age-0 pollock cohorts that fail to make the transition to a predominantly euphausiid diet by September will experience greater mortality through their first winter, and recruit as a weak year class. Conversely, a sparse age-0 cohort such as 1997 may still produce a modestly successful year class when the fish achieve high euphausiid consumption. Winter et al. (2005) found age-0 pollock near the Pribilof Islands to have significantly higher average condition (W/L3) in September 1997 than in September 1996, while Brodeur et al. (2000) did not find a significant difference in condition factor among the years 1994, 1995, and 1996. As age-0 pollock were smaller in 1994 (Brodeur et al., 2002; Table 3), they would still have been suited to attain satiation from smaller (and less energy-rich) prey such as the copepods on which they were predominantly feeding (Ciannelli et al., 2004). Schabetsberger et al. (2003) and Winter et al. (2005) concluded from dietary analyses that smaller age-0 pollock had more food in their guts relative to body weight (i.e. gut fullness) than larger individuals when they consumed larger proportions of small prey items. Yamamura et al. (2002) found that small pollock off the coast of Hokkaido Island (northern Japan) switched from a predominantly copepod diet in spring to a predominantly euphausiid diet by autumn, and attained maximum body condition in August, followed by gradual decline through winter.
While sufficient body condition of age-0 pollock may be a necessary factor for their overwinter survival potential (Sogard and Olla, 2000), it appears not to be a reliable predictor, as early as mid-September, for comparing interannual differences of year-class strength. Experimental studies on age-0 Atlantic cod found body size but not food ration to significantly influence overwinter survival (Brown et al., 1989; Gotceitas et al., 1999). In this study, the influence of size is unclear. Multiplying the age-0 density fractions from Table 2 by the 80 mm proportions from Table 3, 27.4% as many 80 mm age-0 pollock were obtained by 22 September in 1994 as in 1996: 0.93/1 x 0.005/0.017 = 27.4%. That estimate is notably close to the age-3 recruitment abundance ratio between the two years (Table 2), and suggests that the relative growth retardation in 1994 might have resulted in fish having too little feeding history on a euphausiid diet before winter, even if they were big enough to take euphausiids by the time the survey was run. However, equivalent growth calculations for the other years were not consistent with subsequent recruitment. For example, 1997 had 2.9% the number of 80 mm age-0 pollock as 1996 (0.06/1 x 0.008/0.017), yet 45% the cohort strength at recruitment. Possibly, the effect of body size would have become more evident in surveys with older fish.
|
Summary |
|---|
|
TopIntroductionMethodsResultsDiscussionSummaryReferences |
|---|
Three measures of acoustic survey data were identified that together yield a potentially practical index for pollock year-class strength (Table 8): (i) age-0 density; (ii) spatial correlation with euphausiids; and (iii) centricity of the along-transect distributions.
|
Inspection of Table 8 suggests that the measures are additive insofar as year-class strength correlates with the sum of their "scores". As the scores at this level are derived from relatively limited survey data, we restrict them to qualitative interpretation. This index of measures is predicated on conclusions (Swartzman et al., 2002, 2005, and references therein) that the Pribilof Islands area is highly representative for the eastern Bering Sea. Other areas of the eastern Bering Sea, not sampled or not identified, will contribute to each year class, and further validation of any forecast index should include expanded surveys of additional possible pollock nurseries. Expanded time-series should be pursued as well, because the range of contrast in pollock recruitment among the six years of this study was not exceptionally high for the eastern Bering Sea (cf. Ianelli et al., 2003). It is nonetheless compelling to note that the strong age-0 Pribilof cohort in 1996 led to a strong recruitment-age year class, whereas the almost equally strong age-0 Pribilof cohort in 1994 did not. We propose that this index is suitable for measuring short-term recruitment variation, which may be influenced primarily by episodic events. For example, Bond and Overland (2005) correlated strong pollock recruitment on a short-term scale with the occurrence of strong storms in the early summers of age-0 years (including 1996), and attributed the effect to greater mixing of nutrients into the euphotic zone, leading to enhanced prey concentration available to pollock larvae. On a larger scale of oceanographic processes, the Oscillating Control Hypothesis (Hunt et al., 2002; Hunt and Stabeno, 2002) predicts that cold regimes limit zooplankton production and hence prey for age-0 pollock (bottom-up regulation). In warm regimes, zooplankton is plentiful and predation controls the survival of age-0 pollock (top-down). As predator abundance is itself the result of past recruitment, age-0 pollock are in any year constrained, at the minimum, by some variable combination of the two control mechanisms of food supply and predation. An index that integrates several measures may be the effective approach to forecasting cohort success.
|
Acknowledgements |
|---|
We thank M. Brown, A. Cossio, D. Demer, and M. Wilson for providing the acoustic survey data and J. Horne for calculating the Kirchhoff ray-mode model. We also thank A. Bertrand, K. Coyle, D. Gunderson, J. Horne, and two anonymous referees for their critical reading of and suggestions for improvements to the manuscript. The research was sponsored by NOAA's Coastal Ocean Program through Southeast Bering Sea Carrying Capacity, and is contribution of FOCI-S577 to Fisheries-Oceanography Coordinated Investigations. The first author was additionally supported by the Mason Keeler Endowment at the University of Washington.
|
Footnotes |
|---|
1 In this paper, spatial distributions of density always refer to vertical density, i.e. the "curtain" underneath each transect as in Figure 2, rather than a "carpet" covering a surface area. The use of vertical density allows for normalization of different depths of water.
|
References |
|---|
|
TopIntroductionMethodsResultsDiscussionSummaryReferences |
|---|
-
Abella A., Serena F., Ria M. (2005) Distributional response to variations in abundance over spatial and temporal scales for juveniles of European hake (Merluccius merluccius) in the western Mediterranean Sea. Fisheries Research 71:295–310.[CrossRef][Web of Science]
Bailey K.M. (1989) Interaction between the vertical distribution of juvenile walleye pollock Theragra chalcogramma in the eastern Bering Sea, and cannibalism. Marine Ecology Progress Series 53:205–213.[Web of Science]
Bedard J. (1969) Feeding of the least, crested, and parakeet auklets around St Lawrence Island, Alaska. Canadian Journal of Zoology 47:1025–1050.
Begg G.A. and Marteeinsdottir G. (2002) Environmental and stock effects on spawning origins and recruitment of cod Gadus morhua. Marine Ecology Progress Series 229:263–277.[Web of Science]
Blaxter J.H.S. and Batty R.S. (1990) Swimbladder "behaviour" and target strength. Rapports et Procès-verbaux des Réunions du Conseil International pour l'Exploration de la Mer 189:233–244.
Bond N.A. and Overland J.E. (2005) The importance of episodic weather events to the ecosystem of the Bering Sea shelf. Fisheries Oceanography 14:97–111.[CrossRef][Web of Science]
Bradford M.J. (1992) Precision of recruitment predictions from early life stages of marine fishes. Fishery Bulletin US 90:439–453.
Brinton E. (1967) Vertical migration and avoidance capability of euphausiids in the California Current. Limnology and Oceanography 12:451–483.[Web of Science]
Brodeur R.D. (1998) Juvenile pollock studies. In Macklin S.A. (Ed.). Bering Sea FOCI 1991–1997 Final Report, Part 3A(NITS, Springfield, VA) pp. 103–120 NOAA ERL Special Report 167 pp.
Brodeur R.D., Livingston P.A., Loughlin T.R., Hollowed A.B. (1996) Introduction; Ecology of juvenile walleye pollock, Theragra chalcogramma. NOAA Technical Report NMFS 126:1–2.
Brodeur R.D., Wilson M.T., Ciannelli L. (2000) Spatial and temporal variability in feeding and condition of age-0 walleye pollock (Theragra chalcogramma) in frontal regions of the Bering Sea. ICES Journal of Marine Science 57:256–264.
Brodeur R.D., Wilson M.T., Ciannelli L., Doyle M., Napp J.M. (2002) Interannual and regional variability in distribution and ecology of juvenile pollock and their prey in frontal systems of the Bering Sea. Deep-Sea Research II 49:6051–6067.[CrossRef]
Brown J.A., Pepin P., Methven D.A., Somerton D.C. (1989) The feeding, growth and behaviour of juvenile cod, Gadus morhua L, in cold environments. Journal of Fish Biology 35:373–380.[CrossRef][Web of Science]
Burger A.E. and Simpson M. (1986) Diving depths of Atlantic puffins and common murres. Auk 103:828–830.[Web of Science]
Campana S.E., Frank K.T., Hurley P.C.F., Koeller P.A., Page F.H., Smith P.C. (1989) Survival and abundance of young Atlantic cod (Gadus morhua) and haddock (Melanogrammus aeglefinus) as indicators of year-class strength. Canadian Journal of Fisheries and Aquatic Sciences 46:Suppl. 1, 171–182.
Ciannelli L., Brodeur R.D., Napp J.M. (2004) Foraging impact on zooplankton by age-0 walleye pollock (Theragra chalcogramma) around a front in the southeast Bering Sea. Marine Biology 144:515–526.[CrossRef]
Ciannelli L., Brodeur R.D., Swartzman G.L., Salo S. (2002) Physical and biological factors influencing the spatial distribution of age-0 walleye pollock (Theragra chalcogramma) around the Pribilof Islands, Bering Sea. Deep-Sea Research II 49:6109–6126.[CrossRef]
Ciannelli L., Paul A.J., Brodeur R.D. (2002) Regional, interannual, and size-related variation of age 0 year walleye pollock whole body energy content around the Pribilof Islands, Bering Sea. Journal of Fish Biology 60:1267–1279.[CrossRef][Web of Science]
Coyle K.O. and Cooney R.T. (1993) Water column sound scattering and hydrography around the Pribilof Islands, Bering Sea. Continental Shelf Research 13:803–827.[CrossRef][Web of Science]
Coyle K.O. and Pinchuk A.I. (2002) The abundance and distribution of euphausiids and zero-age pollock on the inner shelf of the southeast Bering Sea near the Inner Front in 1997–1999. Deep-Sea Research II 49:6009–6030.[CrossRef]
Croll D.A., Gaston A.J., Burger A.E., Konnoff D. (1992) Foraging behavior and physiological adaptation for diving in thick-billed murres. Ecology 73:344–356.[CrossRef][Web of Science]
Davis N., Myers K.W., Ishida Y. (1998) Caloric value of high-seas salmon prey organisms and simulated salmon ocean growth and prey consumption. North Pacific Anadromous Fish Commission Bulletin 1:146–162.
Decker M.B. and Hunt G.L. (1996) Foraging by murres (Uria spp.) at tidal fronts surrounding the Pribilof Islands, Alaska, USA. Marine Ecology Progress Series 139:1–10.[Web of Science]
Decker M.B., Hunt G.L., Byrd G.V. (1995) The relationships among sea-surface temperature, the abundance of juvenile walleye pollock (Theragra chalcogramma), and the reproductive performance and diets of seabirds at the Pribilof Islands, southeastern Bering Sea. Canadian Special Publication in Fisheries and Aquatic Sciences 121:425–437.
DeRobertis A., Jaffe J.S., Ohman M.D. (2000) Size-dependent visual predation risk and the timing of vertical migration in zooplankton. Limnology and Oceanography 45:1838–1844.[Web of Science]
Drapeau L., Pecquerie L., Fréon P., Shannon L.J. (2004) Quantification and representation of potential spatial interactions in the southern Benguela ecosystem. African Journal of Marine Science 26:141–159.[Web of Science]
Dwyer D.A., Bailey K.M., Livingston P.A. (1987) Feeding habits and daily ration of walleye pollock (Theragra chalcogramma) in the eastern Bering Sea, with special reference to cannibalism. Canadian Journal of Fisheries and Aquatic Sciences 44:1972–1984.
Foote K.G. (1983) Linearity of fisheries acoustics, with addition theorems. Journal of the Acoustical Society of America 73:1932–1940.[CrossRef][Web of Science]
Foote K.G. (1987) Fish target strengths for use in echo integrator surveys. Journal of the Acoustical Society of America 82:981–987.[CrossRef][Web of Science]
Foote K.G. and Traynor J.J. (1988) Comparisons of walleye pollock target strength estimates determined from in situ measurements and calculations based on swimbladder form. Journal of the Acoustical Society of America 83:9–17.[CrossRef][Web of Science]
Francis R.I.C. and Bailey K.M. (1983) Factors affecting recruitment of selected gadoids in the northeast Pacific and east Bering Sea. In Wooster W. (Ed.). From Year to Year. Interannual Variability of the Environment and Fisheries of the Gulf of Alaska and the Eastern Bering Sea(Washington Sea Grant, Seattle, WA) pp. 35–60 208 pp.
Gauthier S. and Rose G.A. (2001) Diagnostic tools for unbiased in situ target strength estimation. Canadian Journal of Fisheries and Aquatic Sciences 58:2149–2155.
Gotceitas V., Methven D.A., Fraser S., Brown J.A. (1999) Effects of body size and food ration on over-winter survival and growth of age-0 Atlantic cod, Gadus morhua. Environmental Biology of Fishes 54:413–420.[CrossRef][Web of Science]
Haralick R. and Shapiro L. (1992) Computer and Robot Vision. (Addison-Wesley, Reading, MA) vol. 1: 672 pp.
Hart J.L. (1973) Pacific fishes of Canada. Bulletin of the Fisheries Research Board of Canada 180: 740 pp.
Hastie T.J. and Tibshirani R.J. (1990) Generalized Additive Models(Chapman and Hall/CRC, Boca Raton) 335 pp.
Hazen E.L. and Horne J.K. (2004) Comparing the modelled and measured target-strength variability of walleye pollock, Theragra chalcogramma. ICES Journal of Marine Science 61:363–377.
Helle K., Bogstad B., Marshall C.T., Michalsen K., Ottersen G., Pennington M. (2000) An evaluation of recruitment indices for Arcto-Norwegian cod (Gadus morhua L.). Fisheries Research 48:55–67.[CrossRef][Web of Science]
Hennemuth R.C., Palmer J.E., Brown B.E. (1980) A statistical description of recruitment in eighteen selected fish stocks. Journal of Northwest Atlantic Fishery Science 1:101–111.
Horne J. (2003) The influence of ontogeny, physiology, and behaviour on the target strength of walleye pollock (Theragra chalcogramma). ICES Journal of Marine Science 60:1063–1074.
Hunt G. L., Epply Z., Burgeson B., Squibb R. (1982) Reproductive ecology, foods and foraging areas of seabirds nesting on the Pribilof Islands, 1975–1979. Environmental Assessment of the Alaskan Continental Shelf Final reports of principal investigators. NOAA, Boulder, CO. Environmental Research Laboratories Biological Studies, 12. 258 pp.
Hunt G.L. and Stabeno P.J. (2002) Climate change and control of energy flow in the southeastern Bering Sea. Progress in Oceanography 55:5–22.[CrossRef][Web of Science]
Hunt G.L., Stabeno P.J., Walters G.E., Sinclair E.H., Brodeur R.D., Napp J.M., Bond N.A. (2002) Climate change and control of the southeastern Bering Sea pelagic ecosystem. Deep-Sea Research II 49:5821–5853.[CrossRef]
Ianelli J. (2005) Assessment and fisheries management of eastern Bering Sea walleye pollock: is sustainability luck? Bulletin of Marine Science 76:321–335.[Web of Science]
Ianelli J. N., Barbeaux S., Walters G., Williamson N. (2003) Eastern Bering Sea walleye pollock stock assessment. Stock assessment and fishery evaluation report. Alaska Fisheries Science Center, NOAA, 7600 Sand Point Way NE, Seattle, WA 98115. 88 pp.
Ings D.W., Schneider D.C., Methven D.A. (1997) Retrospective analysis of year-class rank in Atlantic cod (Gadus morhua L.). Canadian Journal of Fisheries and Aquatic Sciences 54:Suppl. 1, 25–29.
Iwasa Y. (1982) Vertical migration of zooplankton: a game between predator and prey. American Naturalist 120:171–180.[CrossRef][Web of Science]
Kang M., Furusawa M., Miyashita K. (2002) Effective and accurate use of difference in mean volume backscattering strength to identify fish and plankton. ICES Journal of Marine Science 59:794–804.
Kinder T.H., Hunt G.L., Schneider D., Schumacher J.D. (1983) Correlations between seabirds and oceanic fronts around the Pribilof Islands. Estuarine, Coastal, and Shelf Science 16:309–319.[CrossRef]
Kowalik Z. and Stabeno P. (1999) Trapped motion around the Pribilof Islands in the Bering Sea. Journal of Geophysical Research 104:25667–25684.[CrossRef]
Lang G.M., Brodeur R.D., Napp J.M., Schabetsberger R. (2000) Variation in groundfish predation on juvenile walleye pollock relative to hydrographic structure near the Pribilof Islands, Alaska. ICES Journal of Marine Science 57:265–271.
Lang G. M., Derrah C. W., Livingston P. A. (2003) Groundfish food habits and predation on commercially important prey species in the eastern Bering Sea from 1993 through 1996. AFSC Processed Report 2003–04. NOAA, US Department of Commerce. 351 pp.
Leggett W.C. and DeBlois E. (1994) Recruitment in marine fishes: is it regulated by starvation and predation in the egg and larval stages? Netherlands Journal of Sea Research 32:119–134.
Logerwell E.A., Hewitt R., Demer D.A. (1998) Scale-dependent spatial variance patterns and correlations of seabirds and prey in the southeastern Bering Sea as revealed by spectral analysis. Ecography 21:212–223.
Love R. (1971) Measurements of fish target strength: a review. Fishery Bulletin US 69:703–715.
Lovvorn J.R., Baduini C.L., Hunt G.L. (2001) Modeling underwater visual and filter feeding by planktivorous shearwaters in unusual sea conditions. Ecology 82:2342–2356.[CrossRef][Web of Science]
Mackas D.L., Kieser R., Saunders M., Yelland D.R., Brown R.M., Moore D.F. (1997) Aggregation of euphausiids and Pacific hake (Merluccius productus) along the outer continental shelf off Vancouver Island. Canadian Journal of Fisheries and Aquatic Sciences 54:129–136.
Macklin S.A., Hunt G.L., Overland J.E. (2002) Collaborative research on the pelagic ecosystem of the southeastern Bering Sea shelf. Deep-Sea Research II 49:5813–5819.[CrossRef]
MacLennan D.N., Fernandes P.G., Dalen J. (2002) A consistent approach to definitions and symbols in fisheries acoustics. ICES Journal of Marine Science 59:365–369.
MacLennan D.N., Magurran A.E., Pitcher T.J., Hollingworth C.E. (1990) Behavioural determinants of fish target strength. Rapports et Procès-verbaux des Réunions du Conseil International pour l'Exploration de la Mer 189:245–253.
MacLennan D.N. and Menz A. (1996) Interpretation of in situ target-strength data. ICES Journal of Marine Science 53:233–236.[CrossRef][Web of Science]
Merati N. and Brodeur R.D. (1996) Feeding habits and daily ration of juvenile walleye pollock, Theragra chalcogramma, in the western Gulf of Alaska. NOAA Technical Report NMFS 126:65–80.
Miyashita K., Aoki I., Seno K., Taki K., Ogishima T. (1997) Acoustic identification of isada krill, Euphausia pacifica Hansen, off the Sanriku coast, north-eastern Japan. Fisheries Oceanography 6:266–271.[CrossRef][Web of Science]
Miyashita K., Tetsumura K., Honda S., Oshima T., Kawabe R., Sasaki K. (2004) Diel changes in vertical distribution patterns of zooplankton and walleye pollock (Theragra chalcogramma) off the Pacific coast of eastern Hokkaido, Japan, estimated by the volume backscattering strength (Sv) difference method. Fisheries Oceanography 13:Suppl. 1, 99–110.[CrossRef][Web of Science]
National Research Council. (1996) The Bering Sea Ecosystem(National Academy Press, Washington, DC) 307 pp.
Nishimura A., Mito K.-I., Yanagimoto T. (1996) Hatch date and growth estimation of juvenile walleye pollock, Theragra chalcogramma, collected in the Bering Sea in 1989 and 1990. NOAA Technical Report NMFS 126:81–88.
Rose G.A. and Leggett W.C. (1990) The importance of scale to predator prey spatial correlations – an example of Atlantic fishes. Ecology 71:33–43.[CrossRef][Web of Science]
Rose G.A. and Porter D.R. (1996) Target-strength studies on Atlantic cod (Gadus morhua) in Newfoundland waters. ICES Journal of Marine Science 53:259–265.
Russell F.S. and Yonge C.M. (1936) The Seas(Frederick Warne and Co. Ltd, London) 379 pp.
Schabetsberger R., Brodeur R.D., Ciannelli L., Napp J.M., Swartzman G.L. (2000) Diel vertical migration and interaction of zooplankton and juvenile walleye pollock (Theragra chalcogramma) at a frontal region near the Pribilof Islands, Bering Sea. ICES Journal of Marine Science 57:1283–1295.
Schabetsberger R., Sztatecny M., Drozdowski G., Brodeur R.D., Swartzman G.L., Wilson M.T., Winter A.G., Napp J.M. (2003) Size dependent, spatial and temporal variability of juvenile walleye pollock (Theragra chalcogramma) feeding at a frontal region in the southeast Bering Sea. Marine Ecology 24:1–23.[CrossRef]
Schultz L.P. and Welander A.D. (1935) A review of the cods of the northeastern Pacific with comparative notes on related species. Copeia 1935:127–139.[CrossRef]
Schumacher J.D. and Macklin S.A. (2004) Forecasting abundance of walleye pollock: indices for juvenile abundance. In Macklin S.A. and Hunt G.L. (Eds.). The Southeast Bering Sea Ecosystem: Implications for Marine Resource Management (Final Report: SEBSCC), Part 3 pp. 73–128 NOAA COP Decision Analysis Series No. 24 Silver Spring, MD. 192 pp.
Schumacher J.D. and Stabeno P.J. (1998) Continental shelf of the Bering Sea coastal segment. In Robinson A.R. and Brink K.H. (Eds.). The Global Coastal Ocean: Regional Studies and Syntheses(Wiley, New York) pp. 789–822 1062 pp.
Smith R.L., Paul A.J., Paul J.M. (1986) Effect of food intake and temperature on growth and conversion efficiency of juvenile walleye pollock (Theragra chalcogramma (Pallas)): a laboratory study. Journal du Conseil International pour l'Exploration de la Mer 42:241–253.
Sogard S. (1997) Size-selective mortality in the juvenile stage of teleost fish: a review. Bulletin of Marine Science 60:1129–1157.[Web of Science]
Sogard S.M. and Olla B.L. (2000) Endurance of simulated winter conditions by age-0 walleye pollock: effects of body size, water temperature and energy stores. Journal of Fish Biology 56:1–21.[CrossRef][Web of Science]
Springer A.M., McRoy C.P., Flint M. (1996) The Bering Sea green belt. Fisheries Oceanography 5:205–223.[Web of Science]
Stabeno P.J., Schumacher J.D., Salo S.A., Hunt G.L., Flint M. (1999) Physical environment around the Pribilof Islands. In Loughlin T.R. and Ohtani K. (Eds.). Dynamics of the Bering SeaAlaska Sea Grant Press pp. 193–215 825 pp.
Stanton T.K., Chu D., Wiebe P.H., Clay C.S. (1993) Average echoes from randomly oriented random-length finite cylinders: zooplankton models. Journal of the Acoustical Society of America 94:3463–3472.[CrossRef][Web of Science]
Swartzman G. (2001) Spatial patterns of Pacific hake (Merluccius productus) shoals and euphausiid patches in the California Current ecosystem. In Kruse G.H., Bez N., Booth A., Dorn M.W., Hills S., Lipcius R.N., Pelletier D., Roy C., Smith S.J., Witherell D. (Eds.). Spatial Processes and Management of Marine Populations(Alaska Sea Grant Publication, Fairbanks, AK) pp. 495–512 AK-SG-01-02. 720 pp.
Swartzman G. (2004) Pribilof Islands: a microcosm for the southeast Bering Sea. In Macklin S.A. and Hunt G.L. (Eds.). The Southeast Bering Sea Ecosystem: Implications for Marine Resource Management (Final Report: SEBSCC), Part 4 pp. 129–157 NOAA COP Decision Analysis Series No. 24 Silver Spring, MD. 192 pp.
Swartzman G., Brodeur R., Napp J., Hunt G., Demer D., Hewitt R. (1999) Spatial proximity of age-0 walleye pollock (Theragra chalcogramma) to zooplankton near the Pribilof Islands, Bering Sea, Alaska. ICES Journal of Marine Science 56:545–560.
Swartzman G., Brodeur R., Napp J., Walsh D., Hewitt R., Demer D., Hunt G., Logerwell E. (1999) Relating spatial distributions of acoustically determined patches of fish and plankton: data viewing, image analysis, and spatial proximity. Canadian Journal of Fisheries and Aquatic Sciences 56:Suppl. 1, 188–198.
Swartzman G., Huang C., Kaluzny S. (1992) Spatial analysis of Bering Sea groundfish survey data using generalized additive models. Canadian Journal of Fisheries and Aquatic Sciences 49:1366–1378.
Swartzman G. and Hunt G. (2000) Spatial association between murres (Uria spp.), puffins (Fratercula spp.) and fish shoals near Pribilof Islands, Alaska. Marine Ecology Progress Series 206:297–309.[Web of Science]
Swartzman G., Napp J., Brodeur R., Winter A., Ciannelli L. (2002) Spatial patterns of pollock and zooplankton distribution in the Pribilof Islands, Alaska nursery area and their relationship to pollock recruitment. ICES Journal of Marine Science 59:1167–1186.
Swartzman G., Winter A., Coyle K., Brodeur R., Buckley T., Ciannelli L., Hunt G., Ianelli J., Macklin A. (2005) Relationship of age-0 pollock abundance and distribution around the Pribilof Islands, with other shelf regions of the eastern Bering Sea. Fisheries Research 74:273–287.[CrossRef][Web of Science]
Traynor J.J. (1996) Target-strength measurements of walleye pollock (Theragra chalcogramma) and Pacific whiting (Merluccius productus). ICES Journal of Marine Science 53:253–258.
Traynor J.J. and Smith D. (1996) Summer distribution and abundance of age-0 walleye pollock, Theragra chalcogramma, in the Bering Sea. NOAA Technical Report NMFS 126:57–59.
Tveite S. (1984) 0-Group cod investigations on the Norwegian Skagerrak coast. Flødevigen Rapportserie 1:581–590.
Venables W.N. and Ripley B.D. (1997) Modern Applied Statistics with S-Plus(Springer, New York) 549 pp.
Wespestad V.G., Fritz L.W., Ingraham W.J., Megrey B.A. (2000) On relationships between cannibalism, climate variability, physical transport, and recruitment success of Bering Sea walleye pollock (Theragra chalcogramma). ICES Journal of Marine Science 57:272–278.
Winter A. (2005) Comparative analyses of the acoustically-determined abundance and distribution of age-0 walleye pollock in the eastern Bering Sea. PhD dissertation, University of Washington, Seattle, WA. 211 pp.
Winter A., Swartzman G., Ciannelli L. (2005) Early- to late-summer population growth and prey consumption by age-0 pollock, in two years of contrasting pollock abundance near the Pribilof Islands, Bering Sea. Fisheries Oceanography 14:307–320.[CrossRef][Web of Science]
Wyllie-Echeverria T. and Wooster W.S. (1998) Year-to-year variations in Bering Sea ice cover and some consequences for fish distributions. Fisheries Oceanography 7:159–170.[CrossRef][Web of Science]
Yamamura O., Honda S., Shida O., Hamatsu T. (2002) Diets of walleye pollock Theragra chalcogramma in the Doto area, northern Japan: ontogenetic and seasonal variations. Marine Ecology Progress Series 238:187–198.[Web of Science]
Youngbluth M.J. (1976) Vertical distribution and diel migration of euphausiids in the central region on the California Current. Fishery Bulletin US 74:925–936.
Zaret T.M. and Suffern J.S. (1976) Vertical migration in zooplankton as a predator avoidance mechanism. Limnology and Oceanography 21:804–813.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. A. Koslow The role of acoustics in ecosystem-based fishery management ICES J. Mar. Sci., July 1, 2009; 66(6): 966 - 973. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||