Pacific salmon are an important fisheries resource in western Alaska, ranging from Bristol Bay (with the largest commercial sockeye salmon producing system in the world) to Norton Sound and the Yukon River (Fig. 1). Pacific salmon utilize the eastern Bering Sea shelf at different time periods during their marine phase. As juveniles, all five Pacific salmon species consume resources over the shelf during the migration toward the deep waters of the basin. Significant production declines in western Alaska Chinook salmon stocks have prompted the need to improve our understanding of the marine ecology and survival of salmon.
Key dietary differences exist between the five salmon species, which can change with environmental conditions. An understanding of these differences and their drivers are necessary to understand the adaptability and vulnerability of salmon to changes in climate conditions of the Bering Sea. The broad continental shelf of the eastern Bering Sea is a highly productive system. Differences in physical forcing, and persistent species assemblages split the eastern shelf into northern and southern eco-regions at approximately lat. 60˚N. Cross-shelf species assemblages are also apparent, particularly in the southern Bering Sea. These differences are apparent both in fishes and their prey base.
Figure 1. The eastern Bering Sea shelf. Pelagic trawl surveys by the Alaska Fisheries Science Center are outlined in red.
To sample populations of Pacific salmon in the Bering Sea, surface trawl surveys have been conducted by NOAA since the start of the Bering-Aleutian Salmon International Survey (BASIS) research program in 2002. Typical survey effort has ranged from mid-August through early October, using a Cantrawl model 400/601 hexagonal mesh mid-water rope trawl with 1.2-cm cod-end liner, modified to fish at the surface.
Information on salmon diets was collected during both warm (2003-05) and cold (2007-11) periods. Stomachs were collected from juvenile and immature/maturing fish from all five salmon species. Stomach content analysis was performed onboard during the surveys, using protocols established by the Pacific Scientific Research Fisheries Center (TINRO, Vladivostok, Russia). Principle component ordination was used to investigate if a given subset of measured factors has noticeable influence on the dietary patterns of Pacific salmon (Fig. 2). These factors included climate conditions (warm or cold) and life history stage (juvenile or immature/maturing). Vector lengths as shown in Figure 2 represent the strength of the correlation or prey groups to component axes. Vector color corresponds to their respective prey group. Colored dots are the mean species diets by climate period (warm and cold). Climate has the largest influence on diet, as illustrated on the x-axis; species specific feeding strategy has the second largest influence, as illustrated on the y-axis.
In general, piscivory increased with warmer conditions, while the consumption of large crustacean zooplankton increased under colder conditions. Chum, pink, and sockeye salmon experienced the largest dietary shift with climate conditions. Chum salmon consistently incorporate a much higher proportion of gelatinous prey than any other species of salmon and is consistently lower on the y-axis (Fig. 2). Pink and sockeye salmon had a similar feeding strategy, and the largest response to changing climate conditions. Both of these species switch from an almost entirely large crustacean zooplankton diet during cold conditions to a diet dominated by age-0 pollock during warm conditions. This overall strategy of prey switching with availability is reflected in their similar position on the y-axis. Chinook and coho salmon displayed the least amount of variation in prey type under differing climate conditions. They are always predominantly piscivorous, but may switch which species of forage fish makes up the majority of their diet. Chinook had the smallest response to changing climate conditions.
Figure 2. A biplot of principal component ordination of salmon diets on the eastern Bering Sea shelf. The first principal component (PC1) of salmon diets is on the x-axis and the second principal component (PC2) is on the y-axis. Grey circles are individual observations of salmon diets. Colored circles are the average loadings of salmon diets on PC1 and PC2 during warm (2003-05) and cold (2007-11) years in the eastern Bering Sea. Lines indicate the prey contrasts within PC1 and PC2 and prey species are identified along the margin of the figure.
In some cases these changes in diet can be influenced by confounding factors. Shifts in distribution and abundance of both predator and prey should be considered. During warmer years, the distribution of juvenile coho salmon typically shifts further south. The southward movement places them in proximity to an increased abundance of age-0 pollock in the surface waters of the Bering Sea during warmer conditions. The incorporation of age-0 pollock into the diet of coho salmon is more likely due to this migratory shift. While there is a shift in the distribution of juvenile sockeye salmon during warm years to these areas of high age-0 pollock abundance, these salmon are always coming out of Bristol Bay and moving across the southern Bering Sea shelf. Similar to pink salmon, this dietary shift is likely a generalist approach due to prey switching to a more readily available resource. The smallest dietary shift with climate conditions occurred within Chinook salmon. This likely reflects a higher degree of dependence on a particular prey type or prey resource.
Multivariate models will be developed for individual species to further investigate the biological and environmental variables influencing salmon diets in the Bering Sea. This research is expected to provide insight into the adaptability and vulnerability of salmon to changes in the climate conditions of the Bering Sea.