Ocean Indicator Ancillary Data and Future Research Directions
Northwest Fisheries Science Center scientists collect and analyze various data to improve the accuracy of Pacific salmon forecasts.
Forage Fish and Pacific Hake Abundance
We have also explored developing an index that describes food–web interactions between juvenile salmon and their fish predators, chiefly Pacific whiting, aka Pacific hake. We could base this index on interactions between forage fish (e.g., anchovies, smelt, and herrings), juvenile salmon, and hake.
This interaction is somewhat complex and probably non–linear. We hypothesize that hake moves to continental shelf waters during warm–ocean years, where salmon are more susceptible to predation. During cold–ocean years, hake feed in deeper waters offshore, near the shelf break. They are not actively feeding in the shallow continental–shelf waters inhabited by juvenile salmon.
During cold ocean conditions, when zooplankton production is high, small forage–fish biomass increases. This increase in forage–fish abundance allows predators to "see" and prey upon forage fish more often than salmon. Most forage fish populations (smelt, herring, and anchovy) do well during cold conditions but tend to crash during warm conditions. There is a lag of at least one year between boom and bust periods. The interaction among zooplankton production, forage fish abundance, juvenile salmon survival, and hake predation is likely to be non–linear.
We have not analyzed or modeled these interactions. Nevertheless, Figures AD-01 and IUD-02 demonstrate the pronounced interannual differences in the abundance of forage fishes. These are, in part, related to the cycles gauged by the current ocean ecosystem indicators.
Due to funding constraints, we could not conduct any forage fish/predator study cruises in 2012. We believe that this is a promising indicator and hope to resume sampling soon.
In 2011, we continued to find very low Pacific hake abundances (Figure AD-01) compared to 1998, 2003, and 2004. We observed low densities during the cool, negative-PDO phases of 1999-2002 and 2008-2011. Conversely, high abundances occurred during the warm, positive-PDO years of 1998, 2003, and 2004. Probably not coincidentally, these years correspond respectively to "good" and "poor" periods for coho survival.
We expected high abundance levels for hake in 2005 and 2006 (warm years), but this expectation was not met, due possibly to the timing of their northward migration. That is, hake may have moved further north (off Canada) during the warm years of 2004 and 2005, and thus would have been preying on salmon earlier (May) rather than later (Jun-Jul) in the season.
Forage fish show a one year lag between change in ocean phases and population response. We observed anomalously low abundances during the first year of a "cool phase" (1999) and anomalously high abundances during the first year of the "warm phase" (2003). This lag time reflects the time it takes for 0-age fish to grow large enough (i.e., 1-year-old) to be captured by the surface trawl.
The failure of hake to maintain high abundances in 2005 and 2006, and the 1-year lag in the response of forage fish to ocean conditions changes, contributes to the lack of a linear relationship between salmon catches or survival and forage fish or hake densities. The relationship between salmon marine survival and other fishes appears to be very complicated and probably influenced by additional factors.
Forage fish numbers continued to be relatively high in 2011 (Figure AD-02), probably due to relatively good recruitment in 2010. These high densities comprise a positive indicator since juvenile forage fish (ages 0 and 1) are among the favored prey of both coho and Chinook salmon. Salmon were probably not food limited in either 2010 or 2011. High numbers of forage fish in 2011 and a probable cold ocean in spring 2012 (which is good for forage fish recruitment) indicates favorable ocean survival for coho and Chinook salmon in 2012.
The Second Mode of North Pacific Sea Surface Temperature Variation
Changes in sign of the PDO tend to follow an east/west dipole. When the North Pacific is cold in the west, it is warm in the east, and vice versa. Bond et al. (2003) showed that sea surface temperature variability has a second mode, which reflects north/south variations. This pattern first appeared in 1989 and continues to the present.
We have not yet investigated this pattern fully because the negative phase of the first mode (the PDO) indicates favorable conditions in the northern California Current, as does the negative phase of the second mode (called the "Victoria" mode). However, oscillation in the second mode would index good vs. poor ecological conditions between the Gulf of Alaska and northern California. Therefore, this second mode may serve as a better index of conditions for spring Chinook salmon. Conventional wisdom is that spring Chinook resides in the Gulf of Alaska during most of its years at sea.
Based on samples collected along the Newport Hydrographic Line, we developed time series of both total chlorophyll and the fraction of chlorophyll smaller than 10 µm. These data serve as estimates of phytoplankton biomass. We will use both data types to describe interannual variation in the timing of the spring bloom (which can occur between February and April) and summer blooms during July–August upwelling. These measures should provide an index of potential conditions (good vs. poor) for spawning of copepods and euphausiids.
Euphausiid Egg Concentration and Adult Biomass
Euphausiids are an essential prey item for juvenile coho and Chinook salmon. Sampling along the Newport Hydrographic Line has also yielded a time series of euphausiid egg abundance. These data may serve as an adult euphausiid biomass index, proving useful in comparisons of interannual variation in abundance, survival, and growth for these salmon species.
Since 2000, we have also been sampling at night along the Newport Line to capture adult euphausiids. This sampling's long–term goal is to produce an index of euphausiid biomass in the northern California Current. We also measure molting and egg production rates in living animals in anticipation that we can use these data to calculate euphausiid production.
Interannual Variations in Habitat Area
From the salmon trawl surveys conducted in June and September, we are developing "Habitat Suitability Indices." We hope these will prove useful in providing more precise predictors of the potential success or failure for a given year–class of juvenile salmonids. For example, we have determined that chlorophyll and copepod biomass levels are the best predictors of habitat size for juvenile Chinook salmon. Interannual variation in potential habitat area may also serve as a correlate for salmon survival during the first summer at sea.
Salmon Predation Index
A salmon predation index would integrate four variables previously found to influence Columbia River salmon's predation rates in the ocean (Emmett 2006). We based these variables on the following spring (May/June) measurements from 1998 - 2012:
- The abundance of Pacific whiting (hake) off the Columbia River.
- The abundance of forage fish off the Columbia River.
- Turbidity of the Columbia River.
- Columbia River flows.
We estimated predator and forage fish abundances annually from the Predator/Forage Fish Survey. We estimated turbidity using satellite imagery, Secchi disc readings, and transmissometer measurements, we collected from 1998 – 2012. Initial analyses indicated that during years when hake abundance was low and forage fish abundance, turbidity, and Columbia River flows were high, salmon marine survival was high. However, if even one variable had an opposite value, salmon marine survival declined. We stopped sampling of predator and forage fish after 2012.
Potential Indices for Future Development
Some additional indices that we could develop include:
- An index of Columbia River flow.
- Predictors of coho and spring Chinook jack returns.
- Indices based on salmon feeding and growth.
- Indices based on salmon health (diseases and parasites).
- Indices that estimate zooplankton production rates, such as:
- Euphausiid growth rates from direct measurement of molting rates
- Euphausiid growth rates from cohort developmental rates
- Copepod growth rates from direct measurement of Calanus egg production rates
- Copepod growth rates from empirical growth equations