Peter Adams: Our principal interest has been in predator-prey studies, and my comments deal with Recommendation 6. We have been monitoring chinook salmon stomach contents off central California since 1980 and prey concentrations since 1983. We have been using a 24-m headrope mid-water trawl to catch forage fish and MIK nets (3-m Issac-Kidd trawls) to catch euphausiids, both with a great deal of success. However, there is no getting around the fundamental patchiness of both salmon feeding and prey distribution, and this requires samples collected over a long time scale. We have found that an understanding of the feeding dynamics and reason for the shifts in prey distributions underlay much of the changes in salmon distribution. Finally, our neuston sampling has not been very useful in explaining these interactions.
Dick Carlson: I have long been concerned with some of our research pursuits that seek answers amidst many factors interacting at once, that is, a flux of conditions that are not separable, and cannot be examined individually or experimented with except as part of a complex. Some research programs make promises to deliver answers to questions that may not have answers—at least none that are consistent and easily explained or understood. Some of the questions and hypotheses we posed would be difficult to test. One area in which I see considerable potential for developing new information is identifying ocean migration patterns of first-year juvenile salmon. This would require sampling over the potential migration routes often and thoroughly enough to locate stock groups of salmon that are found consistently in certain areas or under certain conditions, and accumulating evidence for a pattern (that the researcher defines). Separate cruises to obtain oceanographic and hydrographic information would allow the fish sampling cruises to better focus on that activity. Recommendation 3 (interaction among stocks) also has some potential for resolution by field sampling. Capture at sea of young salmon from stock groups found consistently together would be evidence for interaction, and opens other avenues for investigation. Stock identification techniques would play a major role in this effort.
Steve Ignell: I support the concept of a monitoring program (Recommendation 1), but it needs to be tightly connected to research questions. Rather than arbitrarily set up various monitoring lines, we need to think very carefully about how these lines will provide data to test specific hypotheses. We also need to link the monitoring program to key index stocks. In Recommendation 2, we need to distinguish between stock identification methods that assign stock origin probabilities to a group of fish (GSI and scale analyses) and methods that identify the origin of individual fish (CWT and otolith marks), as they are used to address different research questions. The mass marking (otolith marks) of stocks of salmon in the Gulf of Alaska offers the first good opportunity to study identifiable cohorts of different stocks over a succession of seasons. For each stock, ocean distribution, energy stores, growth, and maturation can be monitored in the context of biophysical variation. To expand upon the importance of energy stores in salmon (see Recommendation 4), several lines of research suggest that an individual's age at maturation is determined by its energy stores, measured at a particular season of the year. For example, if a salmon's lipid stores in November exceed a "set point" of 10% whole body weight, the salmon will undergo maturation during the succeeding year. Energy stores reflect the nutritional history of a salmon (i.e., how well it fed during the preceding season of growth). Energy stores also reflect the physical environmental (temperature) history, since growth is related to both ration (amount and quality of food) and temperature.
John Karinen: In Question 3, the use of the term "physical forcing" without specification of the range of variables or environmental conditions (e.g., site-specific and regimewide) is inappropriate. Physical factors affecting growth and survival may range from site-specific conditions (tidally modulated boundary effects between water masses of different temperature, salinity, and density) to regimewide interannual, decadal, or less frequent changes in physical conditions. Fish that are successful in finding and following food concentrations along density gradients offshore may survive and grow faster than fish that reside in the center of a water mass where food may not be concentrated. Regimewide changes in physical conditions may change current patterns, temperature distribution, storm tracks, and incidence and persistence of gyres, all of which may affect fish migration routes and perhaps growth and survival. Hypothesis 1 should address site-specific effects rather than regime-wide changes, which have already been shown to affect salmon migration patterns. Do salmon key in on boundary effects and migrate along these food-rich areas? I would change Hypothesis 3 to "Migration paths are selected to maximize survival," as fish are, perhaps, keying on prey abundance and avoiding predators.
Katherine Myers: Much of the difficulty that our work group had in identifying scientific questions and translating these into testable hypotheses arises from a lack of basic understanding of the offshore habitat and life history of Pacific salmon. Our most significant scientific contributions in the near future will likely arise from exploratory and experimental field research on the basic biology and ecology of salmon and associated species in offshore waters. As stated in Recommendation 1, the need for an integrated, coastwide commitment to a fishery-oceanographic monitoring program is closely linked to practical management and natural resource conservation applications. The size and number of adult salmon returning to coastal fisheries each year is largely a mystery because we do not have any information on their oceanic growth and survival. With the end of high-seas driftnet fishing, we have a unique opportunity to discover the natural variation in offshore growth and survival of salmon.
Peter Rand: Many of our hypotheses are mutually exclusive, and will give rise to unique predictions on locations and size (growth) of salmon in space and time. Through modeling exercises, we can codify these hypotheses and make a series of predictions that can be rigorously tested against thoughtful field sampling, ideally employing state-of-the-art sampling technologies, such as acoustics, smart tags, and remote sensing. For example, if we assume that salmon are "navigationally challenged" and that their movements are random, spatial distribution can be projected solely as a consequence of dominant ocean currents. This model can be compared to observed catch distributions. Alternatively, we can assume salmon are minimizing energy losses or maximizing their energy return on foraging investments, which would lead to additional, unique predictions regarding migration behavior and subsequent distribution that can be further tested. Modeling techniques and field sampling are both important and operate synergistically to help move our science forward. This interplay is particularly appropriate for offshore research because of the difficulty and expense of carrying out exhaustive field efforts and experiments across these broad time and space scales.
Bruce Wing: A definition of "physical forcing" (see Question 3 and Hypothesis 1) is critical to designing observational programs. What physical features do we want the oceanographers to monitor and investigate? Physical features such as thermal fronts, the Sitka Eddy, and river plumes have been shown to affect the return migration routes of several species of salmon, and there is no reason to suspect that such features do not affect outgoing migration routes. We need to identify which features beyond and over the continental shelf are influencing migration routes. To address Hypothesis 3, which relates to evolutionary time scales and not to annual variability, will require a lot more basic knowledge of physical and biological conditions than can be gathered in a short time.