The issue of toxic contaminants in the Columbia River basin as an influence on salmon populations has always taken a back seat to various aspects of hydropower development, including turbine-induced mortality, delayed migration, elevated temperatures, etc. It is valuable to see proposed research directed at toxic pesticides, for the intensive agricultural development of basins like the Yakima introduces a mix of potential toxic materials from herbicides, insecticides, and other crop treatments. These are monitored chemically in water by various agencies, but the implications of chronic exposures for salmon remain unclear.

This said, the current proposal has good parts and poor parts, and the ISRP requests a response. This proposal is improved from the previous one by focusing the research on salmon. Laboratory studies to determine the biological effects of pesticides on salmon, particularly the low-level, chronic exposures, seem especially relevant for relating to monitored levels in the spawning, rearing, and migration environment. However, the proposed model may, at this point in time, be more useful as a conceptual tool rather than as a mathematical, predictive tool. Thus, we recommend that the laboratory research, as validated by field assessments, be given priority, while the model development be curtailed.

The technical background and need for the project is quite well presented. The impact of toxic substance on ESA-listed and other species is an important problem and is poorly understood in the Columbia Basin. This proposal seeks to address the problem through a combination of empirical research and modeling, with the outcome to be a model predicting toxic impacts on population dynamics. This is an ambitious but speculative approach. The information required for the model is difficult to obtain for natural populations and habitats. If there is any inadequacy to the background section, it is the omission of references to the work done in the 1960s and 1970s by Charles Warren and his students at Oregon State University. There was a very strong effort there to relate contaminants (mostly pulp and paper manufacturing wastes) to the ecology of salmonid fishes, from the individual up to the population level. An excellent synopsis of the earlier work appears in Warren's book "Biology and Water Pollution Control" (Warren, C. E. 1971. W. B. Saunders, Philadelphia), but even after the book was published there were excellent studies on bioenergetics (e.g., Wayne Wurtsbaugh's research on the effect of ration size on juvenile steelhead). This body of work is well worth checking out.

The rationale is clearly spelled out with regard to large regional programs, but there is little reference to how this work fits into the context of some of the larger subbasin plans (e.g., Yakima). The proposal is broadly related to Objectives and Strategies in the Council's Fish and Wildlife Program. It directly addresses uncertainties identified in the Biological Opinion Remand and the Council's Research Plan. The proposal describes opportunities to link to most of the significant federal and state water quality monitoring efforts. More local monitoring efforts (tribal, county, municipality) are not mentioned, but perhaps these are few and far between. The sponsors plan to use data produced by two monitoring programs in the basin to parameterize the model.

In general, the objectives were clearly explained and sufficiently detailed. The fundamental question is how meaningful the results will be to actual pesticide impacts on populations.

The methods for the basic toxicological studies, likewise, are thoughtfully detailed, although there were a few concerns.

Task 1. The standard 96-hr exposure tests work great in the lab but are not always the best approximation to the real world. Is there any chance of dosing the experimental channels at Manchester with single or binary pesticide mixtures to study direct effects on feeding and growth? Also relative to Task 1, in the lab it might be more realistic to feed the fish a standard ration that represents a "moderate" food level (say, 0.5% body weight per day), as opposed to feeding to repletion.

It appears that the toxicological studies will expose fish at likely lethal levels. This would be a positive control, but the answers are likely obvious. They should consider a graded study, 20, 30, 40% associated with various agricultural surveys (e.g., graded levels of expected cholinesterase inhibition). It would be useful to look at field monitoring studies to obtain exposure levels; this would be more realistic, mirroring likely environmental exposure. A combination of long- and short-term exposures would also be more realistic.

Task 2, Study 1. The sponsors shouldn't overlook the literature on bioassessment metrics such as RIVPACS. There is quite a large body of information on the effects of various pollutants on aquatic invertebrate community composition in Europe and Australia/New Zealand. RIVPACS is mentioned because the Washington Department of Ecology has been developing a reference-site database on bugs for relatively unaltered streams -- many of which are in the interior Columbia Basin. Chuck Hawkins at Utah State University has applied the RIVPACS approach to many sites in Washington and Oregon. He'd be a good contact.

Task 2, Study 2 (relationship between prey quantity and salmon behavior and growth). There is a fairly rich literature on this both in the west (e.g., see some of the papers by Fausch et al. at Colorado State University, Jim Hall and his students at Oregon State, Ken Cummins and Peggy Wilzbach at Humboldt State) as well as in Europe (check out some of the Atlantic salmon literature). Additional background work will help with this task.

Task 2, Study 3. The sponsors will probably find that stomach contents of relatively young fry contain really tiny organisms like mites, copepods, and really small chironomids. Many of these are smaller than early brine shrimp instars. It might be a good idea to have a backup source of really small critters when conducting the small, medium, and large prey study (Daphnia might work).

Several of the methods for the ecological studies deserve greater explanation. How will swimming and feeding behavior be quantified (Objective 1, study 2)? How will variation/uncertainty in performances such as insect response to exposure be incorporated into the model (Objective 2, study 1)? The sponsors have not provided sufficient detail about the bioenergetics model and how it will incorporate the multiplicity of environmental factors the sponsors propose to include in the model. The sponsors do not indicate where the data will come from for Objective 3, Study 1. Methods for objective 3, study 3 are vague, simply calling for the incorporation of new data into the model.

The modeling component of the proposal as a whole was troubling. One really doesn't know how the research relates to what fish are doing in nature. The proposal plans model development as if many ecosystem relationships were known, but they aren't (growth rates in the ocean, etc.). The development of a predictive model, although stylish these days, detracts from this proposal. The ISRP strongly favors a conceptual model (even with some quantification) as a guide to the research, but finds the full quantification into a predictive model to be premature.

Plans for comparing predictions of the model with results in the field are lacking. Perhaps it might be possible to conduct studies with caged fish in streams during the time they're exposed to pesticide runoff, or perhaps there are ways you can use hatchery fish to evaluate some of the model's predictions.

The facilities and personnel are very well qualified to conduct this project. Staff members have good publication records.

The research plans suggest benefits to fish in the long run. The work focuses on Chinook salmon although there will be plenty of opportunities to carry the findings to other species. The lack of information on this potentially important topic makes the laboratory component of the effort a high priority. The benefits from the modeling aspects of the project are uncertain, however. On the one hand modeling effects of chemical contaminants on focal species is needed, but on the other it is not clear that this model can provide much insight into the problem that will be applicable. The information required to parameterize the model is difficult to obtain for natural populations and habitats, and can be highly variable. The sponsors do not explain very well how variability will be dealt with.

A response is needed on the laboratory component, responding to items discussed above. The modeling component does not appear to be fundable, except as a conceptual guide to the research, but the sponsors may respond with better justification. The effort would be better focused on experimental portions and related field assessments in association with chemical monitoring data.