On November 9th and 10th, 2004, the Council and NOAA Fisheries sponsored a reservoir-operations/flow-survival symposium. The ISAB was asked to provide scientific feedback on the Council's proposed operations of Hungry Horse and Libby hydroelectric projects, as described in the Council's Mainstem Amendments, and NOAA Fisheries' questions about Montana's subsequent System Operations Request (SOR, referred to below as the Montana proposal). At issue is the potential trade-off between the probable detrimental biological effects on the resident fishes in reservoirs and rivers at Hungry Horse and Libby dams under flow management with existing BiOp operations and the unquantified potential detriment to anadromous fishes in the lower Columbia River under the Council's 2003 Mainstem Amendment and Montana Systems Operations Request. There is also interest in determining the proper design of an experiment(s) to provide this quantification. The anadromous fish of most concern are ESA-listed Snake River fall Chinook salmon because they are believed to migrate during the August-September period when flows have been augmented by releases from Hungry Horse and Libby dams.
The ISAB found the relevant science to be in flux, providing no unambiguous answers at this time. Flow-survival debates have been prominent in the region since the first Fish and Wildlife Program and its water budget (1982). The current amendment and Montana's proposal have focused attention on one particular manifestation of the issue. Although the likely detrimental effects of current operations on the gross biological productivity (i.e., production through all levels of the ecosystem, including plants, invertebrates, and fish) of the Hungry Horse and Libby reservoirs and downstream rivers are easy to envision and are probably of a substantial magnitude, the way in which current operations will affect the species of greatest concern (sturgeon and bull trout) is actually unknown. The physical changes in flow in the Columbia River below Chief Joseph Dam (the uppermost limit of anadromous fish in the Columbia River) that would result from the Montana proposal are affected by so many factors that predictions remain too uncertain for fine-scaled biological analyses. Little attention has been paid to measuring the physical features of "flow" important to fish migration and survival, such as water velocity or within-day variations due to load following (power peaking). Reliance on broad averages and sweeping generalizations of flow-survival relationships are not adequate for resolving this specific issue. In addition, the effects of altered temperatures on salmon require reevaluation for their relationships to flow, in spite of new thermal modeling efforts.
Although summer-migrating juvenile fall Chinook salmon from the Snake River have been the main concern for downstream effects of the Montana proposal, there is new information about this stock's life history. Some juveniles are holding over their first winter in fresh water and emigrating as yearlings in the spring (termed the "reservoir" life history, also referred to as the holdover life history). Importantly, a disproportionately large percentage of returning adults are originating from these holdovers. This new information leaves us with major unresolved implications for management because populations adopting the holdover life history will not respond to flow changes in the same way that the "ocean-type" summer-migrating fall Chinook normally do. Juvenile fall Chinook salmon have a complex life history, much more so than juvenile spring/summer Chinook salmon, with observations of juvenile fall Chinook salmon emigrating in the Snake and Columbia Rivers throughout most of the year. The intent of flow augmentation is to reduce mortality of smolts by speeding their migration to the ocean. With the recent findings of the large adult contribution from migrants exhibiting the reservoir life history, and also for PIT-tagged late fall migrants (NOAA Fisheries, unpublished data), the strategy of using flow augmentation to speed migration should be reassessed.
In addition, estimates of survival of in-river migrants have included holdovers as an unknown part of the "mortality" experienced during emigration, i.e., a portion of those estimated to have died may actually have survived, but did not pass a tag detection facility downstream during their first year. Resolving the very large uncertainties about actual outmigration survival rates and the contributions of the two distinct life history patterns to the effective production of wild fall Chinook will require new data collected using a new study design. In particular, it will be important to obtain scales from returning adults as they pass Lower Granite Dam to secure a large enough statistical sample for determining frequencies of the two life history types at that stage. This task should be implemented in conjunction with continued studies of PIT-tagged smolts released above Lower Granite Dam, and the exploration of various options for radio tagging migrating smolts to determine their life history behavior, as well as the overwintering locations and mortality patterns for the reservoir type. The importance of these data collection issues for the Snake River fall Chinook goes far beyond the narrow questions concerning the Montana System Operations Request. For example, transporting Snake River fall Chinook salmon subyearling smolts to below Bonneville Dam likely affects the proportion of smolts utilizing the reservoir life history strategy, as well as their size and time of arrival to the ocean. This aspect of transportation needs evaluation.
Traditional flow-survival analyses are currently benefiting from detailed statistical analyses, with the result that different functional mechanisms for fish survival appear to be operating at different flows. There is good physiological evidence supporting delayed effects of hydrosystem passage that are likely expressed in survival in the estuary and ocean. Because adults respond negatively to flow increases, the effects of these increases on them, not just on juveniles, need to be considered as well. No existing models seem adequate for evaluating the flow effects from the Montana proposal.
All indications are that the down-river effects of the shifts in flow associated with the Council's Mainstem Amendments of 2003 will be small. There is some uncertainty whether the effect is positive or negative because flows under the Montana proposal will be slightly greater than BiOp flows at times and slightly less at other times. As a result, the Council's hypothesis that the effects on survival of salmonids in the lower Columbia River will be indiscernible is probably reasonable. Nonetheless, there is reasonable concern for potential cumulative effects of even such small changes.
No "passive" design for measuring downstream effects of the proposed change is likely to be effective in a reasonable period of time. Deliberate experimental flow manipulations of an amplitude considerably larger than flows that will result from the Amendments probably would allow empirical quantification of flow effects. Whether or not present institutional constraints would allow such manipulations, however, is an open question. An experimental trial of Montana's proposed flow regime is unlikely to reveal incremental biological effects within a modest number of years because annual variability in flow and salmon survival due to other causes is much larger than the expected effect of the Montana proposal.