The group concluded that a major reason for obtaining data on survival rates in the three major habitats was to avoid "racheting to extinction." While there appears to be evidence for decadal-scale changes in ocean conditions and overall levels of salmon production, it is not clear what role low frequency changes in climate have on freshwater environments. Further, freshwater productivity may be affected by habitat disruption. Salmon managers need information on freshwater survival to ensure that river habitats maintain the capacity to rebuild stocks. Salmon populations are buffeted by changes in marine habitat and fishing while at sea, and recovery from these stresses would be more difficult if the populations were recovering from low freshwater survival as well. For example, if survival was low in both freshwater and marine habitats simultaneously, it may be difficult to change harvest rates fast enough to maintain a viable population. Alternatively, if salmon were experiencing high survival rates in the sea, compensating for low survival in fresh water, conservation of habitats might be difficult but would be essential for rebuilding efforts. It is therefore necessary to track survival in the three major habitats simultaneously.
There was some debate about the usefulness of setting up a null hypothesis for the ecological work required, given the difficulties of designing field experiments covering oceans and major landforms such as watersheds and estuaries. The following was suggested as a conceptual null hypothesis: "There is no link between freshwater and ocean survival."
It was agreed that an understanding of the roles of the three major habitat types on salmon recruitment requires analysis of long-term series of demographic data. As a first step it was suggested that there are some existing data sources that have not yet been fully utilized. These include the coastwide coded-wire-tag (CWT) database, and scale archives for sockeye salmon. A first step would be to assemble a meta-database on salmon survival that would make information readily accessible to researchers. The databases established for endangered species could be reviewed to avoid duplication of effort. It was recognized that the database varies with species, sampling protocols, and geography, but it was noted that in spite of these variations some coastwide data sets such as those for CWTs and stock identification seem to be working.
While chinook salmon data for disrupted systems such as the Sacramento and Columbia Rivers are relatively good, information from rivers and estuaries farther north, such as the Fraser and Stikine Rivers, are poor. The northern systems offer an opportunity for reference (or "control") data. The information on sockeye salmon from the northern systems, however, is good, and in fact results from these databases tend to heavily influence our understanding of salmonid ecology in the Northeast Pacific. However, it was recommended that investigators continue to "mine" the sockeye salmon database, including the scale archives of the former International Pacific Salmon Fisheries Commission.
Regarding new initiatives, there was at least partial consensus that progress would be made if research were focused on a few watersheds along the coast of the Northeast Pacific. The emphasis of this work would be to establish long-term monitoring programs that would allow assessment of decadal-scale processes in all environments. Research at present is not focused enough to measure survival simultaneously in marine and freshwater habitats, and there is a desperate need to set up coordinated studies addressing salmon survival relative to hydrology and oceanography. A network of salmonid ecologists, oceanographers, and hydrologists in the region might be established to move this idea along. It might be useful to restart some of the older long-term studies on freshwater survival to speed up the process of understanding changes in fresh water due to climate or habitat change.
Several approaches to determine survival in the three major habitat types were discussed. It was pointed out that measuring survival rates should be viewed as the major goal in a research program. Understanding mechanisms will be very difficult and from a management point of view may be a secondary priority. Existing weirs and counting fences in headwaters could be harnessed to measure survival if additional facilities were built just above the limit of tide in the estuaries. Ocean and estuary survival could then be obtained by subtraction. Survival in the estuary would be more difficult to obtain, but with concerted efforts such as those of Parker (1971), the estimates could be done. The idea of tracking an individual fish throughout its life history in freshwater, estuarine, and ocean habitats was raised as an alternate approach to population ecology. Recent developments in telemetry may enable such investigations.
Other possible projects could focus on habitat or landscape changes in the region, but they should be in association with measures of survival. It was thought that investigating the Pacific Northwest Index (sea surface temperature, rainfall, and thermocline depth) might be extended to other coastal regions. There is also scope for development of other indices in fresh water (e.g., snow pack) and estuaries (e.g., salt wedge penetration). Changes in the condition of landscapes in watersheds and estuarine wetlands through time could be assessed by analyzing historical series of aerial photos.
It was concluded that there was a lack of vision for research concerning partitioning of salmon survival in their major habitats. Terrestrial ecologists have a comprehensive network of monitoring plants and animals and their habitats on land, but there is nothing comparable for salmon in rivers, estuaries, and oceans. In general, funding to determine salmon survival is spotty, unstable, and only available in crises. There is an illusion in the public that current levels of activity are resulting in the accumulation of reliable abundance data that can be used to monitor changes in salmon survival. It was proposed that the involvement of nontraditional groups (National Science Foundation, schools, clubs, etc.) might have utility in maintaining continuity in long-term monitoring programs. The reality and need for these types of programs must be pointed out to politicians at all levels of government, to educators, and to our friends and neighbors at the local level.
Parker, R. R. 1971. Size selective predation among juvenile salmonid fishes in a British Columbia inlet. J. Fish. Res. Board Can. 28:1503-1570.