Predicting Survival Rates
            Predicting survival rates of adults from characteristics when they were juveniles is desirable in many situations. For salmon, generally only about 2% of  juveniles survive to adulthood. Characterizing juvenile salmon might help predict the perfomance of individuals in the natural environment and smolt-to-adult-return rates (SARs). Nonetheless, precise survivorship rates for all juvenile stages can be difficult to predict because data generally only exist that estimate direct mortality over segments of the juvenile life stage. Generally, direct survivorship rates for juveniles outside of the riverine migratory stage are unknown. Furthermore, direct survivorship rates of juveniles that are measured may not provide a good estimate of possible subsequent mortality to those same juveniles. Indirect mortality is a non-instantaneous death that can be caused by many possible stressful factors such as extreme environmental conditions, lack of food resources, and sublethal injuries. Indirect mortality can be very difficult to measure in cases like salmon that endure stressful factors in the riverine environment but die in the ocean because of their weakened state.

The Federal Columbia Power Hydrosystem in the Columbia River Basin and Salmon
In general, North Pacific salmon from the Columbia River Basin have suffered declining population sizes from migration across dams. Efforts to counteract these negative effects have been put into barging a proportion of the salmon across the dams. Research has shown that although barging increases survivorship rates of juvenile salmon to nearly 100%. However, these positive results do not show up in the survivorship rates of adults returning to spawn. These results are highly suggestive of latent mortality effects on juvenile salmon.

The Columbia River is the most hydroelectrically developed system in the world. There are over 400 dams in the Columbia River basin, 31 of which are from the Federal Columbia River Power System (FCRPS), and 11 of which are on the main river (Williams et al. 2005). The first dam, Rock Island Dam, was relatively small and built in 1932. Bonneville Dam and Grand Coulee Dam are two of the larger dams which the federal government completed in 1938 and 1941 respectively. Some of the other dams on the Columbia River completed in the 1960s and 1970s are Ice Harbour Dam (1961), Lower Monumental Dam (1969), Lower Granite Dam (1975), and Little Goose Dam (1978).

The Decline of North Pacific salmon

The decline of anadromous salmon in the Columbia River in the 19th centuray was caused by overfishing, and then in the early 20th century, the numbers decreased even further from environmental degredation caused by mining, grazing, logging, irrigation, and the construction of dams. In the late 1960s, the average survival rate of juvenile chinook salmon from the Clearwater, Snake River, and Columbia River conjunction to below Bonneville Dam, which includes four dams along this stretch, was between 40% and 55% (Williams et al. 2001). After the completion of four more dams, this survival rate decreased to an average of 16% from 1975 to 1980 (Williams et al. 2001). Part of the reason for lower survival rates is likely longer migration times. For juvenile salmon to travel past four dams, it took about 10 days in high flow rates and 20 days in low flow rates. Across this same distance, but with eight dams, it takes 15 days in high flow rates, and 40 days in low flow rates (Williams et al. 2005). Efforts by the Army Corps of Engineers to counteract negative effects of the dams on anadromous fish include by-pass systems, spillways, and transportation by barges and trucks. But these efforts to this day have not improved survival rates enough to maintain the target 2% to 6% SARs for recovery of listed stocks (Marmorek et al. 1998). Many species since 1992 have been listed as endangered under the Endangered Species Act, some of which include Snake River sockeye salmon (listed endangered in 1992), Snake River fall and spring/summer Chinook (both listed threatened in 1992), Snake River Steelhead (listed threatened in 2005), upper Columbia River spring Chinook (listed endangered in 1999), and lower Columbia River chum and coho (listed threatened in 1999 and 2005 respectively). The Snake River spring/summer Chinook salmon were reclassified in 1994 as endangered on an emergency basis (from Deriso et al. 2001).

Latent Mortality
“We define latent mortality associated with the FCRPS (for Snake River fish) as any
mortality that occurs after fish pass Bonneville Dam as juveniles that would not occur if the FCRPS dams did not exist. Latent mortality associated with the FCRPS might result from changes in migration timing; injuries or stress incurred during migration through juvenile bypass systems, turbines, or spill at dams that does not cause direct mortality; disease transmission or stress resulting from the artificial concentration of fish in bypass systems or barges (Williams 2001, Budy et al. 2002); depletion of energy reserves from prolonged migration (Congleton et al. 2004); altered conditions in the estuary and plume as a result of FCRPS construction or operation; or disrupted homing mechanisms.” Williams et al. 2005 – page 106

Differential Mortliaty
To quantitatively compare latent mortality of transported juvenile Chinook salmon to that of in-river migrating ones, Marmorek and Peters (1998) developed a simple model for differential mortality (D). First, the survival rate in the ocean (O) can be determined by dividing the SAR by the juvenile survival (S) for transported (T) fish (i.e.  SOT = SART/ST) and for in-river (I) migrating fish (i.e.  SOI = SARI/SI). The ratio of ocean survival of transported fish to that of in-river migrating fish is D = SOT/SOI.  D with a value less than 1 represents higher latent mortality in transported fish than in in-river migrating fish.  D for the SAR from Bonneville Dam to Lower Granite Dam from 1997 to 1999 was 0.62 (Bouwes et al. 2001). For the SAR of transported fish from Lower Granite Dam, the mean from 1997 to 1999 was 0.77. Overall, the geometric mean of D from 1994 to 2003 (excluding 2001) was 0.48 (Berggren et al. 2005). The lower SARs of transported fish relative to in-river migrants is thought to be caused by latent mortality.

Indices of mortality
Determining indices of latent mortality can be difficult. Stress can be exhibited at the organismal level with stress hormones (generalized stress response), or at the cellular level with stress proteins (cellular stress response, Hightower 1991). Stress and resistance to stress has been measured by cortisol, catecholamines, blood glucose, and heat shock proteins to name a few (Bonga 1997, Feder and Hofman 1999). I chose to explore heat shock protein 70kDa as a potential indicator of stress.