Juvenile Chinook salmon Oncorhynchus tshawytscha emigrating from natal tributaries of the Sacramento River must negotiate the Sacramento-San Joaquin River Delta, a complex network of natural and man-made channels linking the Sacramento River with San Francisco Bay. Natural processes and water management actions affect the fractions of the population using the different migration routes through the delta and survival within those routes. However, estimating these demographic parameters is difficult using traditional mark-recapture techniques, which depend on the physical recapture of fish (e.g., coded wire tags). Thus, our goals were to (1) develop a mark-recapture model to explicitly estimate the survival and migration route probabilities for each of four migration routes through the delta, (2) link these route-specific probabilities to population-level survival, and (3) apply this model to the first available acoustic telemetry data of smolt migration through the delta. The point estimate of survival through the delta for 64 tagged fish released in December 2006 (Ŝ delta ¼ 0.351; SE ¼ 0.101) was lower than that for 80 tagged fish released in January 2007 (Ŝ delta ¼ 0.543; SE ¼ 0.070). We attributed the observed difference in survival between releases to differences in survival for given migration routes and changes in the proportions of fish using the different routes. Our study shows how movements among, and survival within, migration routes interact to influence population-level survival through the delta. Thus, concurrent estimation of both route-specific migration and survival probabilities is critical to understanding the factors affecting population-level survival in a spatially complex environment such as the delta.

We monitored growth and life history pathways of juvenile steelhead Oncorhynchus mykiss and compared growth rates between the upper watershed and estuary in Scott Creek, a typical California coastal stream. Growth in the upper watershed was approximately linear from May to December for age-0 fish. For passive integrated transponder (PIT) tagged, age-1þ fish, growth transitioned to a cyclic pattern, peaking at 0.2% per day during February-April, when maximum flows and temperatures of 7-128C occurred. Growth of PIT-tagged fish then slowed during August-September (0.01% per day), when temperatures were 14-188C and flows were low. During each spring, smolts (mean fork length [FL] 6 SE ¼ 98.0 6 1.2 mm) and fry migrated to the estuary; some fish remained there during summer-fall as low flows and waves resulted in seasonal sandbar formation, which created a warm lagoon and restricted access to the ocean. Growth in the estuary-lagoon was much higher (0.2-0.8% per day at 15-248C). Our data suggest the existence of three juvenile life history pathways: upper-watershed rearing, estuary-lagoon rearing, and combined upperwatershed and estuary-lagoon rearing. We present a model based upon the above data that reports size at age for each juvenile life history type. The majority of fish reaching typical steelhead ocean entry sizes (;150-250 mm FL; age 0.8-3.0) were estuary-lagoon reared, which indicates a disproportionate contribution of this habitat type to survival of Scott Creek steelhead. In contrast, steelhead from higher latitudes rear in tributaries during summer, taking several years to attain ocean entry size.

Outmigration survival of acoustic-tagged, hatchery-origin, late-fall-run Chinook salmon (Oncorhynchus tshawytscha) smolts from the Sacramento River was estimated for 5 years (2007–2011) using a receiver array spanning the entire outmigration corridor, from the upper river, through the estuary, and into the coastal ocean. The first 4 years of releases occurred during below-average river flows, while the fifth year (2011) occurred during above-average flows. In 2011, overall outmigration survival was two to five times higher than survival in the other 4 years. Regional survival estimates indicate that most of the improved survival seen in 2011 occurred in the riverine reaches of the outmigration corridor, while survival in the brackish portions of the estuary did not significantly differ among the 5 years. For the 4 low-flow years combined, survival rate in the river was lower in the less anthropogenically modified upper reaches; however, across all regions, survival rate was lowest in the brackish portion of the estuary. Even in the high-flow year, outmigration survival was substantially lower than yearling Chinook salmon populations in other large rivers. Potential drivers of these patterns are discussed, including channelization, water flow, and predation. Finally, management strategies are suggested to best exploit survival advantages described in this study.