- The mechanisms of mortality during critical life stages of fish are not well-understood and, for many species, it is not clear if the mechanisms are similar for naturally and artificially propagated individuals. For Chinook salmon (Oncorhynchus tshawytscha), natural fish potentially face negative interactions, such as competition, and survival disadvantages, such as smaller size, that may limit survival when in association with hatchery fish. To better understand the mechanisms of mortality for hatchery and natural Chinook salmon during the critical early marine residence stage, I: (1) developed a model to discriminate between hatchery and natural juveniles using otolith structure; (2) directly compared migratory patterns of hatchery and natural juveniles; and (3) determined if there was evidence for selective mortality during early marine residence. I followed two cohorts through space and time by collecting juveniles from May-September in the Columbia River estuary and off the coast in September of 2010 and 2011. I compared attributes of those juveniles when they firstentered marine waters with those of survivors after their first summer at sea. I used a combination of genetic stock identification, otolith chemistry and structure, and physical tags to determine stock of origin, size at and timing of freshwater emigration, marine growth, and production type (hatchery or natural). I focused on the subyearling life history of a federally managed genetic stock group (upper Columbia River summer and fall Chinook salmon, UCR Su/F) because: 1) it is an abundant stock group; 2) subyearlings may be more vulnerable to size-selective mortality than yearlings; and 3) it is currently impossible to assess impacts of hatchery production due to low rates of marking the hatchery fish within this stock group. The classification model included two metrics, the presence or absence of a previously unreported transfer check associated with hatchery rearing and variability in otolith increment width, and predicted production type with a 92% jack-knifed accuracy.
Overall, timing of marine entry was similar for hatchery and natural UCR Su/F juveniles, which entered marine waters from May-September with a peak in July and August in both years. Estuarine residence times were brief: 80% of the individuals captured in the estuary had resided in saline waters for < 3 days and mean estuarine residence was significantly greater (7 ± 1.3 d) in 2010 than 2011 (1 ± 0.3 d). The only clear difference was that natural individuals captured in the estuary in 2011 migrated to saline waters earlier (July 13th ± 4 d) than hatchery conspecifics (August 10th ± 6 d). However, the timing of marine entry was similar (July 27th ± 1 d) between hatchery and natural fish collected later in the ocean. This observation could be due to differential survival related to the timing of marine entry. Alternatively, estuarine
collections may not have adequately represented the emigrating population due to rapid emigration.
I documented clear spatial overlap between production types during early marine residence but no difference in median size at marine entry (100 ± 3.5 mm), size at capture (152 ± 4.0 mm), or marine growth (0.94 ± 0.1 %b l d-1). There were also no significant differences in size at marine entry between estuary and ocean collections, which indicates that size-selective mortality had not occurred. Based on both external tags and the otolith classification model, the mean percentage of natural fish in ocean collections was 17% (± 4.8) greater than in the Columbia River estuary; this finding may indicate that estuarine collections are biased to hatchery fish or, more likely, that natural fish survived at higher rates than hatchery fish. Increased survival of natural fish may be related to greater selection pressure during freshwater rearing and prior experience with predators. This study provides the first direct stock-specific comparison of juvenile migratory behavior in natural and hatchery juvenile Columbia River Chinook salmon during early marine residence. Further research is needed to determine if natural fish consistently survive better than hatchery conspecifics and, if so, determine the specific traits and behaviors that afford a survival advantage.