Modeling timing and size of juvenile Chinook salmon out-migrants at three Elwha River rotary screw traps: a window into early life history post dam removal

Chinook salmon (Oncorhynchus tshawytscha) populations express diverse early life history pathways that increase habitat utilization and demographic resiliency. Extensive anthropogenic alterations to freshwater habitats along with hatchery and harvest impacts have led to marked reductions in early life history diversity across much of the species' range. The recent removal of two Elwha River dams between 2011 and 2014 restored access to over 90% of the available habitat that had been inaccessible to Chinook salmon since the early 1900s. This provided an opportunity to investigate how renewed access to this habitat might affect life history diversity. As exotherms, egg-to-fry development, juvenile growth, and movement are influenced by water temperatures. We used spatially and temporally explicit Elwha River water temperature and Chinook salmon spawning location data, in conjunction with spawn timing, emergence, growth, and movement models, to predict observed timing and sizes of juvenile Chinook salmon captured in three rotary screw traps in the mainstem and two tributaries during four trap years. This effort allowed us to test hypotheses regarding Elwha River Chinook salmon early life history, identify potential problems with the data, and predict how emergence and growth would change with increased spawning in the upper watershed. Predicted Chinook salmon emergence timing and predicted dates that juveniles reached 65 mm differed by as much as 2 months for different river locations due to large differences in thermal regimes longitudinally in the mainstem and between tributaries. For 10 out of the 12 trap-year combinations, the model was able to replicate important characteristics of the out-migrant timing and length data collected at the three traps. However, in most cases, there were many plausible parameter combinations that performed well, and in some cases, the model predictions and observations differed. Potential problems with the data and model assumptions were identified as partial explanations for differences and provide avenues for future work. We show that juvenile out-migrant data combined with mechanistic models can improve our understanding of how differences in temperature, spawning extent, and spawn timing affect the emergence, growth, and movement of juvenile fish across diverse riverine habitats.