Get Full Government Meeting Transcripts, Videos, & Alerts Forever!
Get email alerts on the Scott River Study topic
No spam. Unsubscribe anytime.
Study links flow metrics to salmon recruitment in Scott River but finds other factors are also important
Summary
An empirical analysis of the Scott River found that appropriately selected functional flow metrics explained roughly 40% of variation in coho smolt per female and about 60% of Chinook juvenile abundance in the tested models, but authors cautioned that other environmental and oceanic factors also influence recruitment.
A study presented to the work group by Claire Koba used empirical modeling to examine how much variation in salmon recruitment in the Scott River can be explained by functional flow metrics.
Claire Koba, a postdoctoral researcher, said the team used lasso regression and multivariate adaptive regression splines (MARS) to link flow predictors from the Fort Jones gauge to ecological observations (juvenile counts and related metrics). For selected model structures the team reported it could explain about 40% of variation in coho smolt per female and roughly 60% of variation in Chinook juvenile abundance.
"About half of it might be explained with flow and the other half perhaps with other factors," Koba cautioned, noting that small ecological datasets and sensitivity to new observations warrant careful interpretation. She and collaborators emphasized that different life stages and species respond to different flow metrics: fall reconnection timing and fall flow magnitudes were important for coho, while Chinook juvenile abundance in the models was sensitive to wet‑season baseflow and spring recession characteristics.
Method and application: the study aligned seasonal functional flow metrics to life stages (e.g., coho two‑year cycles), used a long Fort Jones record (~80–85 years) to compute predictors, and tested models that included spawner abundance as an additional covariate in some structures. The presenter said the hydrologic modeling used in related HarterLab work supported Siskiyou County groundwater sustainability planning and was incorporated into interconnected surface‑water criteria for those GSPs.
Caveats and next steps: presenters emphasized model sensitivity to sample size and potential overfitting. They recommended adding more ecological observations over time, re‑running the models, and combining flow evidence with other lines of evidence (temperature, habitat, ocean conditions) to inform management decisions.
