Data for Nicola Chinook Ricker stock-recruit model with environmental covariates
Data files
Dec 31, 2021 version files 17.42 KB
-
metadata_DRYAD.xlsx
12.86 KB
-
model_data_unscaled.csv
2.07 KB
-
model_data.csv
2.49 KB
Abstract
- Climate change and human activities are transforming river flows globally, with potentially large consequences for freshwater life. To help inform watershed and flow management, there is a need for empirical studies linking flows and fish productivity.
- We tested the effects of river conditions and other factors on 22 years of Chinook salmon productivity in a watershed in British Columbia, Canada.
- Freshwater conditions during adult salmon migration and spawning, as well as during juvenile rearing, explained a large amount of variation in productivity.
- August river flows while salmon fry reared had the strongest effect on productivity – our model predicted that cohorts that experience 50% below average flow in the August of rearing have 21% lower productivity.
- These contemporary relationships are set within long-term changes in climate, land use, and hydrology. Over the last century, average August river discharge decreased by 26%, air temperatures warmed, and water withdrawals increased. 17% of the watershed was logged in the last 20 years.
- Our results suggest that, in order to remain stable, this Chinook salmon population being assessed for legal protection requires substantially higher August flow than previously recommended. Changing flow regimes – driven by watershed impacts and climate change – can threaten imperiled fish populations.
Methods
This dataset is a combination of several sources:
-Spawner and recruitment data from Fisheries and Oceans Canada, corrected for unmarked hatchery returns and fishing mortality using Regional Mark Processing Centre and Pasific Salmon Commission Chinook Technical Committee data
-Summarized hydrometric data from the Water Survey of Canada
-Smolt-to-age 3 survival data for Chinook released from Spius Creek hatchery
Usage notes
This dataset was used to fit Bayesian stock-recruit models. Before analysis, the hydrometric and smolt-to-age 3 survival variables were centered and scaled to mean=0 and SD=1. We also include an un-scaled data set here for reference. See https://github.com/lukewarkentin/nicola_chinook for complete analysis codes and raw data files.
Important note: For any re-analysis, contact Chuck Parken (Chuck.Parken@dfo-mpo.gc.ca) for any data revisions/updates. CWT data can have minor changes and new analysis should include most up to date data. CWT data is used to estimate smolt-to-age 3 survival and explotiation rate.