TABLE 1.

Summary of the response and predictor variables used in the statistical analyses. See Section 2 for a more detailed description of each variable and an outline of each model. Linear mixed effect models also included Stream ID and Family ID as random effects to account for potential nonindependence between streams and/or siblings

Variable name	Variable description
Relative telomere length	Calculated for each individual as the ratio of telomere repeat copy number to a stable copy number gene (RAG‐1) using quantitative real‐time PCR. Log10‐transformed. Used as both response and continuous predictor variable
Relative TERT expression	Calculated for each individual as the ratio of TERT gene cDNA concentration to the concentration of a stably expressing reference gene (EF1AB) using quantitative real‐time PCR. Log10‐transformed. Response variable
Treatment	“High nutrient” treatment (analogue carcass pellet addition) or “low nutrient” treatment (without pellets). Five streams per treatment. Categorical predictor variable
Sex	Individuals were assigned as male or female using a conventional PCR sexing assay. Categorical predictor variable
Body mass	Somatic mass (±1.0 mg) of each individual measured upon recapture in the field when aged 1+ (i.e., ~15 months old). Continuous predictor variable
Density	Centered_Density was calculated as the number of 1+ salmon captured within each subsection of stream (5 m long) divided by the area (m2) of that subsection. Mean_Density was also calculated for each stream. Continuous predictor variable
Flow Index	Index of water velocity. Centered_Flow Index was calculated for each stream subsection, with values ranging from 1 (100% still water) to 6 (100% torrent). Mean_Flow Index was also calculated for each stream. Continuous predictor variable
Depth	Centered_Depth was calculated for each stream subsection based on the average of 10 depth measurements (m). Mean_Depth was also calculated for each stream. Continuous predictor variable
Granulometric Index	Index of substrate size. Centered_Granulometric Index was calculated for each stream subsection, with values ranging from 1 (100% sand) to 6 (100% bedrock surface). Mean_Granulometric Index was also calculated for each stream Continuous predictor variable