2.4.3 | Cormack–Jolly–Seber model
We estimated the predictors for apparent survival using the
Cormack–Jolley–Seber (CJS) model using the R package “marked ”
version 1.2.6 (Laake et al., 2013). The CJS model enables us to
estimate both capture probability (cp ) and apparent survival
(φ ) by including both emigration and actual survival rates from
capture histories (Lebreton et al., 1992). We constructed models
separately to estimate the factors predicting apparent survival and
capture probability. For the model predicting apparent survival
(φ ), we included FL, body condition, parasite numbers, and their
interactions (i.e. body condition × parasite numbers and FL × parasite
numbers) in June 2020 as predictors, and we constructed models that
included all possible additive and multiplicative interactions among
predictors. Since fish size was the strongest predictor for estimating
the capture probability of salmonids (Saunders et al ., 2011;
Hedger et al., 2018), we only included FL as an explanatory
variable to estimate the capture probability (cp ). The best model
with the lowest Akaike’s information criterion value (AIC: Akaike, 1983)
was selected.