Life stage‐specific, stochastic environmental effects overlay density dependence in an Atlantic salmon population

Abstract

Atlantic salmon populations appear to fluctuate stochastically through time. It is suspected that both density‐dependent and density‐independent factors cause these fluctuations but the relative importance of each, and the life stages at which they operate, is not well known. In this study, a long‐term data set on Atlantic salmon migrants returning to the Foyle catchment, Ireland, was used to determine the role of density‐dependent and life stage‐specific environmental factors regulating population size. A Ricker density‐dependent model showed that spawning adult population size significantly predicted variation in the resultant filial generation; however, a large amount of variation (ca. 68%) remained unexplained. It was shown that environmental factors were significant in explaining some of the remaining variance and that these influences were linked to specific life stages. Three life stages—spawning and incubation, fry emergence and marine survival—were shown to have significant environmental effects that resulted in changes in the returning cohort strength. It is concluded that these life stage‐specific environmental effects are likely to contribute to the stochastic variation in population size resulting from the application of traditional stock–recruitment models. The identification and quantification of these effects should allow improved model accuracy.

Quantifying the relative importance of stock level, river temperature and discharge on the abundance of juvenile Atlantic salmon (Salmo salar)

“…The availability of unbiased long‐term environmental data on river temperature and discharge is a major constraint for understanding environmental effects on juvenile salmonid recruitment (Honkanen et al, 2019; Kanno et al, 2015). Previous studies have therefore often used coarse proxies such as air temperature or precipitation.…”

“…More serious problems can occur where studies fail to account for time varying population characteristics, such as fish size, age or sex ratios (Bacon et al, 2009; Glover, Fryer, Bacon, et al, 2018), and assume fixed population characteristics (Honkanen et al, 2019; Hvidsten et al, 2014). This can introduce temporal biases in ‘apparent’ survival (i.e., trends in stock‐recruitment residuals) that are subsequently correlated with environmental variables that also exhibit temporal trends (Franco‐Villoria, Scott, Hoey, & Fischbacher‐Smith, 2012; Hannaford, 2015; Langan et al, 2001).…”

“…The detailed census data available to this study are rare (Gurney et al, 2010). The use of coarser information where numbers are inferred indirectly from rod or commercial catches (Honkanen et al, 2019; Hvidsten et al, 2014; Neuswanger et al, 2015; Ohlberger et al, 2018) or redd counts (Elliot et al, 1997; Honkanen et al, 2019) typically results in poor estimates of stock size (Dauphin, Prévost, Adams, & Boylan, 2010; Thorley et al, 2005) with consequences for estimating appropriate stock‐recruitment relationships. Substantial time separation between census occasions (e.g., adult–adult) can also introduce noise that obscures stock effects on recruitment (Gurney et al, 2010).…”

“…Studies can also be affected where stock and recruitment census data are separated by multiple years or where census data contain multiple undifferentiated age classes. In these circumstances, environmental effects on survival and recruitment are aggregated over time and cohorts (e.g., Honkanen, Boylan, Dodd, & Adams, 2019; Hvidsten et al, 2014; Jonsson & Jonsson, 2017; Neuswanger et al, 2015). This potentially results in environmental signals going undetected due to noise introduced between census periods and compensatory mechanisms that mask environmental effects on survival.…”

“…The availability of reliable long‐term environmental data is a further constraint on understanding interannual variability in salmonid recruitment. Air temperature and rainfall are frequently used as proxies for water temperature and discharge due to a lack of suitable time‐series to complement fish observations (Honkanen, Boylan, Dodd, & Adams, 2019; Kanno et al, 2015). However at a particular site, air temperature‐water temperature relationships are typically nonlinear, seasonal (Jackson, Fryer, Hannah, Millar, & Malcolm, 2018) and vary with discharge (van Vliet, Ludwig, Zwolsman, Weedon, & Kabat, 2011).…”