Lardeau and Lower Duncan River Juvenile Rainbow Trout Analysis 2014

Data Preparation

The snorkel count data was provided by Redfish Consulting Ltd. The Ministry of Forests, Lands and Natural Resource Operations (MFLNRO) provided the spawner escapement estimates for Gerrard.

Rainbow trout with a fork length \(\leq\) 100 mm are age-1.

Statistical Analysis

Hierarchical Bayesian models were fitted to the snorkel and stock-recruitment data using R version 3.1.0 (Team, 2013) and JAGS 3.4.0 (Plummer, 2012) which interfaced with each other via jaggernaut 1.8.2 (Thorley, 2014). For additional information on hierarchical Bayesian modelling in the BUGS language, of which JAGS uses a dialect, the reader is referred to Kery and Schaub (2011) pages 41-44.

Unless specified, the models assumed vague (low information) prior distributions (Kéry and Schaub, 2011, p. 36). The posterior distributions were estimated from a minimum of 1,000 Markov Chain Monte Carlo (MCMC) samples thinned from the second halves of three chains (Kéry and Schaub, 2011, pp. 38-40). Model convergence was confirmed by ensuring that Rhat (Kéry and Schaub, 2011, p. 40) was less than 1.1 for each of the parameters in the model (Kéry and Schaub, 2011, p. 61).

The posterior distributions of the fixed (Kéry and Schaub, 2011, p. 75) parameters are summarised in terms of a point estimate (mean), lower and upper 95% credible limits (2.5th and 97.5th percentiles), the standard deviation (SD), percent relative error (half the 95% credible interval as a percent of the point estimate) and significance (Kéry and Schaub, 2011, p. 37,42).

The results are displayed graphically by plotting the modeled relationships between particular variables and the response with 95% credible intervals (CRIs) with the remaining variables held constant. In general, continuous and discrete fixed variables are held constant at their mean and first level values respectively while random variables are held constant at their typical values (expected values of the underlying hyperdistributions) (Kéry and Schaub, 2011, pp. 77-82). Plots were produced using the ggplot2 R package (Wickham, 2009).

Abundance

The abundance was estimated using a mark-resight-based binomial mixture model (Kéry and Schaub, 2011, pp. 134-136,384-388).

Key assumptions of the abundance model include:

• Lineal density (fish/m) varies with useable width.
• Lineal density (fish/m) varies randomly with year within river and site.
• Efficiency (probability of capture) varies by visit type (standard versus presence of marked fish).
• Marked and unmarked fish have the same probability of being counted.
• There is no tag loss, mortality or misidentification of fish.
• Sites are closed.
• The actual abundance at a site on a visit is described by an over-dispersed Poisson distribution- The number of marked and unmarked fish observed at a site are described by a binomial distributions.

Stock-Recruitment

The relationship between the number of spawners and the number of age-1 recruits the following spring was estimated using a Beverton-Holt stock-recruitment model.

Key assumptions of the stock-recruitment model include:

• The prior probability for the number of recruits per spawner at low density (R0) is normally distributed with a mean of 500 and a SD of 250.
• The residual variation in the number of age-1 recruits is log-normally distributed.

The mean prior probability of 500 for R0 was based on an average of 8,000 eggs per female spawner, a 50:50 sex ratio, 50% egg survival, 50% post-emergence fall survival and 50% overwintering survival.

Model Code

The JAGS model code, which uses a series of naming conventions, is presented below.

model {

  bEfficiency0 ~ dnorm(0, 2^-2)

  bEfficiencyVisitType[1] <- 0
  for (vt in 2:nVisitType) {
    bEfficiencyVisitType[vt] ~ dnorm(0, 2^-2)
  }

  bDensity0 ~ dnorm(0, 5^-2)

  bDensityWidth ~ dnorm(0, 2^-2)

  sDensityRiverYear ~ dunif(0, 5)
  for (rv in 1:nRiver) {
    for (yr in 1:nYear) {
      bDensityRiverYear[rv, yr] ~ dnorm(0, sDensityRiverYear^-2)
    }
  }

  sDensitySite ~ dunif(0, 5)
  for (st in 1:nSite) {
    bDensitySite[st] ~ dnorm(0, sDensitySite^-2)
  }

  sDispersion ~ dunif(0, 5)

  for (ii in 1:length(Total)) {

    logit(eEfficiency[ii]) <- bEfficiency0 
                            + bEfficiencyVisitType[VisitType[ii]] 

    dEfficiencyMark[ii] <- min(eEfficiency[ii], Marked[ii])
    dMarked[ii] <- max(Marked[ii], 1)

    Resighted[ii] ~ dbin(dEfficiencyMark[ii], dMarked[ii])

    log(eDensity[ii]) <- bDensity0
                        + bDensityWidth * log(Width[ii])
                        + bDensityRiverYear[River[ii],Year[ii]] 
                        + bDensitySite[Site[ii]] 

    eAbundance[ii] <- eDensity[ii] * SiteLength[ii] * SurveyProportion[ii]

    eDispersion[ii] ~ dgamma(1/sDispersion^2, 1/sDispersion^2)
    eTotal[ii] ~ dpois(eAbundance[ii] * eDispersion[ii])

    Total[ii] ~ dbin(eEfficiency[ii], eTotal[ii])
  }
}

model {
  R0 ~ dnorm(8000 * 0.5^4, 250^-2) T(0,)
  k ~ dunif(5 * 10^4, 2.5 * 10^5)
  sRecruits ~ dunif(0, 5)

  for (ii in 1:length(Stock)) {
    eRecruits[ii] <- R0 * Stock[ii] / (1 + Stock[ii]*(R0-1)/k)

    Recruits[ii] ~ dlnorm(log(eRecruits[ii]), sRecruits^-2)
  }
}

Model Parameters

The posterior distributions for the fixed (Kery and Schaub 2011 p. 75) parameters in each model are summarised below.

Figures

Abundance

Abundance - Age-1

Stock Recruitment