Kootenay Lake Gerrard Rainbow Trout Condition and Growth Analysis 2014

The suggested citation for this analytic report is:

Thorley, J.L. (2015) Kootenay Lake Gerrard Rainbow Trout Condition and Growth Analysis 2014. A Poisson Consulting Analysis Report. URL: https://www.poissonconsulting.ca/f/723490014.

Background

The large, piscivorous Gerrard Rainbow Trout of Kootenay Lake support an important recreational fishery. Intermittently, since 1966, length, weight, fecundity and scale information has been collected from individual fish.

The primary questions the following analyses attempt to answer are:

What is the relationship between fork length and a) weight and b) fecundity?

How does body condition vary by year?

How does growth vary by year?

In addition the analysis asks

Are trends in growth and condition related to trends in kokanee abundance?

Methods

Data Preparation

The length, weight, fecundity and kokanee information was provided by RedFish Consulting Ltd. Scale ages were determined by Greg Andrusak (RedFish Consulting Ltd.), Carol Lidstone (Birkenhead Scale Analyses) and Les Fleck. Scale radii at annuli and scale lengths were provided by Greg Andrusak. The data were cleansed, tidied and manipulated prior to analysis using R version 3.1.2 (Team 2013).

Statistical Analysis

Hierarchical Bayesian models were fitted to the count data using R version 3.1.2 (Team 2013) and JAGS 3.4.0 (Plummer 2012) which interfaced with each other via jaggernaut 2.2.8 (Thorley 2013). 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, 41–44).

Unless specified, the models assumed vague (low information) prior distributions (Kery and Schaub 2011, 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 (Kery and Schaub 2011, 38–40). Model convergence was confirmed by ensuring that Rhat (Kery and Schaub 2011, 40) was less than 1.1 for each of the parameters in the model (Kery and Schaub 2011, 61). Model adequacy was confirmed by examination of residual plots.

The posterior distributions of the fixed (Kery and Schaub 2011, 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 (Kery and Schaub 2011, 37, 42).

Variable selection was achieved by dropping insignificant (Kery and Schaub 2011, 37, 42) fixed (Kery and Schaub 2011, 77–82) variables and uninformative random variables. A fixed variables was considered to be insignificant if its significance was \(\geq\) 0.05 while a random variable was considered to be uninformative if its percent relative error was \(\geq\) 80%.

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) (Kery and Schaub 2011, 77–82). Where informative the influence of particular variables is expressed in terms of the effect size (i.e., percent change in the response variable) with 95% CRIs (Bradford, Korman, and Higgins 2005).

Fecundity

The relationship between fork length and the egg count was estimated using an allometric model. Key assumptions of the fecundity model include:

  • The residual egg counts are log-normally distributed.

Condition

The relationship between body weight and fork length was estimated using an allometric mass-length model similar to He and Bence (2008). Key assumptions of the condition model include:

  • The weight varies with the length.
  • The weight varies with the day of the year as a second order polynomial.
  • The weight varies randomly with the year
  • The residual weight is log-normally distributed.

Backcalculated Growth

The relative growth in any given year was estimated from the scale increments using a polynomial model. Key assumptions of the polynomial model include:

  • The scale increment varies with the scale radius as a second order polynomial.
  • The scale increment varies randomly by year.
  • The residual variation which is log-normally distributed varies with the reader’s minimum confidence in the two annuli.

Preliminary analyses indicated that the fork length of the fish at capture and fish ID as a random effect were not informative predictors of scale increment.

Dynamic Factor Analysis

The relative similarities between the yearly condition and growth (scale increment) estimates and the yearly kokanee abundance estimates were estimated using a Dynamic Factor Analysis (Zuur, Tuck, and Bailey 2003). Due to the rotation problem the underlying trends were indeterminate (Abmann, Boysen-Hogrefe, and Pape 2014). The growth estimates were lagged by one year.

Key assumptions of the DFA model include:

  • The time series are described by two underlying trends.
  • The random walk processes in the trends are normally distributed.
  • The residual variation in the standardised variables is normally distributed.

Model Code

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

Fecundity

Variable/Parameter Description
bAlpha Intercept for log(eFecundity)
bBeta Slope for log(eFecundity[i])
Fecundity[i] Fecundity of ith fish
sFecundity SD of residual variation in log(Fecundity)
Fecundity - Model1
model {

  bAlpha ~ dnorm(5, 5^-2)
  bBeta ~ dnorm(1, 5^-2)

  sFecundity ~ dunif(0, 1)
  for(i in 1:length(Fecundity)) {
    log(eFecundity[i]) <- bAlpha + bBeta * Length[i]
    Fecundity[i] ~ dlnorm(log(eFecundity[i]), sFecundity^-2)
  }
}

Condition

Variable/Parameter Description
bAlpha Intercept for eAlpha
bAlphaDayte Effect of Dayte on bAlpha
bAlphaDayte2 Quadratic effect of Dayte on bAlpha
bBeta Intercept for eBeta
Dayte[i] Day of the year on which ith fish encountered
eAlpha Predicted intercept for log(eWeight)
eBeta Predicted effect of centred log(Length) on log(eWeight)
eWeight[i] Prediction weight of ith fish
Length[i] Centered log fork length of ith fish
sAlphaYear SD of effect of Year on bAlpha
sWeight SD of residual variation in log(Weight)
Weight[i] Weight of ith fish
Year[i] Year on which ith fish encountered
Condition - Model1
model {

  bAlpha ~ dnorm(5, 5^-2)
  bBeta ~ dnorm(3, 5^-2)

  bAlphaDayte ~ dnorm(0, 5^-2) 
  bAlphaDayte2 ~ dnorm(0, 5^-2)

  sAlphaYear ~ dunif(0, 5)
  for(i in 1:nYear) {
    bAlphaYear[i] ~ dnorm(0, sAlphaYear^-2)
  }  

  sWeight ~ dunif(0, 5)
  for(i in 1:length(Weight)) {
    eAlpha[i] <- bAlpha + bAlphaDayte * Dayte[i] + bAlphaDayte2 * Dayte[i]^2 + bAlphaYear[Year[i]]

    eBeta[i] <- bBeta

    log(eWeight[i]) <- eAlpha[i] + eBeta[i] * Length[i]

    Weight[i] ~ dlnorm(log(eWeight[i]), sWeight^-2)
  }
}

Age

Variable/Parameter Description
Age[i] Age for ith scale reading
bK Intercept for eK
bKReader Effect of Reader on log(bK)
bLinf Mean maximum length
bT0Reader Effect of Reader on t0
eK[i] Expected growth coefficient for ith scale reading
eT0[i] Expected length at time zero
Length[i] Fork length at capture for ith scale reading
Reader[i] Reader for ith scale reading
sKReader SD of effect of Fish on bK
sLength SD of residual variation in Length
t0 Intercept for eT0
Age - Model1
 model {
  bLInf ~ dunif(700, 1000)
  bK ~ dunif(0, 1)
  t0 ~ dnorm(0, 2^-1)

  bKReader[1] <- 0
  bT0Reader[1] <- 0
  for(i in 2:nReader) {
    bKReader[i] ~ dnorm(0, 1^-2)
    bT0Reader[i] ~ dnorm(0, 5^-2)
  }

  sKFish ~ dunif(0, 1)
  for(i in 1:nFish) {
    bKFish[i] ~ dnorm(0, sKFish^-2)
  }

  sLength ~ dunif(0, 100)    
  for (i in 1:length(Length)) {
    eLInf[i] <- bLInf
    log(eK[i]) <- log(bK) + bKReader[Reader[i]] + bKFish[Fish[i]]
    eT0[i] <-  t0 + bT0Reader[Reader[i]]
    eLength[i] <- eLInf[i] * (1 - exp(-eK[i] * (Age[i] - eT0[i]))) 
    Length[i] ~ dnorm(eLength[i], sLength^-2)
  }
}

Back-Calculated Growth

Variable/Parameter Description
bInc Intercept for log(eInc)
bIncScaleRadius Effect of ScaleRadius on bInc
bIncScaleRadius2 Quadratic effect of ScaleRadius on bInc
bSDInc Intercept for log(eSDInc)
bSDProbability Effect of Probability on bSDInc
eInc[i] Expected increment for ith annulus difference
eSDInc[i] Expected SD for for ith annulus difference
Increment[i] Measured annulus difference
Probability[i] Readers minimum confidence in ith annulus difference
sIncYear SD of effect of Year on bInc
Year[i] Inferred year of second annulus for ith annulus difference
Back-Calculated Growth - Model1
model {

  bInc ~ dnorm(0, 5^-2)
  bIncScaleRadius ~ dnorm(0, 5^-2)
  bIncScaleRadius2 ~ dnorm(0, 5^-2)

  sIncYear ~ dunif(0, 5)
  for (i in 1:nYear) {
    bIncYear[i] ~ dnorm(0, sIncYear^-2)
  }

  bSDInc ~ dnorm(0, 5^-2)
  bSDIncProbability ~ dnorm(0, 5^-2)
  for (i in 1:length(ScaleRadius)) {

    log(eInc[i]) <- bInc + bIncScaleRadius * ScaleRadius[i] + bIncScaleRadius2 * ScaleRadius[i]^2 + bIncYear[Year[i]]

    log(eSDInc[i]) <- bSDInc + bSDIncProbability * Probability[i]

    Increment[i] ~ dlnorm(log(eInc[i]), eSDInc[i]^-2)
  }
} 

Dynamic Factor Analysis

Variable/Parameter Description
bDistance[i,j] Euclidean distance between ith and jth Variable
bTrendYear[t,y] Expected value for tth trend in yth Year
eValue[v,y,t] Expected standardised value for vth Variable in yth Year considering tth trends
sTrend SD in trend random walks
sValue SD for residual variation in Value
Value[i] Standardised value for ith data point
Variable[i] Variable for ith data point
Year[i] Year of ith data point
Z[v,y] Expected weighting for vth Variable in yth Year
Dynamic Factor Analysis - Model1
model{

  sTrend ~ dunif(0, 1)
  for (t in 1:nTrend) {
    bTrendYear[t,1] ~ dunif(-1,1)
    for(y in 2:nYear){
      bTrendYear[t,y] ~ dnorm(bTrendYear[t,y-1], sTrend^-2)
    }
  }

  for(v in 1:nVariable){ 
    for(t in 1:nTrend) {
      Z[v,t] ~ dunif(-1,1)
    }
    for(y in 1:nYear){
      eValue[v,y,1] <- Z[v,1] * bTrendYear[1,y]
      for(t in 2:nTrend) {
        eValue[v,y,t] <- eValue[v,y,t-1] + Z[v,t] * bTrendYear[t,y]
      }
    }
  }

  sValue ~ dunif(0, 1)
  for(i in 1:length(Value)) {
    Value[i] ~ dnorm(eValue[Variable[i], Year[i], nTrend], sValue^-2)
  }

  for(i in 1:nVariable) {
    for(j in 1:nVariable) {
      bDistance[i,j] <- sqrt(sum((Z[i,]-Z[j,])^2))
    }
  }
}

Results

Model Parameters

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

Fecundity

Parameter Estimate Lower Upper SD Error Significance
bAlpha 8.97374 8.93630 9.01392 0.01892 0 7e-04
bBeta 1.74870 1.40080 2.10560 0.17440 20 7e-04
sFecundity 0.07758 0.05576 0.11355 0.01439 37 7e-04
Convergence Iterations
1.01 1000

Condition

Parameter Estimate Lower Upper SD Error Significance
bAlpha 1.32120 1.25860 1.39270 0.03270 5 0.0010
bAlphaDayte -0.01217 -0.02538 0.00124 0.00677 110 0.0739
bAlphaDayte2 0.02240 0.01008 0.03431 0.00610 54 0.0010
bBeta 3.26010 3.18960 3.32970 0.03690 2 0.0010
sAlphaYear 0.10060 0.06010 0.16500 0.02690 52 0.0010
sWeight 0.15795 0.15059 0.16647 0.00400 5 0.0010
Convergence Iterations
1.01 1e+05

Age

Parameter Estimate Lower Upper SD Error Significance
bK 0.17204 0.15224 0.19128 0.01034 11 0.0010
bKReader[2] -0.03520 -0.08910 0.02140 0.02790 160 0.1827
bKReader[3] 0.61110 0.49270 0.72450 0.06070 19 0.0010
bLInf 924.18000 904.69000 951.64000 11.83000 3 0.0010
bT0Reader[2] -0.19280 -0.50200 0.09670 0.15340 160 0.2129
bT0Reader[3] 2.10600 1.52600 2.81600 0.32800 31 0.0010
sKFish 0.32669 0.29840 0.35558 0.01470 9 0.0010
sLength 20.59600 17.63400 23.95700 1.54000 15 0.0010
t0 -0.85200 -1.64400 -0.28300 0.34900 80 0.0010
Convergence Iterations
1.04 80000

Back-Calculated Growth

Parameter Estimate Lower Upper SD Error Significance
bInc 1.59690 1.52200 1.66850 0.03780 5 0.0010
bIncScaleRadius 0.16324 0.13690 0.19196 0.01453 17 0.0010
bIncScaleRadius2 -0.09418 -0.11742 -0.07253 0.01144 24 0.0010
bSDInc -0.96289 -1.00488 -0.91875 0.02153 4 0.0010
bSDIncProbability 0.02469 -0.01958 0.06840 0.02263 180 0.2595
sIncYear 0.14790 0.09810 0.22190 0.03150 42 0.0010
Convergence Iterations
1.01 1e+05

Dynamic Factor Analysis

Parameter Estimate Lower Upper SD Error Significance
sTrend 0.7754 0.4984 0.9807 0.1311 31 0.001
sValue 0.6182 0.5053 0.7812 0.0694 22 0.001
Convergence Iterations
1.02 1e+05

Figures

Condition

figures/condition/length.png
Figure 1. Predicted weight by fork length for a typical year on June 1st (with 95% CRIs).
figures/condition/year.png
Figure 2. Predicted percent change in body weight by year for June 1st(with 95% CRIs).
figures/condition/dayte.png
Figure 3. Predicted percent change in body weight by day of the year for a typical year (with 95% CRIs).

Age

figures/vb/length.png
Figure 4. Predicted fork length by age by scale reader (with 95% CRIs).

Back-Calculated Growth

figures/backcalc/scaleradius.png
Figure 5. Predicted relationship between scale radius and scale increment.
figures/backcalc/year.png
Figure 6. Predicted effect of year on the scale increment of a typical fish.

Dynamic Factor Analysis

figures/dfa/fit.png
Figure 7. Predicted standardised variables values by year (with 95% CRIs).
figures/dfa/mds.png
Figure 8. Non-metric multidimensional plot showing clustering of standardised variables by trend weightings.

Acknowledgements

The organisations and individuals whose contributions have made this analysis report possible include:

  • Redfish Consulting Ltd.
    • Greg Andrusak
  • Birkenhead Scale Analyses
    • Carol Lidstone
  • Reel Adventures
    • Kerry Reed
  • Les Fleck

References

Abmann, C., J. Boysen-Hogrefe, and M. Pape. 2014. “Bayesian Analysis of Dynamic Factor Models: An Ex-Post Approach Towards the Rotation Problem.” Kiel Working Paper No. 1902. Kiel, Germany: Kiel Institute for the World Economy.

Bradford, Michael J, Josh Korman, and Paul S Higgins. 2005. “Using Confidence Intervals to Estimate the Response of Salmon Populations (Oncorhynchus Spp.) to Experimental Habitat Alterations.” Canadian Journal of Fisheries and Aquatic Sciences 62 (12): 2716–26. https://doi.org/10.1139/f05-179.

He, Ji X., James R. Bence, James E. Johnson, David F. Clapp, and Mark P. Ebener. 2008. “Modeling Variation in Mass-Length Relations and Condition Indices of Lake Trout and Chinook Salmon in Lake Huron: A Hierarchical Bayesian Approach.” Transactions of the American Fisheries Society 137 (3): 801–17. https://doi.org/10.1577/T07-012.1.

Kery, Marc, and Michael Schaub. 2011. Bayesian Population Analysis Using WinBUGS : A Hierarchical Perspective. Boston: Academic Press. http://www.vogelwarte.ch/bpa.html.

Plummer, Martyn. 2012. “JAGS Version 3.3.0 User Manual.” http://sourceforge.net/projects/mcmc-jags/files/Manuals/3.x/.

Team, R Core. 2013. “R: A Language and Environment for Statistical Computing.” Vienna, Austria: R Foundation for Statistical Computing. http://www.R-project.org.

Thorley, J. L. 2013. “Jaggernaut: An R Package to Facilitate Bayesian Analyses Using JAGS (Just Another Gibbs Sampler).” Nelson, B.C.: Poisson Consulting Ltd. https://github.com/poissonconsulting/jaggernaut.

Zuur, A F, I D Tuck, and N Bailey. 2003. “Dynamic Factor Analysis to Estimate Common Trends in Fisheries Time Series.” Canadian Journal of Fisheries and Aquatic Sciences 60 (5): 542–52. https://doi.org/10.1139/f03-030.