|Document Type:||Journal Article|
|Title:||Can autocorrelated recruitment be estimated using integrated assessment models, and how does it affect population forecasts?|
|Author:||Kelli F. Johnson, Elizabeth Councill, James T. Thorson, Elizabeth Brooks, R. D. Methot, A. E. Punt|
|Keywords:||autocorrelated recruitment, integrated stock assessment model, statistical catch at age, rebuilding plan, population forecast,|
The addition of juveniles to marine populations (termed “recruitment”) is highly variable due to variability in survival for larvae and early juvenile stages. Recruitment estimates are often large or small for several years in a row (termed “autocorrelated” recruitment). Recruitment may be autocorrelated due to numerous factors, including regime shifts and periodicity in environmental drivers affecting juvenile survival rates. The ability of stock assessments to accurately estimate the magnitude of recruitment autocorrelation, and its effect on the quality of forecasts of spawning biomass, has not generally been analyzed. We used a simulation experiment to evaluate how well Stock Synthesis (an “integrated” age-structured stock assessment modeling framework used extensively in the assessment of fish stocks) estimates autocorrelation in the presence of a range of levels of autocorrelation in recruitment deviations. The precision and accuracy of estimated autocorrelation, and the ability of the stock assessment framework to forecast the true dynamics of the system, were compared for scenarios where the autocorrelation parameter within the assessment was fixed at zero, fixed at its true value, internally estimated, or input as a fixed value determined using an external estimation procedure. Estimates of autocorrelation produced by Stock Synthesis were biased toward extreme values (i.e., towards 1.0 when true autocorrelation was positive and -1.0 when true autocorrelation was negative). Less biased estimates of autocorrelation were obtained by externally estimating it from the recruitment deviations estimated within Stock Synthesis. Ignoring autocorrelation when true recruitment is autocorrelated results in poor forecast interval coverage (i.e., a large proportion of simulation replicates where true biomass is outside the predictive interval for the forecast). However, the “external estimate” of autocorrelation generally improves forecast interval coverage. Collectively, our results suggest that autocorrelation estimates have good statistical performance when calculated from the estimated recruitment deviations. However, estimates are likely to be imprecise whenever there are relatively few years of data to estimate recruitment (i.e., less than 40 years of recruitment estimates).
|Full Text URL:||http://www.sciencedirect.com/science/article/pii/S0165783616301928|
|Theme:||Recovery and rebuilding of marine and coastal species|
Characterize the population biology of species, and develop and improve methods for predicting the status of populations.
Describe the relationships between human activities and species recovery, rebuilding and sustainability.