Utilising mark-recapture data for Bayesian modelling of fish mortality


In this work, the aim was to produce a realistic assessment of yearly mortality of Archipelago Sea pike perch during the period 1997-2012. The utilized data origins from the mark-recapture experiment carried out by the Finnish Game and Fisheries Research Institute (FGFRI). In this mark-recapture experiment, returnings of the marks were based on voluntary tag reporting by the fishermen gaining small monetary rewards. In this study design, the count of returned tags is affected by the size of the release cohort, efficiency of the fishing method used by a fisherman and the fisherman’s willingness to return the tag. In addition, each year a proportion of the tags become detached from fish, which means that those tags cannot be returned. All these factors were taken into account in a hierarchical model, which was developed in the same fashion as the well-known Cormack-Jolly-Seber model. Data from the yearly total catch were not used in this work because those data will be used in the subsequent research utilizing results of this work. The objective of this work was to estimate fishing gear specific catchability coefficients and mortality rates, including natural mortality rate. The amount of data and number of parameters to be estimated set their own limitations, so it was decided to estimate parameters of interest by splitting the data into only three fishing fleets: professional fishermen, recreational net fishermen and recreational line fishermen. The estimability of the hierarchical model developed for mark-recapture data was studied using simulation experiments. One was able to find such a model configuration, where the parameters concerning mortality …

Department of Mathematics and Statistics, University of Jyväskylä

Supplementary notes can be added here, including code and math.

master's thesis animal mortality fish mortality survival modelling Bayesian modelling mark-recapture Pike perch
Juho Kopra
University Lecturer of Statistics

My research interests include Bayesian statistical methods, applied statistics for problems with high societal impact.