The quantitative use of parasite data in multistock modelling of South African sardine (Sardinops sagax)

Differences in parasite infection have previously been used to distinguish between fish stocks. We demonstrate a novel use of parasite prevalence-by-length data to inform quantitatively on stock mixing. An initial two mixing stock hypothesis proved consistent with biological and survey data, suggesting that there are different stocks of sardine off the west and south coasts of South Africa. That hypothesis assumed that only recruits moved from the west to the south stock. However, new tetracotyle-type metacercarian parasite bio-tag data indicate a need to allow older fish to move between the stocks as well. We demonstrate extension of bio-tagging to inform on the plausibility of population structure hypotheses by including parasite prevalence-by-length data in the model's likelihood. Our method enables the estimation of the magnitude of mixing between semidiscrete stocks, providing more precise estimates of annual movement. Such improved precision may be important in better informing future movement hypotheses and thereby management. Our research provides a framework to use to inform quantitatively on stock structure and movement hypotheses for other fish species with bio-tagging data.