Monitoring selenium (Se) concentrations in fish ovaries is an important tool for evaluating the ecological risk posed by Se in aquatic systems. Most guidance recommends sampling fish ovaries as closely as possible to when fish spawn on the premise that Se is mobilized from the liver to the ovary during vitellogenesis, and therefore, sampling ovaries during the early phases of oocyte maturation may underestimate egg Se concentrations at the time of spawning. In this study, we evaluated ovary Se data from two species with synchronous oocyte development (Ptychocheilus oregonensis and Prosopium williamsoni), one species with asynchronous oocyte development (Richardsonius balteatus), and one where the mode of development is unclear (Mylocheilus caurinus). A multivariate analysis of ovary Se as a function of fish sampling location, size, and gonado-somatic index (GSI) demonstrated ovary Se was strongly negatively correlated with GSI in fish species with synchronous oocyte development but only weakly correlated in a species with asynchronous development. In R. balteatus, a relationship between expressible (released) egg Se and remaining ovary Se was observed, with egg concentrations ∼54% of ovary concentrations on average. Overall, our findings suggest that current understanding of the mechanisms by which Se is maternally transferred to oocytes is not entirely correct and raises questions regarding how and when during the reproductive cycle Se is mobilized to ovaries. Further, our findings have significant implications for interpretation of ovary Se monitoring data collected from unripe fish. We developed regression-based models to correct ovary Se data that are biased by sampling females not in spawning condition and demonstrate how this bias can impact evaluation of Se risk to fish.
Read full abstract