Does oyster size matter for modelling trace metal bioaccumulation?

Abstract

The aim of this study was assessing whether trace metal bioaccumulation models for oysters need to account for the effect of size-structured demography. To achieve this, we used a Bayesian generalised additive mixed modelling (BGAMM) approach. This nonparametric regression approach allowed us to estimate potentially nonlinear size effects of the oysters as well as random and structured spatial effects. Native oysters were collected from ten locations around the shoreline of Moreton Bay, a subtropical estuary in southeast Queensland, Australia. The soft-tissue of these sampled oysters were weighed (dry weight basis) and then analysed for Al, Cu, Fe, Mg, Mn and Zn. Of the six trace metals analysed, only Mg was found to have a significant (negative) effect of soft-tissue mass. The correlated spatial effects for Cu, Mn and Zn indicated that the source of these trace metals in the oysters was from anthropogenic inputs while the correlated spatial effects for Al and Mg indicated natural inputs. The correlated spatial effect for Fe was characterised by 'hotspot' concentrations in regions of Moreton Bay where algal blooms have previously occurred and consequently oysters might be useful as an indicator of algal blooms in the Bay. Overall, the absence of an observed size effect for five of the six trace metals indicates that bioaccumulation models do not necessarily need to account for size-structured demography therefore enabling simpler models to be used.