Dietary bioaccumulation of persistent organic pollutants in the common sole Solea solea in the context of global change. Part 2: Sensitivity of juvenile growth and contamination to toxicokinetic parameters uncertainty and environmental conditions variability in estuaries

The amount of potentially toxic chemicals in a fish depends on various environmental factors, such as temperature and feeding ecology, which can be affected by Global Changes (GC). The main objective of the present work was to study the relative influence of temperature, food quality and food availability on the growth and contamination of juveniles of common sole (Solea solea), a marine flatfish species known to be a relevant indicator of the nursery quality. It focuses on two Persistent Organic Pollutants (CB153 and L-PFOS) of legacy and emerging concern, respectively. To achieve this, we used a toxicokinetic (TK) model in which toxicant flows are mechanistically predicted using a bioenergetic model based on the Dynamic Energy Budget (DEB) theory. This modelling framework was applied to juvenile sole from the Gironde estuary (SW France) and allows accounting for the influence of environmental conditions on fish biological processes involved in toxicant fluxes. To compare their respective influence on model predictions of age, length, and contamination at puberty, we included in a global sensitivity analysis: (1) environmental variability gathered from literature for this particular estuary and (2) TK parameters (i.e. assimilation efficiency AE and elimination rate k˙e) variability and uncertainty gathered from literature about each contaminant but for different fish species and experimental conditions. Then, model predictions were confronted to fish contamination measurements from the Gironde Estuary with different combinations of TK parameter values from literature. Results highlighted a key role of diet composition on fish contamination and growth while water temperature only affected growth. It stressed the need to focus on GC impact on benthic communities and their consequences on juvenile fish diet for future work on GC scenarios. Furthermore, for both chemical, the range of variability of TK parameters from experiments led to underestimated fish contaminations. The best model fits were obtained using TK parameter values from model applications: from Mounier et al. (n.d.) for CB153 (Solea solea, experiment, AE=0.8 and k˙e=0 d⁻¹) and from de Vos et al. (2008) for PFOS (food chain of the Western Scheldt estuary, The Netherlands, AE=0.8 and k˙e=0.8 10⁻².d⁻¹).