Bivalve condition index as an indicator of aquaculture intensity: A meta-analysis

Shellfish aquaculture implies the placement of artificial structures in the coastal environment and the alteration of natural bivalve populations, which calls for the establishment of legislative regulatory frameworks based on an ecosystem approach. One of the challenges for policy makers is the need to monitor the effectiveness of management actions. In this study, a meta-analysis across different bays, covering a large spatial scale in Atlantic Canada, was performed to test the response of two potential indicators of aquaculture intensity: (1) the bivalve growth using both Dynamic Energy Budget (DEB) and Scope For Growth (SFG) approaches, and (2) the bivalve condition index (CI = (meat weight/shell weight) × 100). Our underlying premise was that overstocking of bivalves leads to increased competition for food resources, which might ultimately have a significant effect on bivalve growth performance and the CI. Bivalve growth performance for cohorts of Mytilus edulis and Crassostrea virginica were simulated by combining satellite remote sensing (temperature and chlorophyll) with individual based models using both DEB and SFG approaches. These models were calibrated for each cohort, by adjusting the half-saturation coefficient of the food ingestion function term (X_K), which is a common parameter related to feeding behavior in both approaches. A significant correlation between X_K and lease coverage (lease area/bay area, dimensionless) was found for M. edulis. However, because of unrealistic X_K values in some M. edulis cohorts and the lack of consistent simulations for C. virginica precluded using X_K as a reliable indicator of aquaculture intensity. By contrast, according to the observed results CI emerged as a good indicator of aquaculture intensity for both species. A General Additive Model (GAM) for C. virginica provided a regression that included the initial dry meat weight as a linear term and the lease area as a non-linear term, explaining a total deviance of 59.9% in describing final CI values. The GAM for M. edulis included only non-linear terms, lease coverage, and lease area, explaining a total deviance of 61.0%. Since the CI theoretically integrates the effects of changing trophic conditions over time, the good relationship between the CI and lease/bay characteristics provides a scientific framework for its use as a reliable ecological indicator of aquaculture intensity. From an applied perspective, this finding is of relevance because the CI is easy to measure and is widely available in government and industry datasets.