The clam and the dam: A Bayesian belief network approach to environmental flow assessment in a data scarce region

The Volta clam, Galatea paradoxa, is a freshwater macrobenthic bivalve which is endemic to the Lower Volta River in Ghana. The range of occurrence of the clam has been influenced by the flow regime in the Lower Volta which is in turn controlled by operation of two dams located upstream. Previous research has documented the changes to the Lower Volta due to the dams and attempts have been made to design environmental flows (e-flows), freshwater flows to sustain ecosystems, to inform the re-operation of the dams. The past attempts were based on the pre-dam, natural flow regime of the Lower Volta. In this study, a designer e-flow approach is explored using the Volta clam as an indicator species. Using knowledge garnered from various sources on the lifecycle, habitat and the local conditions corresponding to historical and current states of the Volta clams, the factors influencing its extent are visualized and quantified in a Bayesian belief network (BBN). Based on this BBN, an e-flow recommendation for the Lower Volta is for low flows, between 50 m³/s and 330 m³/s, for four months during the Volta clam veliger larva and recruitment life stages which occur in November to March. In addition, it is recommended that full breaching of the sandbar which regularly builds up at the Volta Estuary is done annually and that sand winning on the river bed is prohibited. These e-flow and management recommendations will have consequences for other water users and these have to be investigated, for instance by flow experiments and trade-off analysis. The results show that a BBN is potentially suitable for modelling the linkages between flows, management practices and the status of ecological indicators for the development of e-flows for highly modified rivers in data scarce regions.