Modelling the relative risk of plastic pollution to wildlife when data are scarce: an applied approach in the Mekong and Ganges river basins

Plastic pollution of aquatic ecosystems is a growing environmental problem. Most plastic pollution originates from land-based sources and is transported via rivers, where plastic densities can exceed those at sea. Rivers in southeast Asia are regularly identified among those transporting the greatest plastic loads globally, yet are among the least studied major river systems. This is particularly concerning since they provide habitat for commercially, culturally and ecologically important species. We present a four-step risk framework to assess the relative risk to wildlife at different sections of a river, particularly where data may be sparse. We apply this method to a case study in the Ganges and Mekong rivers. The first step is to estimate litter density and abandoned fishing gear (ALDFG) to determine the relative level of pressure from ingestion or entanglement. Step two uses a risk assessment to determine the likelihood of a species interacting with a given item, and the further likelihood of a negative outcome of that encounter. Step three determines the overlap of the habitat of the species with the litter present in the environment, and the final step integrates all of these factors to present an overall relative risk to migratory species within each watershed. In the Ganges and Mekong rivers, entanglement represents the highest risk to species. The methods presented here estimate the relative risk to vertebrate fauna from litter interactions in rivers where empirical data are not available, aiming to determine (i) where in each river do we expect to find the most plastic and ALDFG and (ii) whether species suspected to be threatened by plastic pollution overlap with these predicted high-risk regions. This approach will aid decision-makers to make informed precautionary assessments bridging data gaps about relative plastic loads and associated risks in aquatic systems, until empirical data becomes available.