查看文章

From 27–29 March 2019, a technical workshop was held at the Australian Institute of Marine Science (AIMS) in Townsville, Queensland, Australia to discuss key issues associated with the derivation of water quality guideline values (GVs). The workshop comprised fourteen participants and was guided by a pre-workshop survey distributed to the workshop invitees as well as to numerous external parties, seeking views on the most pressing issues/questions facing GV derivation. Topics covered include: issues and potential improvements in species sensitivity distribution (SSD) model distributions and model fits; accounting for and reporting uncertainty in SSDs; issues with input toxicity data for SSDs; a discussion on weighting of data points in SSDs; issues associated with small datasets; and finally discussion of some alternative approaches to SSDs for GV derivation. This report summarises the discussion around many of the identified key issues and makes recommendations on the necessary steps to improve the GV derivation methods.

In the context of model SSD distributions and model fits, a particular outcome of the discussions was the desirability of pursuing already initiated collaborations with Canadian scientists with a view to enhancing and adopting their model averaging approach to GV derivations using ssdtools Shiny app recently developed for the British Columbia Ministry of the Environment. This tool represents a promising alternative to BurrliOz, potentially alleviating some of the issues associated with poor model fits whilst representing a structured and objective approach to model fitting. Some concerns were raised that warrant …