Paradigm of pragmatism: managing uncertainty in Australia’s small pelagic fisheries

The two main Australian fisheries for small and medium-sized pelagic fishes are the South Australian Sardine Fishery (SASF) and Commonwealth Small Pelagic Fishery (SPF).  Although small by global standards (combined catch of <50,000 t in 2018), these two fisheries are large from an Australian perspective  and have been subject to extensive public scrutiny, especially from conservation groups and the recreational fishing sector. The SASF was recently certified by the Marine Stewardship Council (MSC), with high scores, no conditions and only three minor recommendations. The SPF is currently seeking MSC accreditation, and appears likely to be certified with some conditions. The harvest strategies for both the SASF and SPF specify that Total Allowable Catches (TACs) are set annually based on estimates of spawning biomass obtained using the Daily Egg Production Method (DEPM).  Traditional population models are not used to set TACs due to data limitations and lack of stakeholder confidence in outputs. Estimates of biomass from other methods (e.g. acoustic surveys) are not available. In this presentation, I describe how the inherent imprecision of the DEPM has been managed in the SASF and the SPF. In short, it has been addressed by 1) using statistical approaches that minimise imprecision and provide conservative (likely negatively biased) estimates of spawning biomass; 2) establishing harvest strategies (control rules) that limit the potential for methodological imprecision to cause catch limits to be set at unsustainably high levels. This precautionary approach has been successful; it was implemented because industry valued stable catches over maximum yields and recognised the need to ensure that the fisheries that were both ecologically and socially sustainable. I also present preliminary results from recent studies designed to evaluate opportunities for 1) increasing the precision of estimates of spawning biomass through improved sampling methods and 2) revising the approach used to monitor stock abundance to focus on the DEPM parameter that can be estimated most reliably (i.e. spawning area).