Biofouling (the undesirable growth and accumulation of organisms on submerged surfaces) presents a set of challenges to the salmon aquaculture industry, with excessive fouling of nets posing risks to fish health, farm infrastructure and the broader environment. To manage this, industry currently undertakes visual estimations of biofouling to decide when net cleaning is required. Currently, these visual surveys are manual and time-consuming, and nets are generally still cleaned on a set schedule, regardless of the biofouling status of individual pens.
This project successfully monitored and characterised biofouling communities across salmon farms in Tasmania's south-east, leading to greater understanding of the environmental and management factors that influence these communities and developing predictive capacity for large biofouling events. This project also employed machine learning to develop a novel AI tool for estimating the extent of biofouling on salmon pen nets in real-time using images collected by existing pen cameras. This AI tool was then further developed to distinguish and quantify the relative abundance of both harmful and benign biofouling taxa.
This project has provided industry with rapid assessment tools for biofouling quantification and management, leading to improved fish health, environmental and operational outcomes. This research has since been shared with New Zealand research counterparts at the Bioeconomy Science Institute and Cawthron Institute.
Further information on the outcomes of this project can be found on the Blue Economy website here.
Dr Robin Cappaert's full PhD thesis titled 'Biofouling associated with salmon aquaculture: dynamic systems and novel methods of assessment' can be accessed here.
