Executive Summary

This document is the final report of a project entitled Environmental Impacts of Sea Cage Aquaculture in a Queensland Context – Hinchinbrook Channel Case Study (SD576/06) conducted by the Australian Institute of Marine Science (AIMS) as part of a research agreement with the Queensland Department of Tourism, Regional Development & Industry (DTRDI), the Queensland Department of Primary Industries and Fisheries (DPI & F) and Lyntune Pty Ltd trading as Bluewater Barramundi.

The Bluewater Barramundi farm is located in an extensive mangrove ecosystem within the Great Barrier Reef World Heritage Area (GBRWHA). This area is a Habitat Protection Zone of the Queensland Great Barrier Reef Coast Marine Park, and is within the Wet Tropics World Heritage Area. The farm is comprised of 32 synthetic mesh cages permanently moored in the main channel of Conn Creek, a side branch of Hinchinbrook Channel. It is approved to hold a maximum tonnage of 450t of barramundi, Lates calcarifer) but usually holds less than 250t. Standard guidelines for seacage farm operations and the appropriate monitoring strategies to track environmental impacts are yet to be defined in tropical regions.

An Interim Report was completed in August 2007. Part A of the Interim Report is a review of literature pertinent to the management of the environmental effects of tropical marine finfish cage culture, with emphasis on studies directly relevant to the Hinchinbrook Channel area. Higher temperatures in the tropics mean biological rate processes are higher than in temperate environments, and many tools developed for managing temperate aquaculture cannot be applied in the northern Australian tropics because of the environmental differences (highly turbid, macrotidal environments) and because of differences in biological communities. Part B of the Interim Report is a desk study of all accessible historical monitoring data required by the licence to operate, placed in the context of previous studies conducted in the Hinchinbrook area. There was insufficient scientific evidence to show that the farm has had a significant impact on the adjacent marine environment since monitoring began in 1998.

For the final report, AIMS was asked to estimate the area of influence of the farm, the carrying capacity of the environment in this area, and the fate of uneaten feed and other aquaculture wastes. AIMS was also asked to synthesise the results of this study to assist predictive modelling of the environmental impacts of sea cage aquaculture in similar areas, and to recommend and design a meaningful continuing monitoring program.

Our results indicate that:

• Water within Conn Creek is well mixed by tidal currents.

• Physical oceanographic models indicate the tidal flushing times in Conn Creek are rapid. Tidal exchange removes 60% of the water within Conn Creek within 12h during spring tides and within 24h on neap tides. This is an effective mechanism for the removal of suspended and dissolved wastes from Conn Creek.

• Seasonal climatic variation is the major factor affecting the water quality of Conn Creek.

• Overall, water quality within Conn Creek, including the farm site, conforms with Queensland Water Quality standards and there is no clear evidence of differences to similar mangrove environments in North Queensland.

• The footprint of the farm on the benthos appears to be restricted to the approval area, based on sediment chemistry and nutrient transformation processes.

• There is no evidence of accumulation of organic waste in the sediments underneath the cages.

• Phytoplankton within the water column of Conn Creek do not have sufficient assimilative capacity to absorb all dissolved wastes from the farm because the volume of the creek is too small.

• Mangroves facing Conn Creek contain N likely to have originated from farm activities, and play a significant role in nitrogen cycling within the ecosystem.

• Tidal flushing of Conn Creek (with Hinchinbrook Channel) is a vital route by which nutrients are exchanged and dissolved oxygen is replenished.

• Our results indicate that there has been no significant impact from the farm’s operation on the adjacent marine environment, despite it being in operation for over 20 years.

We estimate that wastes from the Bluewater Barramundi farm add significantly to the N budget status of Conn Creek, and in a minor way to the C budget. In spite of this, water quality standards established by the EPA for partially enclosed waters of the wet tropics region were only marginally exceeded during the wet season, and are still well within the range of values typical of undisturbed mangrove waterways in North Queensland. In the water column there is a slight degree of enrichment of dissolved N in the immediate vicinity of the farm, but tidal mixing and turbulence rapidly dissipate these nutrients. The ratio of dissolved nutrients (N and P), together with the long turnover time of the dissolved N pool in the water column, both point to nutrients not limiting primary production within the system, and therefore that the assimilative capacity of the Conn Creek water column is saturated. The reason for this is that the water volume of Conn Creek is too small to support sufficient primary producers to absorb all aquaculture wastes.

Flushing by tides is a major physical process for dissipation of aquaculture wastes from Conn Creek. Our results suggest a 60% replenishment of the water within the farm area can occur within a single tidal cycle (~12h) during spring tides. During neap tides however, this increases to two tidal cycles (~24h). The water column of Conn Creek was well mixed, both vertically through the water column and horizontally, from the mouth to the headwaters of the creek. There was, however, a trend toward lower dissolved oxygen concentration in the upper reaches of the creek compared to the mouth, as is typical in these estuarine environments. The tides caused diurnal fluctuations in water temperature, salinity, dissolved oxygen concentration and pH. Based on a hydrodynamic transport model, we predict material originating from the farm potentially disperses as much as 2km both upstream and downstream from the farm, based on the movement of passive, non diffusive, virtual particles (i.e. those not dissipated by chemical or biological processes).

During low tides during the wet season we observed dissolved oxygen concentrations falling below 2mg l^-1 for up to 3h, leading us to believe that oxygen is likely to be a limiting factor for the carrying capacity of aquaculture within Conn Creek. Accordingly, we applied two carrying capacity models based on dissolved oxygen budgets. The model predictions seem reasonable, and are consistent with the upper limits of historical production at the site, but both models suggest that the farm is currently near the maximum carrying capacity of the site. However:

• Model predictions should be interpreted with caution since these models were primarily developed for grouper aquaculture in SE Asia.

• There are no documented cases of mortality as a direct result of oxygen depletion at the Bluewater Barramundi farm, probably because the site is very well flushed.

• There is no published information on the critical oxygen concentrations for barramundi – making it impossible to fully understand the implications of the periodically low oxygen concentrations in Conn Creek.

• Farm management practices (e.g. stocking density, cage location and design, feeding rate and timing of feeding) will all influence carrying capacity, and have not been taken into consideration in our predictions.

We have identified a range of possible monitoring indicators that may be suitable for future environmental monitoring. The costs and benefits of these monitoring strategies should be discussed at a workshop of stakeholders to develop the design of the monitoring program, since in our opinion conventional monitoring strategies are inappropriate in macrotidal tropical mangrove estuaries such as Conn Creek.