Addressing Concerns for Water Quality Impacts
from Large-Scale Great Lakes Aquaculture

Based on a Roundtable
Co-hosted by the Habitat Advisory Board of the Great Lakes Fishery Commission
and
the Great Lakes Water Quality Board of the International Joint Commission

August, 1999


Appendix 11

British Columbia Experiences

by A.J. Castledine
British Columbia Ministry of Fisheries, P.O. Box 9359, FTN Provincial Government, Victoria, B.C., V8W 9M2
E-mail: al.castledine@gems5.gov.bc.ca

Introduction

In 1997, salmon farming production in British Columbia reached approximately 35,000 tonnes valued at about $225 million (Cdn.). Farmed salmon is B.C.'s largest agricultural export. The salmon farming industry has 100 marine sites and 19 freshwater sites - 17 hatcheries and 2 lake net pen sites - for production of 12-14 million smolts. Atlantic, chinook and coho salmon are produced. Some individual farms are licenced to produce in excess of 1,000 tonnes on an annual basis. Freshwater production of trout and charr for the table market is less than 250 tonnes annually. Experimental work on halibut and blackcod is in progress. Direct and indirect employment attributed to the industry exceeds 2,500 full time positions.

Marine feed conversion maximum is 1.5:1 so total feed utilized is about 60,000 tonnes. Feeds are nutrient dense. Delivery systems vary and are designed to optimize uptake. On a mass balance basis, input of nitrogen to coastal waters is a small fraction of inputs from sources such as the Fraser River.

In 1986, farm siting criteria included: avoidance of areas of "significant" fish habitat; distribution of farms to be a minimum of 3 km apart; a minimum of 3 cm/sec average flow; consideration of depth, current velocities, and farm orientation; use of a rudimentary model to predict when assimilative capacity would be exceeded; and low farm production level - circa 200 tonnes.

Farms were located for economic reasons - distance to market, supplies, etc. - as much as for biophysical reasons. There was a lot of feeding the water and not the fish-feed conversions of 6:1 were not unheard of. Accumulation of sediments occurred to the detriment of the farm itself in a few cases. There was no indication that dissolved nutrients were affecting algal production, especially toxic algal blooms. Several studies have shown that the areas impacted by sediments are remediated by natural processes over a period of months to years. Industry moved to cooler northern waters, cage design and anchor systems improved thus allowing use of more active sites.

In the mid 1990s, improvements in feeds and feeding practices - feedback and video monitoring systems - have reduced levels of impact. Fallowing is used when possible as are single year-class sites. Cross (1996) summarized observations from a series of investigations on sedimentation under farms and determined that production has increased several fold on individual farms while sediment accumulation has decreased. These changes are attributed to improvements in feeds and feeding practices as well as better site selection. There is a move to the use of performance-based standards to determine level of production, rather than reliance on regulated production levels.

The ability to determine optimum sites has improved. Circulation dynamics (evaluating backeddy potential and mass transport potential for the site), measuring water column current velocities to ensure optimum dispersion of waste materials and sufficient movement at the sediment-water interface, assessing bottom substrate to determine whether it is depositional or non-depositional, choosing adequate depth to make optimum use of tides and currents, and matching farm size and waste material loadings are all important considerations.

Additionally, currents should not be predominantly onshore; recommended surface currents should be >10 cm/sec, currents at 15 meters should be >5 cm/sec and be >3 cm/sec one meter above the bottom. Water depth should be >30 meters with bottom sloping offshore, biotically sensitive resources (e.g., eel grass, kelp etc.) should be avoided, and sites should not be placed in bays.

A number of models have been created to attempt to predict release of solid and dissolved outputs of farm operations and their effects. In British Columbia, a Modular Aquaculture Modelling System (MAMS)(Chandler and Carswell 1995) was designed to try to integrate and predict the effects of multiple sites to determine whether there is possible interaction between sites from dispersion of metabolic waste products, utilization of oxygen, etc. Further development of the model has been put on hold pending response to the recommendations of the provincial Salmon Aquaculture Review.

The coastal lakes of British Columbia are among the most nutrient poor and unproductive freshwaters in the world (MacIsaac and Stockner 1995). Waters are soft, neutral to slightly acidic with low phytoplankton productivity with phosphorus levels ranging from 3-6 µg/L. Lake and river fertilization is used to increase production of wild fish (e.g., the Department of Fisheries and Oceans may fertilize sockeye lakes with 40-150 mg phosphorus/meter2/year). Lake net pens are used for enhancement of wild salmon and trout. These have low densities and relatively low biomass. Two lakes are used for commercial smolt production, Georgie Lake on Vancouver Island and Lois Lake on the mainland. There is currently a moratorium on the use of additional lakes for smolt production until a response to the Environmental Assessment Review and acceptance of guidelines.

Georgie Lake was intensively studied in 1992 and 1993. Georgie Lake is 8.5 km long with a mean depth of 16.3 meters (maximum 42 m) with a surface area of 4.7 km. Sediments are naturally high in organic matter. Georgie Lake supported production of 8-14 tonnes of Atlantic salmon smolts annually between 1989 and 1994. Maximum feed utilized in this period was 21 tonnes in 1989. This study concluded that synoptic water quality studies are of limited value as effects are ephemeral and easily masked by water flow. Periphyton did accrue on sampling plates at a station as far as 200 meters away from the net pens. Periphyton samplers are a useful integrator of the effects of nutrient addition. Microbial foodwebs were enhanced as were phytoplankton and zooplankton. Nutrient loadings from the net pen operation were at the low end of areal loadings used by Department of Fisheries and Oceans to fertilize coastal sockeye lakes. Development of loading criteria for locating additional net pen sites will require the use of lake specific information.

Current lake net pen guidelines (DRAFT) include: lakes should be larger than 160 ha and preferably be landlocked; be oligotrophic with total phosphorus less than 10 µg/L at spring turnover; fish production should not lead to phosphorus concentrations above 15 µg/L; minimum depth under net pens should be 10 meters; a waste management permit may be required; there will be annual relocation of net pens to decrease sediment buildup; operational environmental monitoring will be instituted; other social and biological criteria will apply (e.g., presence of rare or endangered fish); angler effort; whether the water supply is used for domestic consumption; and absence of sockeye or kokanee salmon.

Summary

Improved husbandry and feeding practices in combination with improved siting has reduced the degree of sedimentation. In marine waters, dissolved nutrients do not appear to influence productivity to any significant extent - there is no documentation that salmon farming influences the development or severity of algal blooms. At current levels of production, net pen operations are not evidently shifting productivity in coastal lakes. The Environmental Assessment Review recommended a move to performance-based standards, rather than a regulated production limit based on biomass modeling to control sedimentation at farm sites.

References

Chandler, P.C.P. and B.C. Carswell. 1995. A modular aquaculture modelling system (MAMS) and its application to the Broughton Archipelago. Report to the British Columbia Ministry of Agriculture, Fisheries and Food.

Cross, S.F. 1993. Oceanographic characteristics of netcage culture sites considered optimal for minimizing environmental impacts in coastal British Columbia. Report prepared for the British Columbia Ministry of Agriculture, Fisheries and Food.

Cross, S.F. 1996. An assessment of marine netcage benthic impacts in the Broughton Archipelago. Report prepared for the British Columbia Ministry of Agriculture, Fisheries and Food.

MacIsaac, E.A. and J.G. Stockner. 1995. The environmental effects of lake net pen reared Atlantic salmon smolts. Report for B.C. Science Council Science and Technology Fund Research Grant #110 (T-6).

Salmon Aquaculture Review. 1997. Http://www.eao.gov.bc.ca