BEHAVIOURAL INTERACTIONS BETWEEN FARM AND WILD SALMON: POTENTIAL FOR EFFECTS ON WILD POPULATIONS

Behavioural interactions between farm and wild fish occur at all three stages mentioned in this review. The ability of farm fish to migrate into rivers following escape from aquaculture in the ocean environment leads to interactions during breeding. As not much is known about salmon during the marine phase of their life cycle, it is difficult to assess how interactions between farm and wild fish will ultimately affect wild populations at this stage. However, farm fish show aberrant migratory patterns, most notably that they may disperse into many rivers and thus may affect more than one wild population. The effects of behavioural interactions between farm and wild fish are most evident during breeding. Farm females and mature male parr represent the most likely means of gene flow from farm to wild populations, which are enhanced by earlier ages at maturity of farm fish because of faster growth rate. However, behavioural interactions on the spawning grounds by large males and females, as well as mature male parr, can negatively influence the reproductive success of wild fish. Pure farm and hybrid offspring in the freshwater environment can effectively compete for food and space with wild individuals, and at this life stage the environmental effects of aquaculture rearing are diminished. Maternal effects heavily influence the success of farm offspring at early juvenile stages, and their survival is usually poor compared to wild fish (e.g. Fleming et al. 2000, McGinnity et al. 2003). In addition, farm juveniles are sometimes less successful at evading predators and are not usually dominant over wild fish in natural environments. While overall trends suggest that farm and hybrid fish may not behave similarly to wild fish, and indeed have lower survival (e.g., McGinnity et al. 2003), variation among studies reflects the context-dependent nature of determining whether farm fish are successful in the wild. Their effect will depend upon a number of factors, including genetic origin, rearing conditions, the number, timing, magnitude and frequency of escapes, and the state of the wild population (Hutchings 1991). Thus, risk assessment will need to focus on those factors mostly likely to generate exposure to the hazard (e.g. escape), and to influence the risk of harm given an escape and the severity of that harm (e.g., Kapuscinski 2005). It will also need to recognize and incorporate various types of uncertainty. A key outcome of this process should be risk reduction planning and implementation.