Iceland request on aspects of the Genetic Intrusion Risk Assessment Framework for salmon aquaculture (GIRAF)

ICES reviewed available information on the latest version of the Genetic Intrusion Risk Assessment Framework (GIRAF). The framework is applied to evaluate the risk of genetic intrusion of aquaculture salmon into wild salmon populations in Iceland.

In general, the types of data collected to support GIRAF are appropriate, although the current focus on sampling effort following escape events is unlikely to detect the presence of all escaped farmed salmon (escapees) in rivers. ICES advises on the establishment of a systematic, in-river monitoring programme independent of escape events to improve data collection. Systematic monitoring will ensure representative sampling that accounts for river habitat, time of year, life-history stages, and the potential large dispersal and uneven distribution of farmed salmon throughout river networks.

The estimated stock size of wild Atlantic salmon currently used in GIRAF appears to include sea-ranched salmon and to be overestimated in comparison with available information on recent returns of wild Atlantic salmon to Icelandic rivers. Since an overestimation of stock size will result in an underestimation of genetic intrusion, ICES advises that assumptions and analytical procedures involved in stock‑size estimation be reviewed to ensure that GIRAF is appropriately risk averse. ICES cautions against including sea-ranched salmon in the wild stock‑size estimate.

ICES notes that prediction uncertainty and potential bias arising from some parameter estimates or missing information are currently not accounted for in GIRAF. Uncertainties in the estimated numbers of escapees are a cause of concern, and data on dispersal and survival of farmed salmon are currently based on a limited number of escape events in Iceland. ICES advises that sensitivity analyses are needed to determine how missing information, or changes in various parameters, impact the outcomes of the risk assessment in the Icelandic context.

Several small rivers and wild salmon populations that do not have a stock‑size estimate are currently not included in GIRAF. Not including all wild salmon populations can introduce bias in the risk assessment and underestimate the genetic intrusion risk to small populations. ICES advises that alternative methods to obtain population size estimates for small rivers be explored to ensure that all wild salmon populations are ultimately included in GIRAF.

ICES advises that the genetic intrusion threshold implemented in GIRAF (4% escapees relative to the wild population size) aligns with the precautionary approach. However, until various sources of uncertainty are explicity incorporated in GIRAF, ICES considers the estimation of the amount of salmon predicted to be safely reared in pens does not yet align with the precautionary approach.