Principal Investigator: Adel Elghafghuf
Co-Investigators: Henrik Stryhn, Raphael Vanderstichel & Sophie St-Hilaire
Sea lice are one of the most significant threats facing salmon farming in eastern Canada. Stocked year-round with thousands of fish in cages, fish farms are prime breeding grounds for lice. Infestations on farms significantly increase the numbers of lice in surrounding waters, above what would occur naturally.
Numerous analytic approaches have been used to estimate and predict the transmission patterns of sea lice and extract useful information on this parasite. The abundance of sea lice on a farm is affected by transmission of the parasite between and within farms. Accounting for the networking effects of parasite transmission in the analysis is challenging, and requires advanced statistical techniques.
In this proposal, we will extend our current project on state-space modeling of sea lice, which considers within-farm transmission of sea lice only, to include external sources of sea lice (by adding transmission of lice between farms). This project builds on our current state-space modeling, and the proposed model will be applied to data from active salmon farms between July 2009 and February 2015 in Grand Manan, Bay of Fundy, New Brunswick, Canada.
H1. The abundance of juvenile sea lice on a farm is strongly associated with the abundance of adult lice on neighboring farms.
H2. The model’s predictive ability will be improved by taking into account variation at the population level, which may be due to the fluctuations of the true sea lice population, and variation at the observation level, which may be attributable to the measurement of outcome.
- To add complexity to the current models by modeling the external infection pressure of sea lice from neighboring farms.
- To apply the model developed in Objective 1 to sea lice data from the Canadian aquaculture industry to quantify internal and external infection pressure of sea lice on salmon farms in a well-defined area.
- Assess the predictive ability of the model.