Skip to main content

    Integrated models for Antarctic krill (Euphausia superba) using survey data from 1981–2014 in Subarea 48.1

    Request Meeting Document
    Document Number:
    D. Kinzey, G.M. Watters and C.S. Reiss (USA)
    Submitted By:
    Mr Doug Cooper (CCAMLR Secretariat)

    The integrated modeling framework for Antarctic krill (Euphausia superba)  has been extended to include estimates of krill growth consistent with survey data and to use multi-nation survey data collected from 1981 to 2014 near the Antarctic Peninsula. Four models of the population dynamics of Antarctic krill in Subarea 48.1 based on different aggregations of the data are described to illustrate the capabilities of the framework. Survey data collected by Germany (RMT8 net sampling from 1981 to 1989), the US AMLR program (IKMT net samples from 1992 to 2013 and 3-channel hydroacoustic transects from 1996 to 2011) and Peru (IKMT sampling in 2014) are organized into different temporal aggregations (annual, seasonal, monthly) and supplied to the framework to estimate population parameters.  One data set contained only the IKMT summer data. Three other datasets each contained all the RMT8, IKMT, and hydroacoustic data arranged in different temporal groupings (annual, summer-winter, monthly). The effects of the different ways of aggregating the data on the model estimates are evaluated both for the original data and using simulated data. Models based only on IKMT summer data or on all data sources aggregated annually produced similar estimates of population parameters. All models fit the simulated data closely. Three of the four models successfully reproduced the original estimates for spawning biomass and recruitment from the simulated data. The model using the most complex separation of the data into 7 different surveys failed to reproduce the correct estimates of recruitment and spawning biomass from the simulated data. Plans for extending the models and data in the upcoming year include modifying the code to make forward projections, incorporating data from the krill fisheries and observer program, and possibly begin to incorporate predator diet compositions.