MODELLING ANTARCTIC KRILL: SCALE, MOVEMENT AND AGE‐STRUCTURE

The organization, example outputs, and future objectives of an integrated, age‐structured model designed to estimate krill population dynamics and productivity are described. The model's capabilities are illustrated using 19 years of survey data collected by the U.S. AMLR Program around the South Shetland Islands, but it is being developed to be applicable to any region where multi‐year data on size compositions from net tows and total biomass from hydroacoustics or net tows are available. The model estimates population parameters from data based on a joint likelihood function and is being developed in stages to incorporate different data sources. Model estimates are incremented for time of year in which the data were collected, with multiple surveys at different times within a year possible. Ages are converted to lengths based on a single age‐to‐length transition matrix. Currently the model assimilates data from research surveys but data from fisheries and eventually from krill predators and environmental time series will also be incorporated. Annual movement by krill among areas is included, but, currently, the model has convergence problems (is able to make point estimates but not variance estimates) when movement is included; converge can be achieved when movement is ignored.