This is a companion paper to Kinzey et al. (2014), who extended the integrated stock-assessment framework for Antarctic krill (Euphausia superba) to assimilate multiple sources of survey data and estimate krill growth. Kinzey et al. (2014) fitted stock-assessment models to survey data that were pooled at annual, seasonal, and monthly time scales. The data were from research cruises conducted by Germany (RMT8 net samples from 1981 to 1989), the U.S. AMLR Program (IKMT net samples from 1992 to 2013 and acoustic biomass estimates from 1996 to 2011) and Peru (IKMT samples from 2014). Results from three of the models provide a robust picture of changes in the krill population. An annual model fitted to IKMT data collected during the austral summer and another annual model fitted to data from all surveys estimated similar absolute biomasses and changes in biomass. A seasonal model estimated similar changes in biomass but lower absolute biomasses than the two annual models. A monthly model estimated different changes in biomass but predicted absolute biomasses that were similar to those from the seasonal model. All four models fitted the observed data reasonably well, but the monthly model failed basic simulation testing. Together, results from the two annual models and the seasonal model suggest that inter-annual variations in cohort strength have dominated krill dynamics in Subarea 48.1. Long-term trends (either decreasing or increasing) in krill biomass have not been apparent over the past 30+ years. Differences in scaling of biomass estimates between the annual models and the more highly resolved models suggest substantial uncertainty in absolute abundance. Management strategies that are based on absolute estimates of biomass (rather than relative estimates of biomass) may therefore be biased, possibly leading to harvests that are inconsistent with the objectives of the Convention. We propose that WG-EMM consider how the integrated assessment framework can be used to provide management advice for the krill fishery.