Earlier initial modelling attempts by the authors are extended to consider the situation where predator survival rates depend on other factors in addition to krill biomass. This is achieved by making the survival rates depend instead on krill "availability", where there is a random component in the relationship between krill biomass and availability. An examination of the consequences in the case of the blackbrowed albatross suggest that these other factors result in yet lesser resilience of the albatross population to the effects of a krill fishery. An approach for estimating the parameters of functional relationships between survival rates and krill biomass (or availability) is introduced. This is based on the method of moments, and is applied to series of survival rate estimates provided by WG-CEMP members for the blackbrowed albatross and Antarctic fur seal. Attempts are made to estimate the levels of krill harvesting which would result in halving the sizes of various krill predator populations, and the precision and robustness to model-misspecification of the associated estimator are investigated. Results in those regards are not too encouraging. Estimates of the functional relationships for the blackbrowed albatross and Antarctic fur seal indicate a surprisingly low resilience of these populations to the effects of a krill fishery. Discussion is needed as to whether this reflects an inappropriate modelling approach, or results from possible negative bias in estimates of survival rates from field data.