Downward adjustments to local catch limits for the krill fishery in Subarea 48.1

Here we compile four vignettes that, together, provide the basis for making downward adjustments to local catch limits for the krill fishery in Subarea 48.1.  We use the term local catch limit to refer to a catch limit that applies to a group of SSMUs (gSSMU), and the work presented here is based on the gSSMUs defined in another compilation of vignettes (AERD 2016a, pp. 3-13).  We propose downward adjustments to local catch limits as one component of a larger strategy for feedback management (FBM) of the krill fishery in Subarea 48.1 (see Watters et al. 2016).  Our proposal to make downward adjustments is founded on 1) a model that quantifies the survival and recruitment rates needed to maintain resilient penguin populations, 2) a leading indicator that predicts recruitment rates of penguin cohorts, 3) feasible methods to estimate the leading indicator from monitoring data, and 4) a decision rule that can be used to decrease local catch limits when penguin cohorts are expected to be relatively weak.  We quantified the survival and recruitment rates needed to maintain resilient penguin populations with a model that was fitted to mark-recapture data on Adélie penguins.  Our results show that resilient populations can be maintained when more than 10% of fledglings recruit back to their breeding populations and adult survival rates are greater than 90%.  We then fitted a Bayesian model to estimates of cohort strength and observations on breeding phenology for Adélie, chinstrap, and gentoo penguins.  Mean age at crèche can predict poor recruitment; when the chicks in a cohort crèche at a relatively young age, cohort strength is expected to be relatively weak.  We are participating in a multi-Member collaboration to monitor, with remote cameras, the mean ages at crèche of Adélie, chinstrap, and gentoo penguins at multiple sites in Subarea 48.1.  We developed a spline-based method to analyze the photographic data collected by this camera network, and camera-based estimates of age at crèche are on par with those based on visual observations taken following CEMP Standard Method A9.  We parameterized a decision rule that, if implemented, will decrease local catch limits when chicks crèche at a relatively young age and penguin cohorts are expected to be relatively weak.  We envision that this decision rule would be applied separately for each gSSMU within Subarea 48.1, using monitoring results for penguins that are known to forage within each gSSMU.  This decision rule is intended to provide the Commission with advance opportunity to proactively manage the krill fishery so that risks to dependent predators are mitigated in a manner consistent with Article II of the Convention.  Retrospective application of the decision rule to historical data suggests that catch limits in the Bransfield Strait would have been decreased about 43% of the time while catch limits for the group of coastal SSMUs in the Drake Passage and around Elephant Island would have been decreased about 32% of the time.  Adélie and chinstrap penguin populations were declining for substantial proportions of the periods over which these retrospective analyses apply.  The retrospective results should only be considered marginal results because the decision rule to decrease local catch limits is only one component of the overall FBM strategy we are proposing for Subarea 48.1.