Skip to main content

    Effects of recruitment variability and natural mortality on Generalised Yield Model projections and the CCAMLR Decision Rules for Antarctic krill

    Request Meeting Document
    Document Number:
    D. Kinzey, G. Watters and C. Reiss (USA)
    Submitted By:
    Sarah Mackey (CCAMLR Secretariat)

    AMLR krill sampling data was supplied to the Generalised Yield Model as the "vector of recruitments" input option to simulate the population dynamics of krill in the Antarctic Peninsula region (Subarea 48.1) under various assumptions. The annual proportions of krill less than 36 mm in length to the total captured in AMLR net samples in four sampling regions of the Antarctic Peninsula were used as proxies for recruitment variability. Simulations were run for 21 years with either no fishing, or with fishing at either the trigger level (gamma = 0.0103), the precautionary catch limit (gamma = 0.093), or half the precautionary catch limit (gamma = 0.0465). Simulations were assigned natural mortalities at either the "base case" value (M = 0.8, or 45% annual survival), "variable mortality" (M with a uniform distribution between 2 and 0.8, or annual survivals varying between 14 and 45%) and "high mortality" (M = 3, annual survival of 5%). CVs of either 0, 10%, 20%, or 30% were added to the observed recruitment values.

    The CCAMLR "depletion" decision rule was more susceptible to being triggered by these modifications to the GYM inputs than was the "escapement" rule. For the base case simulations with M=0.8 and additional recruitment CVs of 0, simulated populations based on recruitment vectors from all four sampling areas were able to support the trigger level of catch with a less than 10% chance of the population spawning biomass falling below 20% of its unfished value, meeting the CCAMLR "depletion" decision rule for recruitment. At the higher catches of the precautionary catch limit, populations based on recruitment vectors from two of the four areas, Elephant Island and the Western Area, were unable to support the catch while maintaining spawning biomasses above 20% of unfished biomass in more than 90% of the trials. As the values for natural mortality and additional recruitment variability were increased beyond the "base case" values, fewer of the simulation scenarios were able to achieve the CCAMLR "depletion" decision rule.