Inferences on the dynamics of Southern Hemisphere minke whales from ADAPT analyses of catch-at-age information

Abstract

The dispute over the last two decades in the IWC Scientific Committee as to whether inferences of utility for management purposes can be drawn from catch-at-age information for Southern Hemisphere minke whales is reviewed, particularly in the context of whether or not such data are able to reveal if this population was increasing prior to the start of major commercial harvests in the early 1970s. Butterworth et al. (1996) developed an ADAPT VPA estimation procedure to address this last question. This paper extends that procedure to take account of assumed separability of the fishing mortality matrix for the periods of the commercial and of the Japanese scientific take (although only for ages above 15 for the former). A base-case estimator is motivated from the many possible variants of the procedure, and applied to catch-at-age and survey abundance estimates for Areas IV and V, both separately and in combination. The survey estimates used include results from both international and Japanese research programmes. Bootstrap methods are used to estimate precision, and a number of sensitivity tests for the Area IV assessment are performed. Estimates are provided of the extent to which this precision is expected to improve given the further data to be collected before the end of the Japanese scientific programme (JARPA) as currently conceived; this is achieved by using the current Area IV assessment as a basis to develop an operating model of the population for evaluation (by simulation) of the information content of future data. The Area IV base-case assessment shows satisfactory behaviour under retrospective analysis, and is consistent with the separability assumptions made. It provides an estimate of 5.5%yr1 (90% confidence interval [1.4%; 9.1%]) for the historic (increasing) trend in minke whale recruitment over the period 1947-1968 prior to the exploitation of this resource. The positivity of this estimate and the associated interval is robust to a number of sensitivity tests. The point estimate of this trend for Area V is larger, but less precisely estimated. Important implications (both qualitative and quantitative) for management of the resource that follow from these results are discussed. The point estimate of age-independent natural mortality M for Area IV is 0.057yr1. The root mean square error of this estimate by the end of the JARPA programme is estimated to be about 0.022yr-1 (much of this reflecting negative bias related to assumptions concerning the slope of the commercial selectivity-at-age vector for large ages). The point estimates of M for Area V, and for the two Areas combined, are lower. A notable result of the Area IV assessment is a marked drop in recruitment from 1970 to the mid-1980s, for which some possible reasons are advanced. Patterns of inter-annual change in recruitment (as distinct from overall trends) are well estimated from the data, indicating that the availability of catch-at-age data leads to the provision of a much finer probe to detect possible links between minke whale dynamics and environmental factors than would survey estimates of total abundance alone.