Natural mortality estimators for information-limited fisheries

The 29 estimators of natural mortality (M) that have been proposed for 'information-limited' fisheries are reviewed, together with a new alternative presented here. Each is applied to 13 example populations for which well-founded estimates are available of both M and the estimators' parameters. None of the 30 can provide accurate estimates for every species, and none appears sufficiently precise for use in analytical stock assessments, while several perform so poorly as to have no practical utility. If the growth coefficient K has been reliably estimated, either M = 1.5 K or Pauly's long-established estimator can provide useful estimates of M, but they fail with species that have long adult lives after swift juvenile growth, with those that never reach their asymptotic lengths and with species that otherwise deviate from archetypal teleost life histories. If a pre-exploitation maximum observed age (T-max) can be established, M can be estimated for both teleosts and sharks using M = 4.3/T-max but that seriously underestimates when the effective sample size (n(e)) is large and overestimates with species showing pronounced senescence. The new estimator presented here addresses n(e) but is upset by even mild senescence. Some estimators of M-at-size, particularly ones recently advanced by Gislason et al. and Charnov et al., also show promise but require further examination. It is recommended that fisheries scientists measure M by more advanced methods whenever possible. If 'information-limited' estimators must be used, their uncertainties should be acknowledged and their errors propagated into management advice.