POPULATION ECOLOGY OF SEALS: RETROSPECTIVE and PROSPECTIVE VIEWS

This review focuses on population ecology, with critical accounts of past work and future possibilities in age determination, body growth and condition, estimating abundances, mortality rates and lifespans, reproduction, comparative life histories, population dynamics, population modelling and seals in ecosystems. We suggest ways to reduce errors in age determination and to improve methods of obtaining and presenting growth data. Generalized von Bertalanffy growth equations are promoted as a basis for analysing species differences and intra-population variation in body lengths. Indices other than blubber thickness may be better for following body condition. Catch-effort and survival-index methods of estimating abundances have limited applicability, total counts are only locally useful, and sample counts may only be accurate for scattered, ice-breeding species. Some new techniques for population indices are promising. Pre-adult mortality remains difficult to assess. Although not always recognized, adult mortality rates do increase with age, as well described by Gompertz functions. Existing estimates of lifespans are unreliable, and a new approach is outlined. There are methodological problems in estimating ages of maturity. Corpora albicantia should not be used for back-extrapolation, and more study is needed of use of teeth annuli as indicators of maturity. Age-specific proportions of females parous based on reproductive tracts may disagree with proportions recruited in breeding groups, suggesting that the former may often be in error. Allometric relationships among body sizes and life-history variables need more reliable data, especially since the residuals of such relationships are of greatest interest. Brain size may be a better scalar. Direct evidence of density dependence in population growth of seals is sparse. Early survival has been more widely shown to be density-dependent, but only among polygynous species where crowding on land may be a byproduct of sexual selection; there is as yet no good evidence of trophic restraints. Evidence of density dependence of ages of maturity is generally unconvincing. Predation, especially by sharks, may be critical in some species. Characteristics of equilibrium populations might profitably be sought in mass remains in middens and historic kill sites. More attention should be paid to the search for density-independent influences. Supposed impacts of fisheries and pollutions are not wholly convincing. Natural epidemics may keep some populations below resource or space saturation, and some high-latitude species may show large year-to-year variations in recruitment and abundances. Evidence for such density-independent effects should be sought in residuals of growth curves and in teeth layers. Although surplus yield and production/biomass models have been tried, realistic pinniped models must be completely age-structured and time-dependent. Simple models have questionably assumed stationarity to derive life-history parameters. The best available estimates of density dependence of such parameters give no resolution when extrapolated toward equilibrium, and only limited efforts have been made to introduce stochasticity. Better data, not improved model structures, are needed for better understanding. Recent work has contradicted the assumed voraciousness of seals, but their system impacts and dependencies are not well understood. Extended Lotka-Volterra equations used to model Antarctic food webs, including seals, are merely heuristic. Fixed seal biomasses enter as top-down, driving functions in a Bering Sea model, which accordingly cannot be used to analyse or manage their populations. Some Soviet models are tantalizing but ill-specified. The introduction of harbor seals in well-chosen lakes might give mote insights into system roles than would more elaborate modelling. We wonder if pinniped ecology is well served by too many enthusiasts operating under too many restraints.