WHY MENOPAUSE

George Williams proposed in 1957 that menopause evolved because, late in life, women have more to gain from child care than from continued f ertility. I develop here a quantitative model of this idea in order to determine whether the proposed benefit is in fact larger than the cos t. To make this work possible, I introduce an age-structured theory of kin selection that allows for a time delay between an act of altruism and the benefit it provides. Using this theory, I show that in age-st ructured populations, conventional inclusive fitness calculations are justified for effects on fertility only when either the population is stationary or there is no delay between cost and benefit. For effects on mortality, conventional calculations also require that donor and re cipient be affected at the same age. I then introduce two versions of Williams' idea. Model I assumes that menopause is maintained because i t reduces the risk of mortality during childbirth, thus increasing the provision of parental care. The analysis demonstrates that this model is incapable of accounting for menopause. Model II assumes that menop ause facilitates parental care by reducing the time during which a wom an is partially incapacitated by the demands of pregnancy and infant c are. This model could not be rejected. However, a definitive test will require parameter estimates that are not yet available.