Humans are an unusually long-lived species when compared to other mammals of a similar size, and even in comparison to our near relative primates. Further, we exhibit an extended period of life following loss of fertility, a rare form of life history that is only observed in a few other species. The grandmother hypothesis is one of the possible explanations for the evolution of extended longevity without fertility; it is a selection effect based on the ability of older individuals to assist in the survival of their descendants. Given the existence of such a mechanism, however, why is it not more widespread?
Data on historical agricultural populations and modern hunter-gatherers show that these groups enjoy significant postfertile periods. Taking an evolutionary approach, the Grandmother Hypothesis proposes that this reproductive inactivity is in fact adaptive. With the sacrifice of continued reproduction, an individual may increase their inclusive fitness by decreasing the interbirth intervals of their offspring. The care that would otherwise be put into one's own children can now be put into weaned (and increasingly independent) grandchildren, allowing their own offspring to reproduce again sooner. Otherwise put, the cost of a reduced relatedness coefficient may be outweighed by an increase in total number of grandchildren resulting from the diverted care.
A valid objection to the Grandmother Hypothesis, however, is if grandmothering can result in a higher fitness, why are significant postfertile life stages so rare? Among vertebrates in the wild, only humans, Globicephala macrorhynchus (pilot whales) and Orcinus orca (resident killer whales), have a significant proportion of individuals with such a life history. In this study, we present a model to investigate this objection. Our model assumes only that individuals transition through various life stages and that there is an average time to conception and gestation. In one of those stages, individuals have the option to provide care for a certain number of their grandchildren thereby allowing their own offspring to reproduce again sooner.
By comparing inclusive fitnesses of individuals that provide intergenerational care with those that instead continue to reproduce into old age, we arrive at a necessary condition for grandmothering to be an evolutionarily stable strategy (ESS). This condition, or stability threshold, relates the number of grandchildren that care must be given to with the ratio of the length of the first two life stages. It tells us nothing about when or how grandmothering may arise initially in a population, but places restrictions on when it will persist. We then make the observation that if a grandmother is to provide care for even one set of grandchildren, their expected postfertile stage must be sufficiently long. More precisely, for grandmothering to be adaptive, it must be the case that postfertile life exceeds the time taken to raise a weaned child to independence. If this were not the case, grandmothers would not be able to shorten their offspring's time between births by caring for some infants themselves. In this way, we derive an eligibility threshold that tells us when grandmothering is a strategy with any possible evolutionary advantage. These eligibility and stability criteria must both be satisfied for grandmothering to evolve and then, most importantly for our purposes, to persist.
Our analyses show that there is conflict between the stability and eligibility thresholds. As it becomes increasingly easier to meet one of them, it becomes increasingly harder to fulfill the other and vice versa. This conflict is, at its core, a grandparent-grandoffspring conflict analogous to parent-offspring conflicts. The result of this is that there is a narrow range over which we should expect grandmothering to evolve and then to persist. In other words, we should in fact expect grandmothering to be rare.