Birth, copulation, and death. That's all the facts
when you come to brass tacks.
- T. S. Eliot, Sweeney Agonistes
As the fish from Monty Python's Meaning of Life could tell you, it's hard to understand the forces shaping lifespan-- not least because many get yanked from the tank prematurely. Natural selection, of course, puts a premium on living long enough to reproduce, and should even pay dividends for seeing the offspring make it to adulthood. But in many species, including humans, the natural lifespan extends far past the reproductive age. It's very interesting in this regard that extended lifespan seems to be relatively recent in human evolutionary history (more on this later).In humans, this added lifetime has been hypothesized to contribute to Darwinian fitness via the "grandmother effect," in which post-menopausal women help out with their grandchildren, and thereby promote the survival of their own genes over more than one generation. This effect should be greatest in cases where parents (or groups) take extended care of their offspring. But this idea has been difficult to test, and in fact baboons and lions, both of which do take care of their young socially, do not display "grandmotherly" lifespans.
An article in the December PLoS Biology takes an negative test of the grandmother effect by looking at lifespan in guppies. Guppies do not take care of their young, so the grandmother effect should not affect their lifespan. Reznick et al. took advantage of closely related guppies which have made major adaptations to high- or low- predation environments. Reznick et al. measured three features of reproduction in these fish- time at first brood; brooding interval; and life after brooding (which turns out to be non-zero). Guppies from high predation environments give birth early and often, and continued reproducing longer than those from low predation environments.
What is cool, though, is that this seems to operate independently of the lifespan after the last brood; so that in fact the guppies adapted to high predation lived longer. It's as if the extra reproductive rounds, necessary for life on the edge, were just plopped into the middle of the guppy lifespan (you can see this in Figure 4 of the paper). More to the point for the grandmother effect, lifespan of individuals after the the last brood was essentially stochastic, and the same in both groups(long-lived or shorter-lived). It follows a random decay, to risk a pun.
So guppies show no grandmother effect, which is predicted, since they don't care for their young. The authors of the paper point out, though, that this kind of actuarial analysis remains very hard to do for longer lived animals; and the positive presence of the grandmother effect in a place where it's expected might require some other approach.
This article was also referenced at the anti-ageing and science blog.