Are scientists smarter than squirrels?
"Monkey-watchers often use the word "aunt" for an adopting female." -- Richard Dawkins, The Selfish GeneThe willingness of animals to adopt and care for orphans has been shaped by past natural selection. Often, Dawkins suggested, adoption represents "misfiring of a built-in rule... a mistake that happens too seldom for natural selection to have 'bothered' to change the rule by making the maternal instinct more selective." This seems a reasonable explanation for the failure of bird parents to kick "brood parasites" out of their nests, a situation I discussed recently.
But this week's paper, by Jamieson Gorrell and colleagues, seems to show that squirrels have a more-sophisticated understanding of selfish-gene theory than I would have expected. "Adopting kin enhances inclusive ﬁtness in asocial red squirrels" was recently published in the new online journal Nature Communications. The authors analyzed five cases of orphaned squirrels being adopted, all by close relatives, and two cases where they were left to die, even though a relative had a territory nearby. In each case, they asked whether adopting would likely increase or decrease the frequency of the adopter's genes in future generations.
Closely related individuals tend to share gene variants (alleles) even if those alleles are rare in the overall population, so adopting a younger sister or a nephew who would otherwise die could increase one's genetic representation in future generations. On the other hand, adding an orphan to one's litter puts one's own offspring at somewhat greater risk. The authors were able to estimate this risk and compare it to the increased survival chances of the adoptee, weighted by its relatedness to the foster mother. If this benefit exceeds the risk, then Hamilton's rule (the fundamental equation of social evolution) predicts adoption. All of the adoptions that did occur met this criterion -- two cases were right on the line -- whereas the two potential adoptions that didn't occur failed the Hamilton's-rule test. Yet another example of squirrels solving challenging problems.
At least, that's what the data seemed to show. But the "relatedness" term in Hamilton's rule isn't necessarily equal to the relatedness we could calculate from a family tree or from genetic similarity. It would be, if helping an orphan had no negative effect on anyone outside one's current litter. But if there are more red squirrels than red-squirrel territories, then a surviving orphan may end up displacing another squirrel. So the question is, how closely related is that displaced squirrel likely to be to the adoptive mother? In the cases studied, 1/4 to 1/2 of the lactating females nearby were kin to the adopting mother. If that's a representative sample, then a surviving orphan might often end up displacing another squirrel that was as closely related to the mother as the orphan was. In such cases, the mother would have exposed her own litter to increased risk, without doing much to increase her genetic representation in future generations. Even so, the adoptive mothers aren't acting as maladaptively as Dawkins suggested (as if they adopted orphans at random), but their behavior wouldn't be optimal (by Hamilton's rule) unless there were unoccupied territories available nearby. Thanks to Dr. Carin Bondar, whose blog alerted me to this interesting paper.
Meanwhile, over at Science, Jeff Smith and colleagues propose "A generalization of Hamilton's rule for the evolution of microbial cooperation." When one cooperative act (releasing an expensive enzyme, say) benefits all microbes nearby, it's common to assume we can add up all the costs and benefits over a population. But what if twice the enzyme gives three times the benefit? The authors developed some high-powered math to deal with such problems and concluded that certain kinds of cheaters would have a harder time getting established than we would have expected from the simpler version of Hamilton's rule. Scientists are definitely smarter than squirrels, but they can't jump as well.