Bacterial husbandry

Even amoeba are more complex than we generally think.  A paper in this week's Nature, written about on the BBC site among other places, describes the social bacterium Dictyostelium discoideum, as having 'a primitive farming symbiosis that includes dispersal and prudent harvesting of the crop.'  That is, they farm bacteria. 

D. discoideum life cycle, from
Weiss and Buchanan, 2009

D. discoideum are a slime mold.  These amoeba live singly apparently quite satisfactorily until they sense a dangerous depletion in the local bacterial food source, at which time they signal to each other that it's time to get together and move to a lusher site.  They do this by amassing into a multi-celled organism, a slug, that then migrates -- and now it turns out that some of the cells take their own food source with them.

When they reach greener pastures, they form a fruiting body, the stage in the life cycle that produces spores. Among other things these creatures do that requires cooperation, a fair number of the cells (20%, according to the Nature paper) of the fruiting body undergo apoptosis, programmed cell death, to sculpt the organism, which means that not every cell -- and the cells come from different clones, so are not all identical -- gets to donate genes to the next generation. Why a single-celled autonomous bacterium would apparently give up so much to contribute to the greater good is a bit of a mystery to strict Darwinians, just as altruism among primates remains a mystery, but it is certainly good for the group.  There is some evidence (from Joan Strassmann's lab, the same lab that reports this farming behavior) that cells from some clones are more likely to 'cheat' than others, by contributing fewer cells to the part of the body that is most likely to undergo apoptosis, but this is apparently a rather complex trait genetically.  Other investigators have looked for this kind of behavior (we think it is a big mistake to use culture-bound terms like 'cheating' -- or even 'farming', which is less loaded but still anthropomorphic -- but that's another matter).

In any case, all of this complex behavior requires signaling and communication -- cooperation -- and this has long been known.  But the fact that they also practice bacterial husbandry has not.

We show that about one-third of wild-collected clones husband bacteria through the sporulation and dispersal process. We call these clones farmers because they carry, seed and prudently harvest their food, but the farming is primitive because no active cultivation is known.

Strassmann et al. confirmed that infection of the amoebae they called farmers was not simply accidental by treating farmers and non-farmers with an antibiotic and then assessing their consequent behavior when placed in an infected site again.  The farmers took up bacteria again, and the non-farmers remained bacteria-free.

Carrying bacteria seems to be a clone-specific trait, and there are other differences between farmer and non-farmer clones, such as that farmers consume less of the bacteria in a given site before they signal that it's ready to go, presumably to retain sufficient bacteria for transport to their next site.  And, farmer clones migrate shorter distances than non-farmers. 

This is all very interesting, in light of recent stories about leaf cutter ant colonies that accommodate their aging colony mates, and bees that can learn, and a recent story about dogs' ability to learn hundreds of words.

With this kind of demonstrated complexity -- 'well over 100 diverse genes' for cheating in tiny slime mold, for heavens' sake! -- it is surprising that anyone could find, much less would advocate the idea that there were genes 'for' behavior.  And that is particularly true in an organism with culture and learning and the plasticity that comes with learning that is known to happen in humans.  At the same time, with many contributing genes, the common story is that there may always be genes that if inactivated can 'break' the trait, or that can have minor effects on it.  Working out the distribution and importance of such effects, and their evolution, is a substantial challenge now that we can identify the variation at the gene level.