Evolution of cooperation in Ottawa
My talk here at the Evolution meetings inn Ottawa was taking 12 minutes in practice but I finished in 10. Nervous? Maybe, because I was arguing that most past measurements of fitness benefits to legumes from rhizobia are suspect, which would throw lots of results into question.
Typically, people who want to compare benefits from two different rhizobial strains inoculate host plants with each strain separately, then compare plant growth. But plants in the field are almost always nodulated by multiple strains. I showed that a plant inoculated with a slow-nodulating but highly efficient strain (which eventually provides lots of nitrogen, relative to its carbon cost) may provide no more growth than a less-efficient but faster nodulating strain, when each is inoculated alone. The problem is that plants with only the slow, efficient strain will be short of nitrogen at first, with compound-interest effects on growth. In the field, though, there will always be some fast-nodulating rhizobia. So it's not a matter of nodules vs. no nodules, but of how efficient those nodules are. So we inoculate plants with the two strains being compared, in different ratios, then measure plant growth as a function of the percent of nodules containing the focal strain. With this method, a computer model and actual data agree that an increase in growth with nodule occupancy by the focal strain indicates a more-efficient strain.
Maren Friesen suggested a similar method in a recent paper in New Phytologist. Unfortunately, her main conclusions were based on the old, one-strain-per-plant method.
Megan Frederickson gave a stimulating talk on cheating in symbionts. We have shown that soybean, alfalfa, and peas all impose fitness-reducing sanctions on rhizobia that fail to fix nitrogen, once established inside root nodules. Frederickson asked what maintains such sanctions? If they work, "cheaters" will be rare, so there will be only weak selection on legumes to maintain sanctions. She suggested that something else (such as a generalized response of sending more resources to parts of a root that supply more nitrogen) must select for the responses we call sanctions.
I agree that cheaters would be rare under consistent sanctions, which could relax selection to maintain sanctions. But sanctions may not be that consistent. See Toby Kiers' "measured sanctions" paper. And maybe only a fraction of less-beneficial rhizobia are cheaters, that is, strains that benefit from diverting resources from nitrogen fixation to their own reproduction. Many may be defective mutants, that fix less nitrogen but don't benefit thereby. Sanctions would keep any individual strain of defective mutant rare, but there could be many different strains.
Frederickson's talk stimulated a lot of useful discussion afterwards and I am looking forward to seeing her paper.