Rhizobia, the bacterial root-nodule symbionts of legumes, vary in net benefits to their hosts.  Some less-beneficial rhizobia are probably defective, not gaining anything from providing their hosts with less nitrogen.  Other rhizobia may be “cheaters”, diverting more resources from nitrogen fixation to their own use, relative to more-beneficial strains.

Legumes may reduce losses to defective or cheating rhizobia through “partner choice” (preventing less-beneficial strains from infecting their roots) or “sanctions” (providing less resources to root nodules housing less-beneficial strains).  As we use the term, “sanctions” does not imply a likely improvement in a strain’s performance, but rather a decrease in the relative fitness of less-beneficial strains, tending to decrease their frequency over years.  Students in my laboratory (Toby Kiers and Ryoko Oono) have previously demonstrated sanctions in soybean, pea, and alfalfa, and we have expressed doubt that plants can reliably identify less-beneficial strains approaching their roots, as partner choice would require.   A recent paper by Annet Westhoek and colleagues, published in Scientific Reports supports both conclusions.

They pointed out that rhizobial strains differing in competitiveness may give the impression that host plants are imposing partner choice.  So they made a nonfixing mutant that had all the same signalling genes as its nitrogen-fixing parent.  Sure enough, pea plants couldn’t tell the difference — the number of nodules each strain occupied was just proportional to its frequency in mixed inocula.  But pea plants did apparently impose sanctions on nonfixing nodules, which only grew half as large as those fixing nitrogen.

I wrote “apparently”, because their methods did not allow them to determine whether the smaller nodules actually contained fewer reproductive rhizobia.  There is at least one report (by Gubry-Rangin and colleagues) of nonfixing nodules being smaller (consistent with our results) but still containing as many viable rhizobia as fixing nodules (in contrast to our results).  Plants that allocate fewer resources to less-beneficial nodules help themselves, but do they also benefit their species, by reducing the release of less-beneficial rhizobia into the soil?

It’s also possible that plants might be able to recognize and exclude some less-beneficial rhizobia, even if they can’t reliably distinguish strains that only differ in nitrogen fixation.

Despite these minor quibbles, I recommend this paper, both for the high-quality data it contains and for literature review, which covers many of the important questions and key papers in this field.