An evolutionary perspective on increasing net benefits to crops from symbiotic microbes.

Plant-imposed, fitness-reducing sanctions against less-beneficial symbionts have been documented for rhizobia, mycorrhizal fungi, and fig wasps. Although most of our examples are for rhizobia, we argue that the evolutionary persistence of mutualism in any symbiosis would require such sanctions, if there are multiple symbiont genotypes per host plant. We therefore discuss methods that could be used to develop and assess crops with stricter sanctions. These include methods to screen strains for greater mutualism as resources to identify crop genotypes that impose stronger selection for mutualism. Single-strain experiments that measure costs as well as benefits have shown that diversion of resources by rhizobia can reduce nitrogen-fixation efficiency (N per C) and that some legumes can increase this efficiency by manipulating their symbionts. Plants in the field always host multiple strains with possible synergistic interactions, so benefits from different strains might best be compared by regressing plant growth or yield on each strain's abundance in a mixture. However, results from this approach have not yet been published. To measure legacy effects of stronger sanctions on future crops, single-genotype test crops could be planted in a field that recently had replicated plots with different genotypes of the sanction-imposing crop. Enhancing agricultural benefits from symbiosis may require accepting tradeoffs that constrained past natural selection, including tradeoffs between current and future benefits.

How do less-expensive nitrogen alternatives affect legume sanctions on rhizobia?

The evolutionary stability of mutualistic interactions involving multiple partners requires "sanctioning"-the ability to influence the fitness of each partner based on its respective contribution. Sanctions must be sensitive to even small differences if even slightly less-beneficial partners could gain a fitness advantage by diverting resources away from the mutualistic service toward their own reproductive fitness. Here, we test whether legume hosts sanction even mediocre N2-fixing rhizobial strains by influencing either nodule growth (which limits rhizobial cell numbers) or carbon accumulation (polyhydroxybutryate or PHB) per rhizobial cell. We also test whether sanctions depend on the availability of less-expensive nitrogen alternatives, either as nitrate or coinoculation with a more-efficient isogenic strain. We found that nitrate eliminated differences in nodule size between the mediocre and more-efficient strains, suggesting that host sanctions were compromised. However, nitrate additions also decreased PHB accumulation by the mediocre strain, which may eliminate any fitness advantages of diverting resources from N2 fixation. Coinoculation with a more-efficient strain could also compromise host sanctions if reduction in fitness from smaller nodules does not offset the potential fitness gain from greater PHB accumulation that we observed in the mediocre strain. Hence, a host's ability to sanction mediocre strains depends not only on alternative sources of nitrogen but also the relative importance of different components of rhizobial fitness.

Resource acquisition and allocation traits in symbiotic rhizobia with implications for life-history outside of legume hosts.

Resources that microbial symbionts obtain from hosts may enhance fitness during free-living stages when resources are comparatively scarce. For rhizobia in legume root nodules, diverting resources from nitrogen fixation to polyhydroxybutyrate (PHB) has been discussed as a source of host-symbiont conflict. Yet, little is known about natural variation in PHB storage and its implications for rhizobial evolution. We therefore measured phenotypic variation in natural rhizobia populations and investigated how PHB might contribute to fitness in the free-living stage. We found that natural populations of rhizobia from Glycine max and Chamaecrista fasciculata had substantial, heritable variation in PHB acquisition during symbiosis. A model simulating temperature-dependent metabolic activity showed that the observed range of stored PHB per cell could support survival for a few days, for active cells, or over a century for sufficiently dormant cells. Experiments with field-isolated Bradyrhizobium in starvation culture suggest PHB is partitioned asymmetrically in dividing cells, consistent with individual-level bet-hedging previously demonstrated in E. meliloti. High-PHB isolates used more PHB over the first month, yet still retained more PHB for potential long-term survival in a dormant state. These results suggest that stored resources like PHB may support both short-term and long-term functions that contribute to fitness in the free-living stage.