Genetic architecture of tripartite interactions: Legumes, rhizobia, and mycorrhizal fungi

Background/Question/Methods

The cost and benefits of pairwise mutualisms can be impacted by the presence of third party species interacting with one or both partners. Multi-player interactions may change the strength of natural selection on these pairwise interactions. One underappreciated dimension of many tripartite interactions is overlapping genetic control in the host. Shared regulatory pathways underlie the interactions of legumes with both rhizobia and arbuscular mycorrhizal fungi (AMF). Understanding the genetic architecture of plant responses to these symbionts is important in understanding the evolutionary outcomes of both mutualisms. If plant genetic variation exists in overlapping genetic regulatory pathways, then we expected to find positive phenotypic and genotypic correlations for the abundance of both symbionts in plants. We present the results of two studies. To see how AMF and rhizobium abundances are correlated in the field, we first collected 179 Chamaecrista fasciculata (partridge pea) from a remnant prairie in Illinois and measured plant biomass and abundance of each symbiont in the roots (nodule number and size for rhizobia and total number of intraradical structures for AMF). Second we grew 31 plant genotypes (USDA accessions) in the greenhouse in field soil, measuring abundance and computed genetic correlations.

Results/Conclusions

We found positive phenotypic correlations in both the field (r = ; p < 0.0001) and greenhouse (r = ; p < 0.0315) between the total number of nodules present on plants and allocation to AMF arbuscules. Neither correlation was explained by plant biomass, suggesting that positive feedbacks between plant condition and symbiont abundance alone did not drive these correlations. In the greenhouse experiment, we partitioned the variance among plant genotypes and found evidence for a positive genetic correlation between nodules and arbuscules (consistent with the phenotypic correlation; r = , p = ). These results are consistent with the presence of plant genetic variation for symbiosis in overlapping regulatory pathways. Positive phenotypic and genotypic correlations suggest that these plant symbioses are more closely linked evolutionarily than previously thought. Historical co-opting of AMF pathways by rhizobium symbiosis is well-known. The existence of shared genetic variation in contemporary natural populations, however, is novel and suggests that evolution of one symbiosis might have dramatic effects on the evolutionary trajectory of the other. For example, increased allocation to rhizobia in high-nitrogen environments would be predicted to lead to decreased investment in AMF. More generally, quantitative genetic approaches provide a framework for understanding the evolution of tripartite interactions in nature.