An emerging view of coevolution in the legume–rhizobium mutualism

Mutualisms are often framed as 'delicately balanced antagonisms' (Bronstein, 1994), with the net fitness benefits to both partners potentially masking underlying conflicts of interest. How commonly symbionts evolve to 'cheat' their hosts and hosts evolve to 'sanction' or 'control' uncooperative symbionts is the subject of debate, especially in legume-rhizobium interactions (Frederickson, 2013; Kiers et al., 2003). This kind of antagonistic coevolution should result in either arms-race dynamics characterized by repeated selective sweeps or fluctuating selection dynamics that leave signatures of balancing selection in host and symbiont genomes (Frederickson, 2013; Kortright et al., 2022; O'Brien et al., 2021). In a From the Cover article in this issue of Molecular Ecology, Epstein et al. (2022) combine GWAS and population genomic analyses to assess the evidence for positive or balancing selection consistent with ongoing, antagonistic coevolution between legumes and rhizobia. They found few genomic signatures of fitness conflicts between mutualistic partners, suggesting that legume and rhizobium fitness interests are largely aligned and symbiotic traits are mostly under stabilizing selection. In combination with other recent work (e.g. Batstone et al., 2020), the results of Epstein et al. (2022) indicate that there is little ongoing fitness conflict between legumes and rhizobia that shapes host and symbiont genomes in this system. It may be time to move beyond symbiont 'cheating' and host 'control' as the dominant paradigm for understanding how partners in mutualism coevolve.