Ecological genomics of mutualism decline in nitrogen-fixing bacteria

Mutualisms, cooperative partnerships between species, are among the most prevalent and economically important biological interactions. The fitness benefits of mutualism are well-known to depend on the context in which the interaction occurs, and recent evidence supports the hypothesis that mutualism evolution is quite sensitive to environmental change. Thus, rapid global changes brought on by human activities could spur the degradation or breakdown of mutualistic interactions. The genomic regions underpinning mutualism evolution, and how natural selection acts differently on these regions depending on the environment, are generally unknown, though such information can shed light on the forces that maintain cooperation in nature. Bacterial mutualists in particular might evolve more quickly than other organisms in response to environmental changes, thanks to short generation times and horizontal gene transfer, providing an ideal system in which to address questions of mutualism evolution. The Heath lab and colleagues have shown recently that, in the ecologically and economically important mutualism between legumes and their nitrogen (N)-fixing rhizobia bacteria, rhizobium partner quality (fitness benefits to plants) has declined in response to 22 years of sustained N-fertilization. This has broad importance because humans have profoundly increased the rate of added N around the world. As a follow-up experiment, I conducted whole-genome sequencing and population genomic analyses of N-evolved versus unfertilized control rhizobia populations. I demonstrate that evolutionary differentiation at a key symbiosis gene region on the symbiotic plasmid (pSym) contributes to the decline of partner quality. Moreover patterns of genetic variation at selected loci were consistent with recent positive selection within N-fertilized environments. By studying the molecular population genomics of a natural bacterial population within a long-term ecological field experiment, I found that: 1) the N environment is indeed a potent selective force mediating mutualism evolution in this symbiosis, 2) rhizobium partner quality in nature is likely to be mediated by key symbiosis genes on the pSym and/or nearby loci, and 3) phenotypic change occurred at selected genes atop otherwise recombining genomes – supporting classical (“eukaryotic”) models of adaptation in bacterial populations