Exposure to novel environmental stressors, such as chemical pesticides, has been shown to reduce population growth rates of plants, animals, and microorganisms; however, these novel stressors could also alter the evolution of exposed populations if genotypes differ in their response to the stress. Moreover, these evolutionary changes could potentially feedback to influence ecological processes including competition, predation, and mutualisms. Here we conducted experiments to examine the ecological and evolutionary consequences of a novel environmental stress, herbicide application, on the soybean (Glycine max) and rhizobia (Bradyrhizobium japonicum) mutualism.

First, using eleven isogenic rhizobium strains, we quantified variation in tolerance to glyphosate, an active ingredient of the herbicide, Roundup ®. We then examined how glyphosate-induced changes in rhizobia influenced soybean fitness and how the effects of glyphosate varied across strains by inoculating replicated soybean seedlings in the greenhouse with each rhizobium strain and imposing glyphosate (presence/absence) treatments. Second, we conducted a field experiment at the Kellogg Biological Station Long Term Ecological Research (LTER) plots to determine whether 22 years of glyphosate application have resulted in evolutionary changes in field rhizobium populations. We hypothesized that rhizobia strains would have evolved increased glyphosate tolerance in herbicide-treated plots. We isolated rhizobia strains from the herbicided and control plots and grew them in liquid media with and without glyphosate to examine their growth using a spectrophotometer. 

Results/Conclusions

Our data suggest that rapid evolution of glyphosate tolerance is likely and that these evolutionary changes could have indirect negative effects on soybean fitness. First, we found that glyphosate reduced the maximum growth rate of many, but not all rhizobia strains. Second, rhizobia collected from LTER plots confirmed that rhizobia populations in herbicide treatment plots have diverged from populations in similar no-herbicide treatments. Surprisingly, plants inoculated with rhizobium populations that had a long history of glyphosate exposure produced more nodules.

We conclude that glyphosate application could potentially influence both the ecology and evolution of field rhizobium populations. Such rapid evolutionary changes could influence soybean yields if glyphosate tolerance is costly in terms of reduced population growth rates and/or reduced nitrogen fixation. Understanding the potential negative effects of glyphosate on rhizobia and the consequences for soybean yields is crucial for sustainable agriculture, especially because the use of herbicide and herbicide-tolerant soybean has increased dramatically in the past decade