Symbiont adaptation to nitrogen-addition is mediated by hosts rather than the direct effects of nitrogen in a model nutritional symbiosis.

Background/Question/Methods

Nutritional symbioses, whereby species trade resources such as carbon (C) and nitrogen (N), form keystone interactions in many ecosystems and play critical roles in Earth’s nutrient cycles. Yet, the persistence of these vital interactions is threatened by a rapidly changing resource environment. Elevated N-deposition due to the application of fertilizer is predicted to lead to the breakdown of nutritional symbioses. Previous work on the model legume-rhizobium symbiosis, wherein rhizobia fix N in exchange for legume-fixed C, found that long-term N-addition caused an evolutionary decline in the benefits rhizobia provide to their legume hosts. However, the selective agents driving this evolutionary shift remain unclear. N-addition could directly select for less beneficial rhizobia, and/or legumes may relax selection on more beneficial rhizobia, allowing less beneficial strains to proliferate. Here, we experimentally evolved replicate populations of rhizobia (28 strains total) with or without legume hosts under both N-addition and N-free conditions across four plant growing seasons, representing hundreds of rhizobium generations. At the end of the evolution experiment, we created soil slurries from each experimental pot, inoculated slurries onto naïve hosts in N-free conditions, and measured host growth.

Results/Conclusions

By independently manipulating different aspects of the selective environment (i.e., legume presence, N-addition), our experimental evolution approach allowed us to identify the drivers of symbiont evolution. When legumes had been present during experimental evolution, our results recapitulated what was observed previously: hosts grew smaller when they were inoculated with rhizobia that evolved under N-addition compared to N-free conditions. However, when legumes had been absent during experimental evolution, we found no difference in growth when hosts were inoculated with rhizobia that evolved under N-addition or N-free conditions. Overall, our results suggest that the evolutionary shift towards less beneficial rhizobia under N-addition is mediated by the indirect effects of hosts rather than the direct effects of N-addition. From a legume's perspective, acquiring N directly from the soil is relatively cheap compared to establishing and maintaining symbiosis with N-fixing rhizobia. Thus, under N-addition, legumes may invest less in symbiosis, potentially relaxing selection on beneficial rhizobia and allowing less beneficial strains to proliferate. Accounting for host-mediated drivers of symbiont evolution is therefore required to make more accurate predictions of the effects of environmental change on nutritional symbioses and their knock-on effects on ecological communities.