Although most invasive species engage in mutualism, we know little about how mutualism evolves as partners colonize novel environments. Selection on cooperation and standing genetic variation for mutualism traits may differ between a mutualism's invaded and native ranges, which could alter cooperation and coevolutionary dynamics. To test for such differences, we compare mutualism traits between invaded- and native-range host-symbiont genotype combinations of the weedy legume, Medicago polymorpha, and its nitrogen-fixing rhizobium symbiont, Ensifer medicae, which have co-invaded North America. We find that mutualism benefits for plants are indistinguishable between invaded- and native-range symbioses. However, rhizobia gain greater fitness from invaded-range mutualisms than from native-range mutualisms, and this enhancement of symbiont fecundity could increase the mutualism’s spread by increasing symbiont availability during plant colonization. Furthermore, mutualism traits in invaded-range symbioses show lower genetic variance and a simpler partitioning of genetic variance between host and symbiont sources, compared to native-range symbioses. This suggests that biological invasion has reduced mutualists’ potential to respond to coevolutionary selection. Additionally, rhizobia bearing a locus (hrrP) that can enhance symbiotic fitness have more exploitative phenotypes in invaded-range than in native-range symbioses. These findings highlight the impacts of biological invasion on the evolution of mutualistic interactions.

We generated a collection of Ensifer medicae (rhizobia) strains from field soils and used PCR to screen each strain for the presence of hrrP, a gene that can change how cooperative rhizobia are with their legume hosts. We provide the results of the PCR screen for each strain in our collection. A subset of rhizobia from this collection was inoculated onto Medicago polymorpha plants in a greenhouse experiment, and we measured plant growth benefits from inoculation (leaf count, shoot mass) and symbiosis traits (number of root nodules, nodule size, and number of colony-forming units of E. medicae within the nodules). We provide raw plant-level and nodule-level data from this experiment.

We provide four datasets as csv files and one R code file that contains code for analyzing the data in these files. The first two columns of each data file contain the metadata, "Column" and "Description", which contain details for understanding the remaining columns. The contents of each file are as follows:

• 1_PCR_data.csv contains the hrrP genotype information for each E. medicae strain in the study.
• 2_map_data.csv contains geographic information about the populations from which we sourced E. medicae strains and M. polymorpha plants.
• 3_CFU_data.csv contains colony count data from culturing of nodule contents in the greenhouse inoculation experiment.
• 4_greenhouse_data.csv contains plant-level trait data from the greenhouse inoculation experiment.
• Wendlandt_et_al_2020_Evolution_code.R contains R code for importing, processing, and analyzing data from the above four datasets. It also contains code for producing figures from these data.