Understanding what controls dispersal of invasive species is key for predicting which areas on the landscape are susceptible to invasion and for addressing the vectors contributing to spread. Increasing evidence suggests that anthropogenic vectors may be responsible for the spread of many invaders; however long-distance dispersal is exceedingly difficult to measure using ecological methods. Here we use a population genetics approach to understand historical gene flow within a species invaded range for a noxious rangeland weed in California, Medusahead. Medusahead is particularly problematic in rangeland, because it reduces forage quality for cattle and impacts native diversity and ecosystem function. Its seeds have long barbed awns that allow it to attach and disperse effectively on cattle or vehicles. We use ddRAD (double digest restriction associated DNA) techniques, which are particularly effective at recovering thousands of SNPs (single nucleotide polymorphisms) in non-model organisms; and we assess the genetic structure and genetic diversity of Medusahead collected from 15 populations throughout CA. We hypothesize that (1) Medusahead is dispersal limited in its invaded range and (2) anthropogenic factors, such as road density and cattle grazing, promote dispersal.
Results/Conclusions
Medusahead exhibited modest amounts of genetic variability in California, which is consistent with previous research suggesting multiple introductions since it first established in the US in the late 1800s. Population structure analysis indicated that individuals grouped into four main clusters. The populations from southern CA were not very diverse, all individuals in each population were assigned to the same cluster, which is in line with the idea that Medusahead has only recently invaded southern CA. However, most populations from northern CA were composed of individuals assigned to different clusters, and admixture (mixed genetic makeup) was detected for many individuals. Despite this genetic structure, we found no evidence for isolation by distance, suggesting that Medusahead does not appear to be dispersal limited. Genetic diversity within populations (theta) was positively correlated with road density around the site, but was not significantly related to whether or not the site had been grazed. Overall, results suggest that Medusahead is effectively dispersing long distances throughout the state and that vehicles are contributing to its dispersal. Thus, this approach has identified a characteristic of sites with high invasion risk and has identified a vector that we can begin to address with preventative management.