Nuclear and chloroplast microsatellites show multiple introductions in the worldwide invasion history of common ragweed, Ambrosia artemisiifolia

Abstract

Background: Ambrosia artemisiifolia is a North American native that has become one of the most problematic invasive plants in Europe and Asia. We studied its worldwide population genetic structure, using both nuclear and chloroplast microsatellite markers and an unprecedented large population sampling. Our goals were (i) to identify the sources of the invasive populations; (ii) to assess whether all invasive populations were founded by multiple introductions, as previously found in France; (iii) to examine how the introductions have affected the amount and structure of genetic variation in Europe; (iv) to document how the colonization of Europe proceeded; (v) to check whether populations exhibit significant heterozygote deficiencies, as previously observed.

Principal findings: We found evidence for multiple introductions of A. artemisiifolia, within regions but also within populations in most parts of its invasive range, leading to high levels of diversity. In Europe, introductions probably stem from two different regions of the native area: populations established in Central Europe appear to have originated from eastern North America, and Eastern European populations from more western North America. This may result from differential commercial exchanges between these geographic regions. Our results indicate that the expansion in Europe mostly occurred through long-distance dispersal, explaining the absence of isolation by distance and the weak influence of geography on the genetic structure in this area in contrast to the native range. Last, we detected significant heterozygote deficiencies in most populations. This may be explained by partial selfing, biparental inbreeding and/or a Wahlund effect and further investigation is warranted.

Conclusions: This insight into the sources and pathways of common ragweed expansion may help to better understand its invasion success and provides baseline data for future studies on the evolutionary processes involved during range expansion in novel environments.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. Map and nDNA genetic composition (based on Structure) of the studied populations.

For each population, the pie represents the membership coefficients to the five clusters inferred in Structure.

Figure 2. Figure 2. Within-population genetic diversity.

Relationship between nDNA allelic richness and cpDNA mean pairwise number of differences between individuals (the number of haplotypes in the population is indicated within brackets). The correlation was significant (P = 0.041).

Figure 3. Figure 3. Bayesian analysis performed in Structure.

A) on the overall nDNA dataset, relationship between K, lnP(D) and ΔK. B) On the overall nDNA data set, cluster partitioning of the populations at consecutive K-values from K = 2 to K = 5. Each vertical line represents one individual and the colors represent the membership coefficients to the K clusters. The clustering solutions inferred by Instruct and Structurama were highly similar. Colors are the same as in Fig. 1. C) On the North American (K = 3) and European (K = 7) datasets.

Figure 4. Figure 4. Mean pairwise F ST indices (± S. D.) estimated at nDNA loci between populations from different regions.

Regions include Central Europe (IT9 to PO1; 9 pops.), Eastern Europe (UKR to RU5; 5 pops.), western North America (Utah to Minnesota; 3 pops.) and eastern North America (Missouri to Bronx; 5 pops.).

Figure 5. Figure 5. Median-joining network of cpDNA haplotypes.

Ten haplotypes counting only one individual each (Table S1) were discarded, so that the network includes 23 haplotypes and 600 individuals. The size of each pie is proportional to the frequency of the corresponding haplotype. The colors indicate the geographical origin of the populations displaying each haplotype. Light green: western North America; light blue: eastern North America; dark green: Eastern Europe; dark blue: Western Europe. Purple: Argentina; White: Asia (China, Japan, Korea); Yellow: Australia. Black dots stand for unsampled haplotypes and each segment joining haplotypes represent one mutation. The two ellipses indicate the two areas of the network discussed in the text.

Figure 6. Figure 6. Relationship between mean pairwise differentiation indices at nDNA (F ST) and at cpDNA (N ST) loci.

The correlation was significant (P = 0.002).

Figure 7. Figure 7. Principal Coordinate Analyses.

A) At nDNA loci. B) At cpDNA loci. Dots: North America; open squares: Europe; crosses: non-European invasive populations. The percentages of variance explained by each axis are indicated within brackets.