The genetic analysis indicates low genetic differentiation among the populations studied since most of the specimens evaluated share the most frequent haplotype for both markers. The haplotype H1 occurred in most of the individuals for the ITS and trnL-trnF regions, respectively, covering, besides the specimen from Mexico, the four Brazilian states analyzed. A phylogeographic structure was not observed, since the same haplotype could be found in specimens from distinct countries, while distinct haplotypes were obtained from geographically close specimens.
In terms of diversity, aquatic plants generally have a lower number of species than terrestrial plants, and a low intraspecific variability (Nakamura and Kadono 2000; Amsellem et al. 2000; Lambertini et al. 2010; Lucio et al. 2019). Low intraspecific genetic variability had also been shown for Urochloa arrecta populations colonizing one reservoir in South Brazil (Diamante et al. 2020). In this study, a larger number of individuals was used as well as a wider sampling area than Diamante et al. (2020), including the states with the highest concentration of Urochloa (Latini et al. 2016), and the low genetic differentiation among populations from different locations was still observed. Somaclonal variation might be an explanation for the presence of different haplotypes, since there are records of this variation, which becomes heritable, in plants of various groups (Kaeppler et al. 2000; Bairu et al. 2011).
Some studies observed that the processes involved in species invasion can result in genetic differences between the introduced population and that from its native location (Bossdorf et al. 2005; Dietiz and Edwards, 2006). Samples of U. arrecta in this study had the same haplotype as the sample from Mexico. However, when compared to the sample from Zimbabwe, the native location, there was a haplotype difference, in which Zimbabwe sample showed a unique haplotype, different from those present in the Brazilian territory. However, there are no other sequences available for this species sampled in its native location. Therefore, it is not yet possible to infer whether the population introduced into Brazil has a low or high genetic variation when compared to specimens from the native region.
As noted in previous studies, it is common for introduced populations to show lower genetic diversity in the introduced area when they do not experience environmental pressures (Lucio et al. 2019; Diamante et al. 2020). That could explain the low genetic variability observed in this study, as well as the mode of dispersion by asexual structures, since Urochloa is able to propagate by stem fragments and stolons (Michelan et al. 2010; Amorim et al. 2015) and by rhizomes, or any other fragment that can be transported by water flow (Pott et al. 2011; Michelan et al. 2018). Furthermore, experiments related to the regeneration of U. arrecta showed that it recovers very efficiently through asexual fragments (Michelan et al. 2010). In contrast, genetic diversity can be greater than expected when the number of introduced individuals is high or a high number of introduction events occur (Estoup et al. 2016).
Our results point out that propagule dispersal is a likely means of reproduction of this species and this could explain the high frequency of a unique haplotype for both markers. In cases where the pattern of low variability occurs, besides being indicative of the presence of a single founder genotype, it may also indicate that the maintenance of the population occurs through vegetative reproduction and dispersal of propagules to different locations (Lucio et al. 2019). In addition to clonal reproduction, we speculate that the introduction of a small number of individuals may also help to explain the low genetic variability found within a given population. This process acts as a genetic bottleneck, limiting the genetic variation in invasive populations compared to populations from natural sites (Estoup et al. 2016; Diamante et al. 2020). In the case of specimens of Urochloa of this study, most of them were sampled in the upper Paraná River basin, where they could disperse by river flow, being found as far as the Itaipu Reservoir, where they share the same haplotype, although some sampling points within this reservoir were located more than 80 km apart (Diamante et al. 2020).
Although we have a limitation regarding the number of individuals in some of the areas sampled, as in the state of Mato Grosso, the presence of a widely dispersed haplotype in a large area is consistent with the expectation of clonal reproduction (Santamaría, 2002). The reproductive system of a plant determines opportunities for adaptive evolution, since it has an important influence on different population genetic parameters, including genetic recombination, effective population size, gene flow and partitioning of genetic diversity within and between populations (Barrett et al. 2008).
The success of exotic species, such as Urochloa arrecta, can be limited by the combination of different factors acting as filters (biotic, abiotic and dispersal), in which biotic resistance could be the main explanation for failure (Pearson et al. 2014); however, the species may be favored in disturbed habitats. In addition, geographic factors may contribute to the existence of clones. For example, habitat connectivity provided by the main river channel, as occurs in the Paraná River and the Itaipu reservoir, could be the explanation for the low genetic variability, at least for the specimens collected in this basin. The presence of few or unique haplotypes in a population, as evidenced in our study, does not preclude high invasiveness. In fact, some studies report that not only genetic diversity, but also phenotypic plasticity, has been an important factor to explain invasion success in plants (Richardson and Pysek, 2006).
Another evidence that can be highlighted, from the comparison of haplotype networks, is that the trnL-trnF region showed a higher haplotype diversity when compared to the ITS. This can be explained by the fact that the chloroplastidial DNA region are susceptible to accumulating mutations, since regions with a higher degree of variation have been found in the chloroplast genome (Collevatti et al. 2003; Cloutier et al. 2005). This region has already proven useful in studying the genetic variability of different aquatic plants (Salariato et al. 2010; Silva et al. 2015; Machado et al. 2016; Lucio et al. 2019; Diamante et al. 2020).
In summary, we observed that populations of U. arrecta are highly homogeneous when comparing trnL-trnF and ITS markers. These results suggest an important role of asexual reproduction for this plant, which probably spreads through fragmentation toward distant ecosystems. In addition to the data obtained, molecular studies, including other DNA regions, such as microsatellite markers, will be necessary to comprehend the low genetic variability which occurs within the Brazilian territory.