Introduction

Numerous aquatic macrophytes have high ecological plasticity (Thiebaut et al. 2021) and developed characteristics that allow fast dispersion, growth and reproduction (Spencer and Bowes, 1990; Hussner et al. 2021). For these reasons these plants comprise a group with many potentially invasive species (Pieterse and Murphy, 1990) and constitute a threat in ecological and economic terms to a variety of aquatic ecosystems (Kolada and Kutyla 2016; Wahl et al. 2021). Particularly in Brazil, 12 species of freshwater aquatic exotic species have been identified (Latini et al. 2016).

Among these species, representatives of Urochloa stand out for their invasive potential. These plants were disseminated in Brazilian territory as forage (Latini et al. 2016). Urochloa arrecta is a species that invades different types of aquatic ecosystems, such as ponds, canals, rivers, reservoirs and tidal habitats (Thomaz et al. 2009; Amorim et al. 2015; Diamante et al. 2020; Sato et al. 2021; Santini et al. 2022). Although Africa is the native range of this species, it is also distributed in Asia, Oceania, Europe, North and South America (GBIF.org, 2021), including anthropized areas in several Brazilian states (Fares et al. 2020a, b; Flora do Brasil 2020).

Representatives of this genus (including U. arrecta) have been responsible for negative impacts such as the reduction of the diversity of aquatic communities, replacement of native flora, increase of biotic homogenization and the blocking of canals, bridges, and water intake ducts for hydroelectric power plants (Michelan et al., 2010a; Carniatto et al. 2013; Amorim et al. 2015; Latini et al. 2016; Galvanese et al. 2022). Furthermore, U. arrecta has high plasticity and can grow as an epiphytic life form on drifting floating macrophyte mats, which enhances its dispersion and potential impacts (Michelan et al. 2018). The establishment of U. arrecta is facilitated in ecosystems where riparian vegetation is absent, since the shade caused by native vegetation can be a limiting factor for its development (Evangelista et al. 2016).

To understand the mechanism of invasion, the correct identification of invasive aquatic plant species and information on their genetic variability is fundamental (Naugzemyz et al. 2021). This information is also important to implement management strategies, as well as research into potential control agents, including biological ones (Pitelli et al. 2008). Molecular tools have been useful for the correct identification and analysis of genetic variability in macrophyte populations, including Urochloa in the Upper Paraná River basin (Machado et al. 2016; Lucio et al. 2019; Diamante et al. 2020) and in other watersheds (Collevatti et al. 2003; Cloutier et al. 2005; Azevedo et al. 2008). A previous study with U. arrecta in Itaipu reservoir (Brazil) showed a high degree of genetic homogeneity among populations in this ecosystem (Diamante et al. 2020). However, it remains to be confirmed whether this result applies or not to samples obtained in a variety of ecosystems at a larger spatial scale.

The use of the nuclear genome, specifically the internal transcribed spacer (ITS) region of nuclear ribosomal DNA, is common for molecular phylogenetic studies of plants (Torres González and Morton, 2005). Moreover, this technique has been used on different taxonomic levels of grasses (Wang et al. 2017), in addition to studies concerning molecular discrimination and identification (Diamante et al. 2020). The use of DNA chloroplast (cpDNA) allows the identification of structural and sequence polymorphisms, which may be useful for genetic studies, as well as contributing to phylogenetic reconstruction and estimation of the divergence time between species (Salariato et al. 2010; Silva et al. 2015; Machado et al. 2016; Pessoa-Filho et al. 2017) in addition to providing insights into plants population genetics and their evolution (Moore et al. 2010).

Given the potential negative effects of introduced species on native flora and the potential economic consequences of uncontrolled growth of invasive alien species, it is important to understand the processes involved in the occupation and especially of diversification of invasive populations. In this study we evaluated the genetic diversity, at the haplotype level, of Urochloa arrecta in different ecosystems located in a large spatial scale (c. 500,000 km2) within the Brazilian territory, using the internal transcribed spacer (ITS) of nuclear DNA and the intergenic spacer located between the tRNALeu and tRNAPhe genes of chloroplast DNA (cpDNA).

Material and methods

Collecting specimens and extracting DNA

Urochloa arrecta specimens (n = 62) were collected in the states of Paraná (PR) (n = 38), Mato Grosso do Sul (MS) (n = 10), São Paulo (SP) (n = 13) and Mato Grosso (MT) (n = 1) (Fig. 1). The most distant sampling points were ca. 1,000 km apart and the coordinates of each sampling point were determined with GPS (Garmin eTrex® 30). For the collection in the state of Mato Grosso do Sul, the samplings were conducted inside an Environmental Protection Area and the authorization to collect the specimens was granted by IMASUL (No. 71/405723/2018), considering Brazilian legislation. For the other regions no authorization was required. The testimonial material of this study is deposited in the herbarium of the State University of Maringá (HUEM) and herbarium of the Federal University of Mato Grosso do Sul/ Campus—Pantanal (HCOR). The habitats where we collected samples (including the Environmental Protection Area) are altered by human interferences at different degree. Impacts include alteration of hydroperiods, deforestation, agriculture, eutrophication, and river damming (Agostinho et al. 2004; Thomaz et al. 2009; Evangelista et al. 2016).

Fig. 1
figure 1

Sampling points of Urochloa arrecta obtained in this study

Full size image

In the field, leaf tissue samples from each U. arrecta specimen were preserved in a falcon tube with TE buffer pH 8.0 (Tris HCl (1M), EDTA (0.5mM)). Total DNA from the samples was extracted using the Promega Wizard® Genomics extraction kit according to the manufacturer's instructions. To estimate the concentration of DNA present in each sample, 1.0% agarose gel electrophoresis was performed and compared with λ-phage DNA of known concentration.

Molecular markers

In this study, we used ITS and trnL-trnF loci. Both regions were partially amplified using the primers ITS 5a–F (5′-TATCATTTAGAGGAAGGAG-3′) and ITS 4-R (5′-GCATATCAATAAGCGGAGGA-3′) (Baldwin, 1992) and trn-c-F (5′-GGAAATCGGTAGACGCTACG-3′) and trn-f-R (5′-ATTTGAACTGGTGACACGAG-3′) (Reid et al. 2006), respectively. The polymerase chain reaction (PCR) was conducted in total volume of 25 μL, containing Tris-KCl 1x reaction buffer [20mM Tris-HCl (pH 8.4), 50 mM KCl], MgCl2 (1.5 mM), primers (2.5 µM each), dNTPs (0.1 mM each), and 0.5 U Platinum™ Taq DNA Polymerase. The same markers (ITS and trnL-trnF loci) were used by Diamante et al. (2020) in another study of the same species conducted in much smaller spatial scale, which allows comparisons between both investigations.

The amplification reaction program for the ITS locus consisted of 35 cycles, with a temperature of 94 °C at 30 seconds for denaturation, 30 seconds of annealing at 55 °C, and 1 min at 72 °C for polymerization. After completion of the 35 cycles, there was a 5 min time at 72 °C for extension of incomplete fragments. For the trnL-trnF locus, the amplification reaction program started with 4 min at 92 °C, followed by 40 cycles of 15 seconds at 94 °C for denaturation, 30 seconds of annealing at 59 °C and 2 minutes at 72 °C for polymerization, with a final step of 10 min at 72 °C for extension of incomplete fragments. Subsequently, the PCR products were analyzed in 1% agarose gel and purified following the protocol of Rosenthal et al. (1993). For the sequencing reaction the Big Dye Terminator kit was used. The sequencing was performed at ACTGene Análises Moleculares Ltda, Rio Grande do Sul, using the AB-3500 automated sequencer.

Genetic analysis

The obtained sequences were edited and aligned by the algorithm ClustalW (Thompson et al. 1994), using the programs BioEdit (Hall 1999) and MEGA 7 (Kumar et al. 2016), respectively. Sequence similarity values were obtained by comparison with GenBank data, obtained using Blastn.

In addition to the sequences obtained in this study, we also obtained sequences from Urochloa arrecta available in GenBank from the state of Paraná, Mexico and Zimbabwe (Supplementary Material—Table 1). Distance p values were calculated between sequences by the program MEGA 7. 0 and the nucleotide and haplotype diversity indices were obtained by DnaSP v5 (Librado and Rozas 2009). It was possible, by analyzing the alignments of both genes, to estimate the frequency of each haplotype in the different sampled populations. A haplotype network was generated using the program PopArt (Leigh and Bryant 2015). Access to the genetic heritage of the specimens was authorized through the electronic system Sisgen—National System for the Management of Genetic Heritage and Associated Traditional Knowledge (number A0D62EA) and the sequences obtained were deposited in GenBank: MW853712—MW853763; ON710859—ON710862 (ITS) and MW863262—MW863307; ON778459—ON778462 (trnL-trnF).

Results

Fifty-six ITS sequences with 517 base pairs, after alignment and editing, and fifty trnL-trnF sequences (844 bp) were obtained for Urochloa arrecta specimens from the states of Paraná, São Paulo, Mato Grosso do Sul and Mato Grosso. Although all samples were tested with both markers, the number of sequences differs between them because for some samples the amplification was not successful. For the ITS region, 14 variable sites and 7 haplotypes were identified, while for trnL-trnF, 14 variable sites and 10 haplotypes were identified. For both markers, H1 was the most frequent haplotype, shared by individuals from the states studied (Table 1). The haplotype diversity indices (h) and nucleotide (π) were, respectively, 0.1946 and 0.00105 for ITS and 0.5044 and 0.00110 for trnL-trnF.

Table 1 Distribution of haplotypes obtained from sequences of the ITS and tnrL-trnF regions of Urochloa arrecta per location
Full size table

The genetic similarity of the ITS sequences of Urochloa obtained in this study for each haplotype ranged from 98.65 to 100% with GenBank sequences from U. arrecta. The p-distance values obtained also confirm this genetic similarity, ranging from 0 to 1.8% between the collected specimens and those obtained from the database. The analysis of the trnL-trnF region also confirms the genetic proximity between the specimens, since the sequences obtained in this study for each haplotype found were 98.64–100% with sequences from the GenBank of U. arrecta, the p-distance values ranged from 0 to 0.4% between the collected specimens and those obtained from the database, confirming this similarity.

The haplotype network obtained with the ITS marker (Fig. 2A) shows that all specimens collected in the state of São Paulo (n = 13) have the same haplotype (H1), shared with samples obtained from GenBank (Itaipu Reservoir (PR) and Guaraguaçu River (PR)) and with individuals from other Brazilian localities (Paraná, Mato Grosso and Mato Grosso do Sul). The state that showed the greatest genetic variation was Paraná, with the presence of 4 of the 7 haplotypes obtained in Urochloa. The sequence comes from Zimbabwe, a country in Africa, where Urochloa is native, constituted a distinct haplotype (p distance of 0.4% to 1.8% relative to the others) (Supplementary material—table 2). Among the sequences obtained in this study, the most distinct haplotype was H6, with a p-distance of 1.4–1.9% relative to the other haplotypes.

Fig. 2
figure 2

Haplotype network of U. arrecta, based on sequences of the nuclear ITS region (2A) and the chloroplast trnL-trnF region (2B)

Full size image

On the other hand, the haplotype network obtained with the trnL-trnF marker (Fig. 2B) shows that not only state of Paraná presented the greatest haplotype variation, with the presence of 7 of the 10 haplotypes obtained in Urochloa, but specimens from the states of Mato Grosso do Sul and São Paulo also showed genetic variation. Mato Grosso do Sul presented 5 of the 10 haplotypes obtained and state of São Paulo presented 3 haplotypes. The sequences available in GenBank (from state of Paraná and Mexico) shared the most common haplotype (H1) for this marker with 100% similarity with the other specimens (Supplementary material—Table 3). Among the sequences obtained in this study, the most distinct haplotype was H5 (from state of Paraná), showing a p-distance of 0.4% to 0.7% compared to the other haplotypes.

The states of Paraná and Mato Grosso do Sul stand out for the presence of several unique haplotypes, both for the ITS region and for trnL-trnF.

Discussion

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. 2010b; 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. 2010b). 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 in 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 (Leal et al., 2022), as the ones where we collected our samples. The predominance of habitats that have been altered by human interference could help explaining the existence of populations that are similar in terms of genetic, probably well adapted to disturbances, and indeed we were not able to identify specific haplotypes in particular regions. 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 the phenotypic plasticity has been an important factor to explain the invasion success of these plants (Richardson and Pysek, 2006). Among these phenotypic characteristics, we can cite the U. arrecta aggressive potential regarding competition for nutrients, fast growth and allocation of individual resources when compared to other species (Bianco et al., 2015). Compared to plants native to aquatic environments, U. arrecta also stands out as it presents high resilience to drought (Michelan et al. 2010b; Leal et al. 2022).

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.