Cytogeography of Naturalized Solidago canadensis Populations in Europe

Autopolyploidization has driven the successful invasion of Solidago canadensis in East Asia. However, it was believed that only diploid S. canadensis invaded Europe, whereas polyploids never did. Here, molecular identification, ploidy level, and morphological traits of ten S. canadensis populations collected in Europe were compared with previously identified S. canadensis populations from other continents and S. altissima populations. Furthermore, the ploidy-driven geographical differentiation pattern of S. canadensis in different continents was investigated. All ten European populations were identified as S. canadensis with five diploid and five hexaploid populations. Significant differences in morphological traits existed among diploids and polyploids (tetraploids and hexaploids), rather than between polyploids from different introduced ranges and between S. altissima and polyploidy S. canadensis. The invasive hexaploids and diploids had few differences in latitudinal distributions in Europe, which was similar to the native range but different from a distinct climate-niche differentiation in Asia. This may be attributed to the bigger difference in climate between Asia and Europe and North America. The morphological and molecular evidences proved the invasion of polyploid S. canadensis in Europe and suggest that S. altissima may be merged into a complex of S. canadensis species. Our study may be concluded that geographical and ecological niche differentiation of an invasive plant driven by ploidy depends on the degree of difference in the environmental factors between the introduced and native range, which provides new insight into the invasive mechanism.


Introduction
Solidago canadensis L. is an invasive plant native to North America and is now widely distributed in Asia, Europe, and Australia [1,2]. It is a polyploid complex (S. canadensis L. complex) composed of multiple ploidy subspecies or variants, with three main ploidy levels in its native and introduced range: diploid (2n = 18), tetraploid (2n = 36), and hexaploid (2n = 54) [3][4][5][6], all of which are harmful weeds [6][7][8]. The hybridization, variable chromosome number, and morphological characters of genus Solidago make its classification extremely complex, and distinguishing them can be based on only a few quantitative traits [3,9]. Among them, the taxonomic status of S. altissima and S. canadensis has been controversial, with some arguing that the two should be one species, and S. altissima should be considered as a variety of S. canadensis [10][11][12], and some suggesting that S. altissima is a separate species by morphological analysis [9,13]. Most European populations of both species are described as S. altissima [1,12], but macro-morphological analysis suggests that S. canadensis is a common species in Europe [13,14]. The main differences between the two are the length and arrangement of epidermal hairs on stems and leaves, the shape and margins of leaves, and micromorphological features of the leaf epidermis and the rhizome system [15], but such distinguishing traits are sometimes inaccurate in complex hexaploid S. altissima in Belgium and suggested that all previously reported diploid populations were S. canadensis [13,52]. However, given the uncertainty of the distinction between S. altissima and S. canadensis, this population may be also a polyploid population of S. canadensis.
Hence, molecular marker and ploidy determination were performed on ten new populations in Europe to determine whether these populations are S. canadensis, as well as their ploidy level. Molecular identification results and morphological traits of S. canadensis populations and populations previously identified as S. altissima by phenotypic characteristics were compared to clarify the relationship between them. The morphological differences of different geo-cytotype populations were compared. In addition, we investigated the niche differentiation of different ploidy populations in Europe, and compared the pattern with those in North America and Asia, to further improve its geographical distribution pattern in the Northern Hemisphere and invasion mechanism.

Molecular Markers and Phylogenetic Analysis
The results of molecular markers showed that all ten populations from Europe were S. canadensis, and the ribosomal ITS sequences of EU10, EU11, and EU14 had one base mutation each compared to the sequence from GenBank, while the sequences of the other seven populations were completely consistent with S. canadensis (Table 1); the trnH-psbA intergenic spacer of EU02 had one base mutation, while the sequences of the other nine populations were completely consistent with S. canadensis in GenBank (Table 2). The ribosomal ITS and trnH-psbA intergenic spacer of the populations phenotypically identified as S. altissima were > 99% similar to S. canadensis (Table S3). Therefore, we preliminarily consider that S. canadensis and S. altissima are one species by molecular level. The ribosomal ITS and trnH-psbA intergenic spacer of 159 populations were spliced to construct a phylogenetic tree, and all populations except RUS02 were clustered together with the sequences of S. canadensis from GenBank and separated from other closely related species (Figure 1). The S. canadensis populations from Europe, other S. canadensis populations, and S. altissima populations were clustered together in the phylogenetic tree ( Figure 1). The ribosomal ITS and trnH-psbA intergenic spacer of the populations phenotypically identified as S. altissima were > 99% similar to S. canadensis (Table S3). Therefore, we preliminarily consider that S. canadensis and S. altissima are one species by molecular level. The ribosomal ITS and trnH-psbA intergenic spacer of 159 populations were spliced to construct a phylogenetic tree, and all populations except RUS02 were clustered together with the sequences of S. canadensis from GenBank and separated from other closely related species ( Figure 1). The S. canadensis populations from Europe, other S. canadensis populations, and S. altissima populations were clustered together in the phylogenetic tree ( Figure  1).

Ploidy Level of S. canadensis Populations in Europe
The peak positions of five populations in Europe were 2.6 to 2.7 times that of the internal reference diploid population, all of which should be considered as hexaploidy [6,13,53]; five overlapped with the peak of the diploid population, which were diploid (Figure 2b-k; Table S4). Based on the results of molecular markers and ploidy determination, we concluded that these populations in Europe consisted of five diploid populations and five hexaploid populations of S. canadensis.

Ploidy Level of S. canadensis Populations in Europe
The peak positions of five populations in Europe were 2.6 to 2.7 times that of the internal reference diploid population, all of which should be considered as hexaploidy [6,13,53]; five overlapped with the peak of the diploid population, which were diploid (Figure 2b-k; Table S4). Based on the results of molecular markers and ploidy determination, we concluded that these populations in Europe consisted of five diploid populations and five hexaploid populations of S. canadensis.  [54], was used as an internal reference, therefore DNA content of the diploid population (CA09) is 2.15 pg. (b-k) The relative nuclear DNA content of each population, native diploid population (CA09) used as the internal reference.

Comparison of the Morphological Traits of Different Geo-Cytotype Populations
The plant height, stem diameter, leaf length, and leaf width of the introduced populations were significantly higher than those of the same ploidy of native populations (p < 0.05), and the polyploids were significantly higher than the diploids (p < 0.05; Table 3; Figures S1-S6). The number of epidermal hairs of native populations was significantly higher than that of the introduced populations of the same ploidy (p < 0.05), but there was no significant difference between different ploidy in both the native and introduced range. The number of ray florets and disc florets did not differ significantly between different geo-cytotypes, but the length of the ray floret and height of involucre were significantly higher in the polyploids than in the diploids (p < 0.05; Table 3). A comparison of the aspect ratio of leaves and the ratio of plant height to stem diameter at the adult stage showed that the leaves were narrower in the native populations than those of the same ploidy of introduced populations, and narrower in the diploids than polyploids (p < 0.05; Figure  3a,b). The ratio of plant height to stem diameter was greater in the native populations than those of the same ploidy in the introduced range (p < 0.05), and the diploids were  [54], was used as an internal reference, therefore DNA content of the diploid population (CA09) is 2.15 pg. (b-k) The relative nuclear DNA content of each population, native diploid population (CA09) used as the internal reference.

Comparison of the Morphological Traits of Different Geo-Cytotype Populations
The plant height, stem diameter, leaf length, and leaf width of the introduced populations were significantly higher than those of the same ploidy of native populations (p < 0.05), and the polyploids were significantly higher than the diploids (p < 0.05; Table 3; Figures S1-S6). The number of epidermal hairs of native populations was significantly higher than that of the introduced populations of the same ploidy (p < 0.05), but there was no significant difference between different ploidy in both the native and introduced range. The number of ray florets and disc florets did not differ significantly between different geo-cytotypes, but the length of the ray floret and height of involucre were significantly higher in the polyploids than in the diploids (p < 0.05; Table 3). A comparison of the aspect ratio of leaves and the ratio of plant height to stem diameter at the adult stage showed that the leaves were narrower in the native populations than those of the same ploidy of introduced populations, and narrower in the diploids than polyploids (p < 0.05; Figure 3a,b). The ratio of plant height to stem diameter was greater in the native populations than those of the same ploidy in the introduced range (p < 0.05), and the diploids were significantly greater than the polyploids (p < 0.05; Figure 3a,b and Figures S1-S6). This suggests that the differences in morphological characters of different S. canadensis populations are mainly caused by both cytotype and geographical factors. the native diploid populations clustered together. The S. canadensis populations were separated with the closely related species S. gigantea, S. simplex, and S. decurrens (Figure 3c). It showed that there were no significant differences in a biological trait between the polyploids of S. canadensis and S. altissima, and the cluster analysis also indicated that the polyploids of S. canadensis and S. altissima clustered together (Table 3; Figure 3). These results further suggest that S. canadensis and S. altissima belong to the same species.  The polyploid populations from different introduced regions were not significantly different for each trait, but were significantly different from the native polyploids (p < 0.05; Table 3). Based on the cluster analysis of 11 phenotypic traits, the North American polyploid population, the Asian polyploid population, the European hexaploid population, and S. altissima were clustered together as one group, where the introduced and native polyploid populations divided into two branches. The Chinese cultivated species "huangyinghua", the European diploid populations, the Russian diploid population, and the native diploid populations clustered together. The S. canadensis populations were separated with the closely related species S. gigantea, S. simplex, and S. decurrens (Figure 3c). It showed that there were no significant differences in a biological trait between the polyploids of S. canadensis and S. altissima, and the cluster analysis also indicated that the polyploids of S. canadensis and S. altissima clustered together (Table 3; Figure 3). These results further suggest that S. canadensis and S. altissima belong to the same species.

Geographical Distribution in Europe and Global Differentiation of Ecological Niches of Different Ploidy Populations
We have previously demonstrated that S. altissima and S. canadensis are one species by molecular markers and comparison of morphological traits ( Figure 1; Table S3), and therefore integrated S. altissima in the literature into S. canadensis species for analysis. The results showed that the diploid population in Europe was mainly distributed in central Europe, while the hexaploid was relatively widely distributed, with a relatively wide range  Figure 4A,B). The latitudes of the two ploidy populations did not differ in general, but the mean July temperatures of the distribution areas differed, ranging from 15 to 21 • C for diploids and 17 to 25 • C for hexaploids ( Figure 4B,C). results showed that the diploid population in Europe was mainly distributed in central Europe, while the hexaploid was relatively widely distributed, with a relatively wide range of latitudes and temperatures (Figure 4a,b). The latitudes of the two ploidy populations did not differ in general, but the mean July temperatures of the distribution areas differed, ranging from 15 to 21 °C for diploids and 17 to 25 °C for hexaploids (Figure 4b,c).
No significant geographical differentiation was observed between diploid and polyploid populations in Europe, as in North America, while a highly significant geographical differentiation was found between the diploid and polyploid populations in Asia, with polyploids distributed at lower latitudes ( Figures S7 and 5a). However, the mean July temperature was significantly higher in the polyploid distribution than in the diploid, both in the native and different introduced ranges (Figure 5b).  No significant geographical differentiation was observed between diploid and polyploid populations in Europe, as in North America, while a highly significant geographical differentiation was found between the diploid and polyploid populations in Asia, with polyploids distributed at lower latitudes ( Figure S7 and Figure 5a). However, the mean July temperature was significantly higher in the polyploid distribution than in the diploid, both in the native and different introduced ranges (Figure 5b).

The Relationship of S. canadensis and S. altissima
Verloove et al. [13] found hexaploid S. altissima in Europe and identified it as different from S. canadensis mainly based on morphological comparison and ploidy determination. The study used diploid S. canadensis as a reference, whereas our study showed that morphological differentiation between diploids and polyploids of S. canadensis was obvious, and together with the absence of molecular marker identification, the results of this study are difficult to confirm. Semple and Uesugi [55] compared S. altissima, S. canadensis, S. chilensis, and S. gigantea for phenotypic comparison, where the phenotypic characters of S. altissima and S. canadensis were partially overlapping, but the study also did not distinguish ploidy levels. It has been shown that S. altissima and S. canadensis do not differ in functional traits, biomass production and allocation, and ploidy levels [47,56,57]. S. canadensis and S. altissima were also not significantly different in their ecological distribution in Europe [15]. The main difference between S. canadensis and S. altissima is the length and arrangement of the epidermal hairs of the stems and leaves [15]. However, the present study (Table 3; Figure 3) and several previous studies have shown that the differences in phenotypic traits among different populations were influenced by ploidy and environment [43,44]. These findings, in combination with the results of this study, suggest that these phenotypic differences between S. altissima and S. canadensis result from polyploidy and the evolution of adaptation to the environment.
Furthermore, the ribosomal ITS and trnH-psbA intergenic spacer sequence similarity between S. canadensis and S. altissima was greater than 99%, and the DNA content of hexaploid S. canadensis was the same as that of hexaploid S. altissima found in Europe ( Figure  1) [13]. The DNA content of the diploid S. canadensis found by Szymura et al. [47] was also not significantly different from that of the diploid S. altissima. Morphological comparisons showed that the differences in morphological characters between polyploid S. canadensis and S. altissima were not significant (Table 3; Figure 3). In summary, we concluded that S. canadensis and S. altissima should be considered as one complex species (S. canadensis L. complex) from the comparison of molecular identification, morphological characters, and DNA content.

The Relationship of S. canadensis and S. altissima
Verloove et al. [13] found hexaploid S. altissima in Europe and identified it as different from S. canadensis mainly based on morphological comparison and ploidy determination. The study used diploid S. canadensis as a reference, whereas our study showed that morphological differentiation between diploids and polyploids of S. canadensis was obvious, and together with the absence of molecular marker identification, the results of this study are difficult to confirm. Semple and Uesugi [55] compared S. altissima, S. canadensis, S. chilensis, and S. gigantea for phenotypic comparison, where the phenotypic characters of S. altissima and S. canadensis were partially overlapping, but the study also did not distinguish ploidy levels. It has been shown that S. altissima and S. canadensis do not differ in functional traits, biomass production and allocation, and ploidy levels [47,56,57]. S. canadensis and S. altissima were also not significantly different in their ecological distribution in Europe [15]. The main difference between S. canadensis and S. altissima is the length and arrangement of the epidermal hairs of the stems and leaves [15]. However, the present study (Table 3; Figure 3) and several previous studies have shown that the differences in phenotypic traits among different populations were influenced by ploidy and environment [43,44]. These findings, in combination with the results of this study, suggest that these phenotypic differences between S. altissima and S. canadensis result from polyploidy and the evolution of adaptation to the environment.
Furthermore, the ribosomal ITS and trnH-psbA intergenic spacer sequence similarity between S. canadensis and S. altissima was greater than 99%, and the DNA content of hexaploid S. canadensis was the same as that of hexaploid S. altissima found in Europe (Figure 1) [13]. The DNA content of the diploid S. canadensis found by Szymura et al. [47] was also not significantly different from that of the diploid S. altissima. Morphological comparisons showed that the differences in morphological characters between polyploid S. canadensis and S. altissima were not significant (Table 3; Figure 3). In summary, we concluded that S. canadensis and S. altissima should be considered as one complex species (S. canadensis L. complex) from the comparison of molecular identification, morphological characters, and DNA content.

Ploidy and Environment Determine the Morphological Diversity of S. canadensis
The growth and morphological traits of plants are important indicators for the efficient utilization of resources and tools to indicate the competitive ability of plants [58]. The general effect of polyploidy is an increase in plant size; polyploids typically have taller, more robust plants that produce higher biomass, larger rhizome systems, and larger flowers and seeds [59][60][61][62]. For example, introduced tetraploids of S. gigantea were taller than the native diploids [8]. Introduced polyploids of Fallopia have been successfully invaded due to taller plants, higher biomass, and larger rhizomes [59]. Our results showed that plant height, stem diameter, leaf length, and leaf width were significantly higher in polyploids of S. canadensis than in diploids in different regions (Table 3). Epidermal hairs can resist biotic and abiotic stresses in the environment [63]. S. canadensis has a variety of natural enemies in its native range, but lacks natural enemies in the introduced range [11,64], so the native S. canadensis may resist insects and other enemies by having more epidermal hairs. The polyploid S. canadensis had a larger flower head than the diploids due to polyploidization ( Table 3). The flowers of tetraploids and triploids of Leptospermum scoparium JR and G. Forst were significantly larger in diameter than diploids [65], and the flowers of Escallonia rubra tetraploids were also larger than diploids [66]. The greater flower heads help to attract pollinators, increase the probability of cross-pollination, ensure sufficient seed production, and improve the adaptability of S. canadensis [67]. Differences in phenotypes of S. canadensis between geo-cytotypes were observed, with the native polyploid being taller and having a more robust stem than the diploids, while the introduced polyploids had taller and stronger stems than the native polyploids (Figure 3), suggesting that S. canadensis has enhanced growth capacity through pre-and post-adaptation, resulting in greater invasive ability.
Polyploids can promote biological invasion by accelerating the differentiation of morphological traits through pre-and post-adaptive evolution [22]. Invasive plants usually have characteristics such as high phenotypic plasticity, high reproductive capacity, and suitability for dispersal [68]. The change in phenotypic expression of a genotype in response to environmental factors is known as "phenotypic plasticity" [69,70]. In Europe, S. canadensis has a relatively wide niche with regard to soil properties, which can be related to both local adaptation and phenotypic plasticity [56,71]. Studies have proved that plasticity in S. canadensis plays a central role in shade adaptation [72] and salt tolerance [73], but these studies did not compare phenotypes of different ploidy populations. Our study and Cheng et al. [6] proved that the polyploids had a higher phenotypic plasticity than diploids. Studies on S. gigantea have shown that only tetraploids have successfully invaded Europe, that native tetraploids produce more branches and rhizomes than diploids, and that population size, density, and total plant biomass were greater in introduced populations than in native populations [8,74]. A study of Mimulus guttatus DC. showed that there were no significant differences in phenology, floral traits, sexual, and nutritional reproduction between the North American (native range), Scottish, and New Zealand populations (introduced range), but the introduced populations had twice as many flower-bearing upright side branches as those of native populations, due to the fact that different introduced populations have evolved by adaptation and enhanced reproduction [75]. The morphological characteristics of polyploid populations of S. canadensis in different introduced ranges did not differ significantly from each other, but they differed from the native populations, especially diploid populations ( Table 3), indicating that polyploids underwent rapid adaptive evolution after the invasion in different regions, adapted to the local ecological environment, and finally invaded successfully.

Polyploid S. canadensis Has Been Successfully Invaded and Widely Distributed in Europe
Studies have investigated the cytogeography patterns of S. canadensis in its native range in North America, and Asia where it is invasive, and diploids, tetraploids and hexaploids were found [6,76]. However, few studies have investigated the distribution of ploidy levels in S. canadensis in Europe, and previous studies have suggested that only diploid is present in Europe [5,6,16,47,50]. In general, diploids are less abundant in the invasive range than polyploids and are present in areas where climatic conditions are similar to those of native populations [77]. For example, there are diploids, tetraploids, and hexaploids of S. gigantea in the native range, but only tetraploids have invaded Europe [41]. Studies on C. maculosa have also shown that the proportion of polyploids is much higher in North America, where it is more invasive than in its native range in Europe [40]. Previous studies have shown that only polyploids of S. canadensis have successfully invaded in Asia [6], and five hexaploid populations and five diploid populations were found in Europe in our study (Figure 1), suggesting that polyploids may also have been widely distributed in Europe.
The polyploid populations of S. canadensis in Asia have changed their isothermal niche by adapting to summer heat stress and are distributed at a lower latitude compared to diploids [6]. Distributions of S. canadensis were limited climatically in Central Europe, and were also correlated with proxies of human pressure [78]. The latitudinal difference between diploid and hexaploid populations found in Europe was not significant, but hexaploids have a wider latitudinal range of distribution (Figure 3). Heat tolerance is usually a key factor for the successful invasion of invasive plants, and greater heat tolerance of polyploids could help them expand their invasion range [33]. The mean July temperature in the distribution areas of hexaploids was significantly higher than that of diploids (Figure 3d), suggesting that hexaploids in Europe have also evolved adaptively and are better adapted to temperature changes. Currently, S. canadensis is widely distributed in central Europe [47,52,57], and some studies also predict that S. canadensis will invade further south in Europe [46], which may be related to the greater adaptability and heat tolerance of polyploids [6]. Certainly, the number of populations we studied is not large enough, so future attention should be paid more to the ploidy distribution of S. canadensis in Europe and its invasion in the south.

Ploidy-Driven Ecological Niche Differentiation of S. canadensis in the Northern Hemisphere Depends on Environmental Factors
Solidago canadensis is widely distributed in North America, including the temperate continental climate zone in the northeast and the subtropical humid climate zone in the southeast [6]. The geographical distribution and ecological niches of diploids and polyploids have diverged significantly after the invasion, which is mainly due to the deviation of the climatic ecological niches of polyploids. Diploids are still distributed in regions with similar climates to their native regions, such as Europe and the Russian Far East. Polyploids have invaded China and Europe, mainly distributed in the subtropical monsoon climate zone in China, and in the Europe range with a climate similar to that of North America ( Figure 5). This is mainly due to the weak adaptation of diploids, while polyploids are more environmentally adapted and can successfully invade a wide range of climatic regions [79].
Whether invasive plants show ecological niche differentiation after the invasion is influenced by several factors [80,81]. Polyploidy can facilitate the adaptive evolution of plants, which is important for the spread of plant invasions [23,82]. There is no generalized pattern of niche differentiation during the establishment and evolution of polyploids, and different lineages of the same polyploid plant may coexist with different patterns of climatic niche evolution (niche conservatism, contraction or expansion) [37]. Whether polyploids develop niche differentiation depends on the history of the species: e.g., age of the polyploid, multiple origins, or the number of ploidy levels, among others [83]. In addition, the climate is the main factor driving the distribution of plant species [84], which has an important impact on the ecological adaptability of different cytotypes [85]. S. canadensis invaded East Asia and Europe at about the same time, but the geographical differentiation patterns of different ploidy populations after the invasion were different. The climate of the distribution area of S. canadensis polyploids in China was different from that in North America and Europe, and the mean July temperature was significantly higher than that of the polyploid population in North America and Europe ( Figure 5). Polyploids in East Asia have evolved through adaptation to enhance heat tolerance to expand into subtropical regions, while diploids are unable to adapt to high temperatures and can only be distributed in mid-latitudes, thereby promoting geographical differentiation of ploidy [6]. As the climate was similar to that of the native range, there was also no clear ploidy-driven geographical differentiation in Europe. In summary, we suggest that it is mainly climatic factors that lead to the adaptive evolution of polyploids of S. canadensis in different introduced ranges, hence to different patterns of ecological niche differentiation. Our study indicated that the ploidy-driven adaptive evolution of invasive plants depends on the degree of difference between the environmental factors of the introduced range and the native range.
Even though there was no difference in latitude between the distribution of different ploidy populations in North America and Europe, the mean July temperature of polyploids was still higher than that of diploids ( Figure 5), indicating that there is a difference in the required optimum ecological niches of different ploidy populations, and temperature is an important factor influencing the geographical distribution of S. canadensis. Under the environmental conditions of global warming, the future distribution of S. canadensis is likely to further expand from the temperate zone to pantropical regions driven by polyploidization.

Experimental Materials
The ten populations collected from Europe (Table S1) were performed for ploidy determination and molecular identification. Molecular identification results were used to compare and construct phylogenetic trees together with 142 S. canadensis populations from other regions and seven populations identified as S. altissima by phenotypic characteristics as previously described by Cheng et al. [6].
Biological characteristics were measured for ten populations from Europe and 93 other different ploidy populations of S. canadensis collected from the native regions (US and Canada) and introduced regions (China, Russia), and 8 populations identified as S. altissima, and one population each of S. gigantea, S. simplex, and S. decurrens (Table S1). The detailed descriptions of material collection and identification had been presented by Cheng et al. [6].

DNA Extraction and Molecular Identification
To confirm the correct identification of the ten populations, we sequenced the ribosomal ITS (608 bp) and trnH-psbA intergenic spacer (213 bp) of three individuals from each population. To this end, the total DNA from the dried leaves was extracted using a Plant Genomic DNA kit (DP305, Tiangen Biotech Co., Ltd., Beijing, China). A pair of primers (upstream: 5 -CGTAACAAGGTTTCCGTAG-3 , downstream: 5 -TTATTGATATGCTT AAACTCAGCGGG-3 ) were used to amplify ITS rDNA gene and another pair of primers (upstream: 5 -CCGCCCCTCTACTATTATCTA-3 , downstream: 5 -TCTAGACTTAGCAGCTATTG-3 ) were used to amplify psbA-trnH intergenic spacer. Polymerase chain reaction (PCR) was carried out in 50 µL containing 2 µL of 10 µmol/L primers, 10 × buffer 5 µL, 25 mM Mg 2+ 4 µL, 2.5 mmol/L dNTPs 4 µL, 50 ng/µL DNA templates 1 µL, and 5U Taq DNA polymerase 0.5 µL. The PCR cycle was 95 • C for 4 min followed by 30 cycles of 95 • C for 30 s, 52 • C for 60 s, 72 • C for 60 s, and a final extension step at 72 • C for 7 min. PCR products were separated on 1% agarose gels and stained with ethidium bromide. The PCR products were purified and sequenced.

Construct Phylogenetic Tree
The ribosomal ITS and chloroplast psbA-trnH were sequenced and edited using the BioEdit v 7.0.5.3 software. As the reference for alignment, we extracted the data for S. canadensis (ITS, HQ142590.1; psbA-trnH, KX214929.1) from GenBank and aligned them with the ITS and psbA-trnH sequences of our samples using the Clustal X 1.83 software.
The ITS1, 5.8S, and ITS2 gene boundaries and psbA-trnH were determined from ITS and psbA-trnH sequences of the Solidago genus in GenBank. To further verify the species identity of our samples, we conducted a BLAST search in the database of the National Center for Biotechnology Information (NCBI) to determine the species with the most similar identities. The ribosomal ITS and chloroplast psbA-trnH of S. decurrens, S. gigantea, S. flexicaulis, S. juncea, S. petiolaris, S. sempervirens, S. shortii, S. fistulosa, and S. simplex were downloaded from the GenBank. The ribosomal ITS and chloroplast psbA-trnH were spliced by SequenceMatrix 1.78. In addition, we built phylogenetic trees for all 152 S. canadensis populations, 7 S. altissima populations, and closely related species with the ITS-psbA-trnH sequences, using the MEGA 4.1 Beta 3 software based on the maximum likelihood (ML) analyses, using a GTR substitution model as determined. We assessed branch support with 1000 bootstrap replicates.

Determination of Ploidy Level
We grew up plants from each seed family in the greenhouse. Fresh leaf material was held in self-sealing plastic bags containing silica and transported to the laboratory for ploidy level determination. We applied a modified flow cytometry method by Galbraith et al. [86] for the determination of ploidy level. Briefly, 0.5 g fresh leaf tissue was chopped up with a razor blade in 2 mL of ice-cold Galbraith buffer (PH 7.0, 45 mM MgCl 2 , 30 mM sodium citrate, 20 mM 4-morpholinepropane sulfonate, 0.5% w/v polyvinylpyrrolidone, and Triton X-100 1 mg/mL) in a vitreous Petri dish held on a chilled brick. The resultant fine slurry was filtered through a 50 µm microfilter into 2 mL polystyrene tubes and incubated at 4 • C for 5 min. The final filtrate was treated with 10 µL/mL of RNase A for 10 min at 4 • C. Then, 200 µL/mL of propidium iodide (PI) staining solution was added and incubated at 4 • C for 30 min in darkness. The fluorescence strength of PI-DNA conjugates was measured by flow cytometry (BD FACS, Becton Dickinson, Franklin Lakes, NJ, USA) using a 488 nm laser and 590/40 emission filter [87]. The fluorescence can be converted to chromosome number using standard population with known ploidy [6,76]. Therefore the native diploid population (CA09, 2n = 18) which chromosome number has been assessed using mitotic root-tip squashes [6], was used as an internal reference to measure the ploidy of ten populations. The ratio of DNA content of different ploidy was determined previously by the determination of ploidy level of a large number of populations [6]. Glycine max Merr. 'Polanka is used as an internal reference to measure the DNA content of diploids [54]. Flow-cytometry data acquisition and analysis was done with the software FloMax version 2.0 (Quantum Analysis, Münster, Germany). Only histograms with CVs smaller than 5% and with a minimum peak height of 50 particles were accepted. Other histograms were discarded and corresponding samples were designated no signal [88].

Measurement of Morphological Traits
Populations of S. canadensis, S. altissima, S. gigantea, S. simplex, and S. decurrens (Table S1) were planted in a common garden in Pailou Teaching and Research Station of Nanjing Agricultural University (32 • 02 N, 118 • 37 E), Nanjing, China. This region is characterized as being a subtropical monsoon climate, with an average annual rainfall of 1090 mm and the mean annual temperature is 15 • C (lowest in Feb at 2.7 • C and highest in July at 28.1 • C). They were watered, weeded, and sprayed with pesticides regularly. After plant establishment, three individuals of each population were selected, and three replicates of each individual were used to measure the 7 vegetative traits: height, stem, leaf length, leaf width, number of epidermal hairs of the main vein (per 2 mm length), number of epidermal hairs of the lateral vein (per 2 mm length), and number of epidermal hairs of the netted venation (per 4 mm 2 ) ( Table 3). One leaf from each plant was randomly removed from the upper, middle, and lower parts of the plant to measure the leaf length and width and epidermal hair-related traits. In the observation of leaf epidermal hairs, the apical part of the leaf, the middle part of the leaf, and the base of the leaf were measured as replicates.
During the flowering period, 56 populations of S. canadensis were selected, and populations of other species remain unchanged. Three individuals of each population were selected, and three replicates of each individual were used to measure the 4 floral traits: length of the ray floret, height of the involucre, number of ray florets, and number of disc florets (Table 3).
Morphological traits of populations of different sources, ploidy, and species were compared, and each morphological index was standardized by the maximum method for cluster analysis. Although a diploid population was previously found in Russia [6], it was investigated as an individual population due to its proximity to Asia and distance from other European populations.

Geographical Differentiation Analysis of Different Ploidy Populations
The geographical differentiation of different ploidy populations in Europe and other regions was analyzed by combining with our previous research [6] and literature records about S. canadensis and S. altissima (Table S2; Figure S7). Climatic factors of all occurrences were extracted from worldclim (https://www.worldclim.org/ (accessed on 1 December 2021)) for the analysis of differences in climatic factors between the distribution ranges of different ploidy populations. In addition, the ecological niche differentiation pattern of different ploidy populations in Europe, North America, and Asia (the Russian population merged into Asia) are compared by comparing latitude and mean July temperatures of their distribution sites.

Data Analysis
The average and square deviation of morphological traits and climate factors for different geo-cytotypes of S. canadensis in different regions were calculated, and the significance of the differences (p < 0.05) between geo-cytotypes in different regions was analyzed using one-way ANOVA with Least Significant Difference (LSD). All data are expressed as the mean ± standard error (Mean ± SE). Statistical analyses and plotting were implemented in R v4.1.2 [89].

Conclusions
Our study provides morphological and molecular evidence that the hexaploidy of S. canadensis has successfully invaded in Europe and S. altissima and S. canadensis should be considered as one complex species. There were significant differences in morphological traits among different geo-cytotypes. Polyploidy has evolved adaptively in different introduced ranges, enhancing the competitive ability of S. canadensis to achieve its successful invasion. Ploidy-driven ecological niche differentiation of S. canadensis in the northern hemisphere depends on environmental factors, with no significant differentiation in North America and Europe, and significant differentiation in Asia, suggesting that the ploidydriven adaptive evolution in the introduced range depends on the degree of difference from the environmental factors of the native range. Compared with the diploid, the polyploid of S. canadensis has a higher mean July temperature in its distribution area and is more adaptable to temperature changes. The distribution area of S. canadensis will gradually expand globally in the future, and the changes in the distribution area of the polyploids should be paid attention to in time.
Author Contributions: Z.T.: methodology, data curation, formal analysis, investigation and writing original draft; J.C., J.X., J.Z. and X.Y.: methodology, investigation, data curation; D.F.: data curation, review and editing; Z.Z., Y.Z. and Z.M.: resources; S.Q.: conceptualization, funding acquisition, project administration, supervision, writing, review, and editing. All authors have read and agreed to the published version of the manuscript.