1. Introduction

Biological invasions threaten ecosystems, negatively impacting global ecology, the environment, and the economy (Hao and Ma, 2023; Li et al., 2023). Generally, areas at high elevations face a lower risk of invasion by non-native species (Alexander et al., 2016; Iseli et al., 2023). Invasive species typically undergo directional ecological filtering during their upward expansion along elevations (Alexander et al., 2011). In the harsh environmental conditions of higher elevations, this filtering effect becomes more pronounced, leading to a gradual decrease in the species richness of invasive plants (Steyn et al., 2017; Larson et al., 2021). Consequently, non-native plants are generally less common at higher elevations under natural conditions (Pauchard et al., 2009; McDougall et al., 2011). However, this situation is changing, as the global climate changes and human activities (e.g., tourism development, road construction, and agricultural development) increase at high elevations (Petitpierre et al., 2016; Dainese et al., 2017; Larson et al., 2021).

Studies have shown that human disturbance and human-mediated dispersal significantly influence the distribution of alien species in mountain regions (Pauchard et al., 2009; Alexander et al., 2016; Rocabert et al., 2024). Due to this increased human interference, mountain ecosystems face growing invasion pressure (Luo et al., 2020; Larson et al., 2021). This trend is reflected not only in the increase in the number of invasive species in mountain ecosystems, but also in the continuous expansion of their distribution range (Alexander et al., 2009; Guo et al., 2018; Iseli et al., 2023). For example, in the Andes, hiking infrastructure has been shown to facilitate the upward expansion of non-native species by serving as a conduit for both propagule introduction and increased human disturbance in otherwise undisturbed areas (Alvarez et al., 2022). In this process, both hikers and livestock can serve as important vectors for the propagation of non-native plants (Liedtke et al., 2020; Koyama et al., 2024). Studies have shown that in mountainous areas, both large-scale disturbances (such as roads) and small-scale disturbances (even centimeter-scale gaps) can promote the establishment and expansion of invasive species (Milbau et al., 2013; Lembrechts et al., 2014, 2016). Moreover, studies have indicated that invasive species richness has increased over time because of human activities and climate warming (Khuroo et al., 2011; Pollnac et al., 2012; Seipel et al., 2012; Sandoya et al., 2017). In contrast, for a certain invasive alien species with widespread invasion, the distribution pattern in global mountain ecosystems remains unclear, as does the impact of human activities on its invasion in these ecosystems.

Species face highly complex and diverse environmental conditions when they expand into mountain ecosystems (Iseli et al., 2023). Consequently, species that achieve widespread invasion in mountain ecosystems often benefit from enhanced phenotypic plasticity or genetic diversity, which enables them to adapt to varying conditions at different elevations (Ohsawa and Ide, 2008; Jia et al., 2019). Previous studies have shown that many invasive plants exhibit marked phenotypic plasticity along elevation gradients in mountain regions (Masaaki et al., 2017; Steyn et al., 2017). For example, they frequently adjust traits such as plant height and leaf area to adapt to local variations in light and temperature at different elevations (Bustamante et al., 2017; Midolo and Wellstein, 2020). Additionally, the dispersal-related traits of plants (e.g., seed size, shape, number) are closely related to elevation (Rathee et al., 2021; Rosbakh et al., 2022). Seeds in high-elevation areas are typically smaller and more elongated, which helps them disperse more effectively (Wang et al., 2021). However, some invasive plants have developed larger seed mass at higher elevations, potentially to enhance seedling establishment and survival by providing more stored resources (Pluess et al., 2005). In theory, reductions in population size that accompany expansions into high-elevation regions may lead to decreased genetic diversity, and certain traits, particularly seed dispersal capacity, may also weaken with increasing elevation (Li and Feng, 2009; Reisch and Rosbakh, 2021). Such changes should be highly unfavorable for species invasion in mountain ecosystems, indicating that mountain ecosystems may act as a barrier to the invasion and expansion of alien species (Tecco et al., 2015; Liu et al., 2021b). However, the increasing number of invasion events in these areas seems to conflict with this expectation, especially because many invasive species already have low genetic diversity because of founder effects (Shen et al., 2014; Ju et al., 2022). This suggests that the barrier effect of mountains may be broken due to human activities, enabling invasive species without genetic diversity advantages to expand widely in mountain ecosystems. However, this issue has not yet been reported in relevant studies.

Galinsoga quadriradiata, an annual herbaceous plant native to Central and South America, has expanded to various regions worldwide (Yang et al., 2018b; Zhang et al., 2021). The plant is considered a malignant invasive weed in agriculture, producing a large quantity of lightweight seeds with high dispersal potential (Liu et al., 2021b). Research has shown that with increasing elevation, both its dispersal ability and genetic diversity tend to decline (Liu et al., 2021b). Although G. quadriradiata benefits from symbiosis with arbuscular mycorrhizal fungi in harsh environments, this association decreases with elevation in mountain ecosystems, possibly explaining why its dispersal ability weakens at higher elevations (Liu et al., 2021a). However, numerous studies suggest that with intensified human disturbance and global climate change, the trend of exotic species expanding to higher elevations may gradually strengthen (Alexander et al., 2016; Yang et al., 2018b). Despite its tropical origins, G. quadriradiata is widely distributed across various climate zones and elevations in China. However, little is known about how it has adapted to this non-native habitat.

In this study, we asked which factors drive the global and regional distribution of Galinsoga quadriradiata at different elevations. We examined (1) whether human activities (i.e., roads) promote G. quadriradiata invasion, (2) how seed dispersal-related traits of G. quadriradiata change along elevation gradients, and (3) whether G. quadriradiata has adapted to high-elevation environments through phenotypic plasticity or genetic variation.

2. Materials and methods 2.1. Influence of human activities

On the global scale, we assessed the elevational distribution of Galinsoga quadriradiata and its relationship with human activities. Initially, we selected the Alps, Rockies, Andes, and Himalayas as the main study sites. However, due to limited occurrence data availability for the Himalayas, we expanded the study area to cover all of China. Given China's numerous mountain ranges and plateaus, as well as the well-defined invasion gradient of G. quadriradiata, this broader scope allowed for a more comprehensive evaluation of the impact of human activities on the distribution of invasive species (Liu et al., 2016, 2021b).

To assess the impact of roads and human activities on the invasion, we combined data from published literature (Zhang et al., 2021), the Global Biodiversity Information Facility (https://www.gbif.org/en/; GBIF, 2016), and field survey records from this study to summarize global occurrences of Galinsoga quadriradiata. Subsequently, we converted the occurrence data into raster data (6 arcminutes) and counted the number of occurrences within each raster. We obtained the relevant global datasets from the following sources: elevation data from the WorldClim 1.4 database (https://www.worldclim.org/; Hijmans et al., 2005); road density data from the GRIP dataset of the Global biodiversity model for policy support (GLOBIO, https://www.globio.info/global-patterns-of-current-and-future-road-infrastructure; Meijer et al., 2018), which covers 222 countries and includes over 21 million kilometers of roads, providing an accurate representation of road density worldwide; and the human influence index (HII) from the Last of the Wild, Version 2 (http://sedac.ciesin.columbia.edu/, accessed on 27 July 2021). This dataset is created from nine global data layers covering human population pressure (population density), human land use and infrastructure (built-up areas, nighttime lights, land use/land cover), and human access (coastlines, roads, railroads, navigable rivers), providing a comprehensive reflection of the intensity and spatial distribution of human activities globally. The resolution of each dataset is 2.5 arcminutes. We resampled these datasets to match the resolution of the occurrence data.

2.2. Field survey and sample collection

Focusing on localized studies within China, we conducted field surveys and greenhouse experiments to investigate the invasion mechanisms of Galinsoga quadriradiata across elevation gradients. This investigation specifically focused on the impact of both phenotypic plasticity and genetic adaptation.

Field surveys and sample collection (both seed and leaf samples) were conducted in July and August of 2017–2019. We surveyed 77 populations of Galinsoga quadriradiata across a range of elevations (from 42 to 3443 m a.s.l.) in China (Fig. 1 and Table S1). We randomly chose 4–20 mature individuals from each population to collect seed samples from 33 populations and leaf samples from 53 populations. From the sampled plants, 10 mature and intact leaves were collected per plant and placed in plastic bags with silica gel for drying, then stored at −80 ℃ in a freezer until measurement. All seed samples were placed in envelopes and stored at 4 ℃ in a refrigerator until use.

2.3. Measurement of seed dispersal-related traits

To calculate hundred seed weight (HSW), we weighed 100 randomly selected ripe seeds with at least 10 replicates from the collected seeds (only disk achenes) of each population. We used WinSEEDLE Pro (WinSEEDLE™, Régent Instruments Inc., Québec, QC, Canada) to measure seed (only disk achene) total length (TL), achene length (AL), pappus length (PL), and pappus width (PW) (Fig. 1). For each elevational population, 30 ripe seeds with 8 replicates were randomly selected for measurement.

The mass-area ratio (MAR) has been demonstrated to reliably indicate the dispersal ability of wind-dispersed species within the Asteraceae family (Matlack, 1987). A greater seed dispersal ability is indicated by a lower MAR value. We calculated MAR using a method previously described (Liu et al., 2021b) (1):

MAR=mπR2

(1)

where m represents the seed mass, and R represents the pappus radius of the seed.

From the aforementioned 31 populations (excluding two populations due to a lack of seed samples), 30 seeds (only disk achenes) with similar size and complete morphology were randomly selected from each population for measurement of seed settlement velocity. The release height (h) was set at 1.5 m. The seed was released from the top of a glass tube with a 5-cm inner diameter, and the time (t) it takes for the seed to reach the ground was recorded with three repetitions per seed. The seed settlement velocity was calculated using the following formula (2):

v=h/t

(2)

Generally, the lower the settlement velocity of seeds, the stronger their ability for long-distance dispersal.

Horizontal dispersal distance of seeds was measured using a wind tunnel. The seeds were released at a height of 150 cm, and the horizontal dispersal distance was measured with a wind speed of 2.4 m/s (Li et al., 2022). Thirty seeds were measured per population, with each seed being measured three times. The wind speed was set following the prevailing wind speed in the study area.

2.4. Genetic variation

Leaves from 53 populations were selected for genetic analysis from all collected samples (Table S1). Genetic diversity of Galinsoga quadriradiata was analyzed using inter-simple sequence repeat (ISSR) markers. Total DNA was extracted by the CTAB method (Gawel and Jarret, 1991). ISSR primers were synthesized (Shanghai Sangon Biological Engineering Technology & Service Co., Ltd) according to a publicly available primer sets from the University of British Columbia.

DNA from 336 samples of 53 populations was amplified using seven primers (UBC 816, UBC 817, UBC 818, UBC 826, UBC 847, UBC 855, and UBC 857) in ISSR analysis. Polymerase chain reaction (PCR) for ISSR analysis followed previous studies (Liu et al., 2021b). PCR products were separated on a 1.2% agarose gel. Images were captured and analyzed using the Tanon 3500 gel imaging system.

2.5. Greenhouse experiment

Seeds from 27 populations were selected for greenhouse experiments (Table S1). Ten seeds were randomly chosen from each population for cultivation. Germination rate of these seeds was recorded within two weeks for each population. Germinated seedlings were transplanted into plastic pots measuring 8 cm in diameter and 8 cm in height. For each pot, a single plant was cultivated with nutrient soil. Seedlings were watered daily to avoid drought stress. We recorded the temperature, humidity and flowering time (FT). During peak growth, we measured chlorophyll content (CHL) in three healthy, mature leaves randomly selected each plant by using a portable chlorophyll meter (SPAD-502, Spectrum Technologies, Inc., Plainfield, IL, U.S.). Simultaneously, we measured leaf area using the WinFOLIA system from 3 to 5 leaves from each plant. Leaves were then dried at 65 ℃ for 48 h until a constant weight was achieved to calculate the specific leaf area (SLA, the ratio between leaf area and leaf mass). Plant height was measured at one month and two months after transplantation to calculate the relative growth rate (RGR, the ratio of the difference in plant height to the difference in time). Plants were harvested after three months, and plant height (H), number of capitula (NC), and branch number (BN) were recorded. Roots, stems, leaves, and flowers were harvested separately. Root length (RL) was recorded and wet weight measured for roots, stems, leaves and flowers. Root mass (RM), stem mass (SM), leaf mass (LM), and seed mass (SDM) were measured after drying in an oven at 65 ℃ to constant weight.

We calculated other growth traits such as aboveground mass (AM, including stem and leaf mass), total mass (TM, including stem, root, seed and leaf mass), leaf mass ratio (LMR, the ratio between leaf mass and total mass), root mass ratio (RMR, the ratio between root mass and total mass), stem mass ratio (SMR, the ratio between stem mass and total mass), inflorescence mass ratio (IMR, the ratio between seed mass and total mass), and water content (WC, the ratio of the difference between aboveground mass_wet and aboveground mass_dry to aboveground mass_wet).

2.6. Data analyses

To examine the influence of human activities, we analyzed trends in HII, road density, and the number of Galinsoga quadriradiata occurrences with elevation in four study regions using both linear and nonlinear regression methods. Furthermore, we employed a structural equation model (SEM) to assess the contributions of HII, road density, and elevation to the number of G. quadriradiata occurrences on a global scale.

For data from the field survey and greenhouse experiment, one-way ANOVA was used to compare the seed dispersal-related traits of Galinsoga quadriradiata among populations at different elevations. Linear regression analysis was employed to examine the variation patterns of seed dispersal-related traits and growth traits of G. quadriradiata across elevations. We also performed Redundancy Analysis (RDA) to analyze the correlations between seed dispersal-related traits and climate factors, while Pearson correlation analysis was used to examine the relationships between seed traits and soil factors. The climate data was obtained from the WorldClim 1.4 dataset (https://www.worldclim.org/) with a resolution of 2.5 arcminutes, and the soil factors included four variables from the Harmonized World Soil Database v.2.0 (https://gaez.fao.org/pages/hwsd).

For the statistical analysis of genetic variation, only distinct, reproducible and well-resolved PCR fragments were included. ISSR bands were scored as either present (1) or absent (0), constructing a binary matrix. We utilized the Frequency-Based function in GenAlEx v.6.502 (Peakall and Smouse, 2012) to calculate genetic diversity indices at the population level, including the percentage of polymorphic loci (PPL), the Shannon's information index (I), and the expected heterozygosity (H_e), with higher values indicating greater genetic diversity. Analysis of Molecular Variance (AMOVA) was employed to compare the amount of total genetic variation partitioned within and among populations. Nonlinear regression analysis was used to assess the variation pattern of population genetic diversity with elevation.

Finally, we generated the distribution map using ArcGIS 10.2. Additionally, all data analysis and figure creation were performed in R v.4.1.3 (R Core Team, 2022; https://www.r-project.org/).

3. Results 3.1. Influence of human activities

Road density, human activities, and the number of Galinsoga quadriradiata occurrences shared similar patterns across different elevations around the world (Fig. 2a). In the Alps and China, the number of G. quadriradiata occurrences was most concentrated in low-elevation areas, as was the frequency of human activity. In the Rocky Mountains and the Andes, the number of G. quadriradiata occurrences peaked in mid-elevation areas, as did the frequency of human activity. Notably, human activity and road density data were extracted based on actual distribution points of G. quadriradiata (not the overall conditions of the mountain ranges), indicating that the distribution of this species along an elevational gradient is closely related to local human activity intensity and road density. Global analysis indicated that road density (β = 0.11, P < 0.001) and human activity (β = 0.08, P < 0.001) contribute positively to the number of G. quadriradiata occurrences (Fig. 2b).

3.2. Seed dispersal-related traits

Seed dispersal traits, including TL, AL, PL, PW, HSW, and MAR, differ between populations of Galinsoga quadriradiata (Table 1), indicating that dispersal ability may differ between these populations. In addition, G. quadriradiata seed traits were correlated with elevation (F = 52.2511, P < 0.001). For example, HSW and MAR increased at higher elevations, whereas TL, AL, PL, and PW decreased (Fig. 3), resulting in reduced dispersal ability.

Generally, lower settlement velocities had greater long-distance seed dispersal. In this study, seed settlement velocity was positively correlated with elevation, whereas horizontal diffusion distance was negatively correlated with elevation (Fig. 4). This indicates that as elevation increases, the long-distance dispersal ability of the seeds of G. quadriradiata decreases.

Seed traits—TL, SL, PL, PW, and HSW were positively correlated with soil water capacity (AWC) and topsoil organic carbon (TOC), whereas MAR was negatively correlated with these traits (Table 2). These findings suggest that seed dispersal capacity increases in areas with high soil water capacity and topsoil organic carbon.

Most variance in seed traits was explained by RDA1 (97.93%) (Fig. 5). Seed traits were also correlated with environment factors. Specifically, seed traits were correlated with bio1 (F = 48.9056, P = 0.001), bio5 (F = 5.7662, P = 0.028), bio7 (F = 10.5175, P = 0.007), bio10 (F = 4.8670, P = 0.039), bio12 (F = 12.9534, P = 0.006), bio14 (F = 9.3636, P = 0.006), and bio16 (F = 4.4394, P = 0.047). MAR was negatively correlated with these climate factors, suggesting that seed dispersal capacity increases at higher temperatures and precipitation levels.

3.3. Genetic diversity

A total of 116 polymorphic loci were obtained from the ISSR analysis. Genetic variation among populations and within populations accounted for 60.88% and 39.12% of the total genetic variation, respectively (Table 3).

PPL of the 53 populations ranged from 2.59% to 43.97%, with an average of 22.69%. I ranged from 0.014 to 0.189, with an average of 0.116. H_e ranged from 0.010 to 0.122, with an average of 0.077. There is a nonlinear relationship between genetic diversity of Galinsoga quadriradiata and elevation, characterized by an initial decrease followed by an increase of genetic diversity indices along the elevation (Fig. 6).

3.4. Greenhouse experiment

Under common conditions, significant differences were detected among populations of Galinsoga quadriradiata in most phenotypic traits; however, no significant differences were observed in SLA and RL (Table S2). The similarities and differences in phenotypic traits exhibited by G. quadriradiata from different populations suggest the presence of a complex adaptive mechanism involving both phenotypic plasticity and genetic variation.

SM, SDM, RM, NC, RGR, RMR, and IMR are negatively correlated with the source population's elevation, while LMR is positively correlated with the source population's elevation (Figs. 7 and S1). This suggests that as elevation increases, the energy allocated to roots and reproduction in G. quadriradiata decreases, while the energy allocated to leaves increases.

4. Discussion

Mountains, characterized by extreme heterogeneity in climate and human disturbance, provide a unique "natural laboratory" for both observational and experimental studies to reveal the invasion process (Pauchard et al., 2009; Fuentes-Lillo et al., 2021). Our study mapped the global and regional elevational distribution of the invasive plant Galinsoga quadriradiata and determined the factors that drive its distribution. At the global scale, we found that human activities (especially road density) greatly increased the invasion risk of this species in mountain areas, highlighting the key role of anthropogenic pathways in breaking through natural ecological barriers. At the regional scale (China), we observed significant trait variation among populations of G. quadriradiata across different elevations in China, likely due to the combined effects of phenotypic plasticity and genetic differentiation. The low genetic diversity and reduced seed dispersal ability of G. quadriradiata are unfavorable for its future expansion across high mountainous terrain. However, the plant adapts by adjusting reproductive allocation to ensure higher survival rates in high-elevation areas. Our combination of global and regional research scales allowed us a more comprehensive understanding of the invasion patterns and drivers of G. quadriradiata expansion across different elevations.

4.1. Influence of human activities

In this study, we determined how road density and human activities affected Galinsoga quadriradiata elevational distribution on a global scale. We found that the number of the invasive G. quadriradiata plants decreased with increasing elevation, which confirms previous studies (Seipel et al., 2012, 2016; Yang et al., 2018a). Under normal circumstances, the harsh climate and limited human activity at high elevations pose barriers to the expansion of non-native species (Watermann et al., 2020; Iseli et al., 2023). However, in some mid-elevation areas of certain mountain ranges, an unexpected pattern emerged: frequent human activity led to a peak in the number of G. quadriradiata (Fig. 2). This phenomenon indicates that human activity and road density have, to some extent, reversed the typical trend of decreasing species abundance with elevation. This clearly suggests that while elevation plays a key role in determining plant distribution, human activity within the region should not be underestimated, as it can significantly influence species distribution (Dainese et al., 2017). Specifically, soil disturbance and vegetation destruction caused by road construction and tourism development have facilitated the establishment and expansion of alien plant species (Flory and Clay, 2009; Son et al., 2024). Moreover, human activities, both intentionally and unintentionally, have transported seeds or reproductive bodies, further accelerating their expansion (Mack and Lonsdale, 2001; Pickering and Mount, 2010; Huiskes et al., 2014).

The findings of this study are consistent with global-scale conclusions that the expansion of invasive plants benefit from human activities (Pickering and Mount, 2010; Iseli et al., 2023). We further reveal the specificity of this relationship along elevation gradients: although climatic conditions in high-elevation regions create natural barriers, localized human activity can serve as "stepping stones" for species invasion (Lu et al., 2018; Bănăduc et al., 2022). It is possible that G. quadriradiata, aided by human activities, initially establishes itself in accessible areas at low and mid elevations, then expands throughout mountain ecosystems (Liu et al., 2021b). Therefore, remote mountainous areas traditionally regarded as "low-risk zones" could still become potential gateways for invasive species if coupled with localized high-intensity disturbances (Pauchard et al., 2009). Future research should focus on the interactive effects of human activities and natural factors to more accurately predict the distribution of invasive species in mountain ecosystems.

4.2. Phenotypic trait variation of Galinsoga quadriradiata

The ability of seed plants to disperse over long distances determines, to some extent, the expansion of their range (Song et al., 2020). Generally, seeds with lower settlement velocities spend more time airborne and can travel longer distances (Andersen, 1992). As Galinsoga quadriradiata expands to higher elevations in China, its long-distance dispersal capability decreases, indicating that high mountains have a blocking effect on its dispersal (Liu et al., 2021b). Therefore, the dispersal process of G. quadriradiata may be greatly hindered when it expands from the low elevation area in eastern China to the high elevation areas in southwestern and northwestern China (Liu et al., 2016, 2021b). Other studies also observed a similar phenomenon in alpine plants, where high-elevation environments restrict the dispersal distance of plant seeds, thereby hindering species expansion (Naoe et al., 2016).

Elevation is not a singular geographic concept, but rather a comprehensive geographic variable that encompasses factors such as climate and soil (Mayor et al., 2017). Our findings in China reveal that in environments characterized by higher temperatures and abundant precipitation, the seed dispersal ability of Galinsoga quadriradiata is enhanced. This observation indicates that seed dispersal ability changes with elevation, suggesting that climate and soil may be potential factors driving this variation (Hernandez et al., 2023; Ni and Vellend, 2024). These results further substantiate the reliability of our conclusions. However, we observed that as the elevation increases, the seed weight of G. quadriradiata also increases (Chen et al., 2022). This may be a strategy used by invasive plants to span high elevation areas and increase the colonization and germination advantages of seeds by increasing seed weight (Pluess et al., 2005). Similar adaptive changes have also been observed in other plants, such as Parthenium hysterophorus (Rathee et al., 2021).

Although phenotypes are determined by genes, they are also directly influenced by the environment (Sommer, 2020). Compared to low-elevation areas, high-elevation areas experience lower atmospheric pressure and temperature, increased irradiance, and higher wind speed, leading to greater daily fluctuations in these environmental factors (Pandey et al., 2006). For example, in high-elevation regions, plant height, branching number, and pappus width of Eupatorium adenophorum have been shown to be higher than those in low-elevation regions (Zhang et al., 2009). Our greenhouse experiments indicate that the harsh survival conditions in high-elevation areas inhibit the growth and reproduction of Galinsoga quadriradiata, further revealing a potential inhibitory effect of high mountainous terrain on the growth and expansion of G. quadriradiata. To adapt to harsh environments, G. quadriradiata allocates more energy to growth than to reproduction (Xie et al., 2024).

4.3. Genetic variation of Galinsoga quadriradiata

Our findings on the genetic diversity of Galinsoga quadriradiata in China are consistent with those found in previous research (Li, 2016; Yang, 2019). Compared to other invasive Asteraceae species, such as Dittrichia viscosa (PPL, 94.28%) (Sevindik et al., 2023), Silybum marianum (PPL, 83%) (Jamshidnia et al., 2023), and Solidago canadensis (PPL, > 78%) (Pan et al., 2014), genetic diversity is relatively low in G. quadriradiata. This may be attributed to genetic drift and founder effects during G. quadriradiata invasions (Hagan et al., 2024). Moreover, the different degrees of human disturbance in various habitats where G. quadriradiata is distributed may led to an uneven distribution of alleles among populations, potentially causing localized differences in the frequency of environmentally related alleles and subsequently contributing to the lower genetic diversity of populations (Li, 2016). Generally, populations with high genetic diversity tend to have a higher degree of evolution and adaptability, making them more likely to survive in new environments (He et al., 2024). In contrast, invasive species with low genetic diversity may face the risk of reduced population fitness due to inbreeding depression (Liu et al., 2005). However, recent research has shown that low genetic diversity does not limit population expansion for most invasive plants (Irimia et al., 2023; He et al., 2024). Even though we found that G. quadriradiata has low genetic diversity, it has a wide distribution in China, spanning across different climatic zones. These results, along with previous research findings, suggest that the level of genetic diversity in invasive species is not necessarily correlated with their invasion success (Jiang et al., 2023).

We also found that the genetic diversity of Galinsoga quadriradiata showed a nonlinear relationship with increasing elevation in China. Typically, populations at lower elevations exhibit higher genetic diversity (Ohsawa and Ide, 2008; Reisch and Rosbakh, 2021). This is because low-elevation areas tend to have warmer and more humid ecological environments, which provide abundant resources and favorable conditions for species growth and reproduction (Ohsawa and Ide, 2008). Such environments can support larger population sizes, thereby reducing the effects of genetic drift and promoting the maintenance of genetic diversity (Schierenbeck, 2016). However, there are also viewpoints that populations in medium or high elevation regions exhibit greater genetic diversity (Shen et al., 2014; Ju et al., 2022). Specifically, we found that the genetic diversity of G. quadriradiata decreased with increasing elevation below 1500 m, meaning that its growth is more restricted at middle elevations. However, at elevations above 1500 m, the genetic diversity of G. quadriradiata showed an increasing trend. This phenomenon may be related to the fact that many samples from high elevation areas (greater than 2000 m in particular) in this study were collected from regions around roads on the Qinghai-Tibet Plateau. These areas were influenced by frequent human activities such as tourism and goods transportation, which may enhance the genetic diversity of G. quadriradiata due to its multiple introductions into higher elevation areas (Luo et al., 2020; Mairal et al., 2022a, 2022b). Therefore, the variation in genetic diversity of G. quadriradiata across different elevations, especially the increase in genetic diversity in high elevation areas like the Qinghai-Tibet Plateau, suggests that this region may face a relatively high risk of biological invasion (Dlugosch and Parker, 2008).

5. Conclusions

Our study combined global and regional scale analyses to reveal the invasion patterns of Galinsoga quadriradiata at different elevations, with a focus on the roles of human activities, phenotypic plasticity, and genetic diversity. On a global scale, we found that human activities, especially increased road density, have significant positive contributions to the species' expansion in mountainous areas. In China, we observed a significant decline in the seed dispersal ability of G. quadriradiata with increasing elevation, indicating that high-elevation areas act as a natural barrier to its expansion. However, G. quadriradiata has adapted to harsh high-elevation conditions by adjusting its growth allocation strategy. Moreover, genetic analysis revealed that its expansion has not been limited by low genetic diversity. The global scale analysis provides general patterns and a basis for risk assessment, while the regional scale study reveals key phenotypic and genetic mechanisms. This multi-scale research enhances our understanding of invasion mechanisms. In conclusion, the elevational invasion pattern of G. quadriradiata is shaped by multiple factors, particularly human activities and phenotypic adaptability, rather than being constrained by low genetic diversity. Therefore, we recommend enhanced monitoring of the species' expansion in high-elevation areas with high levels of human disturbance in the future.

Acknowledgements

We thank Xiaoyan Chen, Yingbo Yang, Wengang Zhang, and Ruiling Liu who helped to collect the materials used for the experiments. This work was supported by the National Natural Science Foundation of China (32271584 and 31600445); the Natural Science Basic Research Plan in Shaanxi Province of China (2020JM-286); the Fundamental Research Funds for the Central Universities (GK202103072, GK202103073); the National College Students' Innovative Entrepreneurial Training Plan Program (202310718085); and Special Research Project in Philosophy and Social Sciences of Shaanxi Province (2022HZ1795).

CRediT authorship contribution statement

Yu Chen: Writing – original draft, Methodology, Data curation, Conceptualization. Xingjiang Song: Writing – original draft, Methodology, Investigation, Data curation. Gang Liu: Writing – review & editing, Supervision, Methodology, Investigation, Funding acquisition. Jia Wang: Investigation, Data curation, Conceptualization. Chunling Zhang: Investigation, Data curation. Xiaojian Chang: Investigation, Data curation. Jiabin Zou: Methodology, Investigation. Zhihong Zhu: Methodology, Investigation.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.pld.2025.05.012.