Increasing invasions by alien plants in grasslands and subsequent alteration of species composition and ecological processes is a global problem, particularly in the grasslands of human dominated landscapes (Cilliers et al. 2008; Seastedt and Pyšek 2011). Annual invasive weed Parthenium hysterophorus L. (Asteraceae; hereafter Parthenim), a native of Mexico, is one of the rapidly spreading invasive plant species found in grasslands and other ecosystems in tropical and subtropical regions of the Old World. It is already well established in > 45 countries other than its native range (Maharjan et al. 2020; Mainali et al. 2015; Mao et al. 2021). It is an invasive weed of the global concern due to its’ significant negative ecological and socio-economic impacts (Bajwa et al. 2016; Mao et al. 2021). Parthenium weed grows well in disturbed habitats such as the grasslands in urban and periurban areas, and various other natural habitats such as grassland, open woodlands, flood plains and river banks (Chhogyel et al. 2021; Gupta and Narayan 2006; Shrestha et al. 2019a; Tamado and Milberg 2000). High invasiveness of this weed has been attributed to its prolific seed production, persistent soil seedbank, intermediate C3-C4 mechanism of photosynthesis, high tolerance to abiotic stresses, and allelopathic potential (Bajwa et al. 2016). A typical mature Parthenium plant produces about 15000 seeds (Haseler 1976). It has been found that the large percentage of buried seeds of Parthenium can survive for one to several years (Butler 1984; Navie et al. 1998) and a study has reported that seeds of Parthenium can contribute up to 84% of the germinable soil seed bank (Ayele et al. 2013).

Parthenium weed is attributed in loss of crop and livestock production (Bajwa et al. 2019) and displacement of native flora (Hejda et al. 2017; AMollot et al. 2017) by modifying soil nutrients (Rokaya et al. 2020; Timsina et al. 2011). This often leads to the potential elimination of many useful forage plants and thus decrease pasture productivity (Navie et al. 1996). Additionally, Parthenium weed has a range of impacts on human health causing allergy, contact dermatitis, bronchitis, asthma and fever (Allan et al. 2019). A recent study in Africa has also shown that allelochemicals released from root of Parthenium can modulate oviposition behaviour and survivorship of malaria mosquito vector Anopheles gambiae (Milugo et al. 2021). Although a diversity of ecological and socio-economic impacts of this weed has been reported from different countries, there is geographic bias in studies. Most of the studies were undertaken in India, Australia and Pakistan (Maharjan et al. 2020). These are also the countries where national efforts to manage Parthenium weed is relatively high. One of the reasons for the low national efforts for the management of this weed could be due to inadequate empirical studies showing its impacts (Maharjan et al. 2020). As for example, Parthenium weed is already widespread in Nepal, but there are virtually no national efforts to control this weed (Shrestha et al. 2015; Shrestha et al. 2019a). Therefore, it is imperative to generate additional data on the impacts of this weed. This allows us to inform and encourage concerned national authorities to formulate appropriate management plan.

Previous studies in Nepal have revealed that Parthenium weed alters species composition of native plant species (Rokaya et al. 2020; Timsina et al. 2011), changes soil nutrient content (Timsina et al. 2011), causes health hazards to human (Shrestha et al. 2015), and affects peoples’ livelihood and food security (Shrestha et al. 2019b). Although the impacts on aboveground vegetation have been examined, the impacts of Parthenium weed on soil seedbank remains much unexplored in Nepal, except a study by Rokaya et al. (2020) examining how soil seedbank changes over the time in highly invaded sites. As all plants growing in a particular area do not contribute to soil seedbank, examining the soil seedbank allows us to identify such plant species which contribute to soil seedbank (Coffin and Lauenroth 1989). Soil seedbank studies also suggests regeneration potential and future composition of the plant community, especially to those that rely mostly on non-vegetative means of reproduction. Comparing the species composition at different levels of Parthenium invasion helps to evaluate its exact and potential impacts on plant community composition. In the present study, we specifically answered following research questions: (1) Do plant species richness and composition vary among sites of the different levels of Parthenium invasions? (2) Is there any difference in species diversity (species richness and composition) between aboveground and below ground? (3) What is the extent of contribution of Parthenium to the soil seedbank? To answer above-mentioned questions, we sampled vegetation in low, medium and high Parthenium invaded sites, compared above ground and below ground plant species, and estimated germinable soil seedbank density in central Nepal.

Study area

The study area is located in the Siwalik region of Makawanpur district in south-central Nepal (Fig. 1). Geologically, the Siwalik region represents Sub-Himalayan zone (southernmost front of Himalaya sensu lato) with sandstone, siltstone, mudstone and boulder conglomerate (Upreti 1999). The area lies at the eastern part of the east-west oriented Chitwan Valley which is the largest inner Tarai Valley in central Nepal, and extends into Makawanpur, Chitwan and Nawalpur districts. The region receives monsoonal rainfall and has a dry subtropical climate with hot-humid summers (monsoon) and cold-dry winters. Annual rainfall is about 2430 mm (> 80% during four monsoon months [June–September]) and mean temperature ranged from 14°C (January) to 28°C (June) (based on the weather data recorded at the nearest weather station Nepal Forestry Institute at Hetaunda; Marahatta et al. 2009). Sal (Shorea robusta) forest is the dominant vegetation of the Valley whereas small patches of tropical deciduous riverine forests, Dalbergia-Acacia forest, and riverine grasslands are also found in the Valley.

Figure 1. Location of Makawanpur district in Nepal (α), the study location in Makawanpur district (β) and the sampling sites in the grasslands (c).

The sampling sites are located in Hetaunda Municipality which is the district headquarter and the most populated and urbanized city of Makawanpur district. The sampling sites (27° 25’ N latitude, 85° 03’ E longitude, 450–490 m asl elevation) were the grasslands located in the property of Hetaunda Cement Factory (two of the three sites) and Hetaunda Industrial Area (one of three sites) (Fig. 1). The grassland had been grazed throughout the year by domestic animals, and fodder collection activities were high during the summer. These sites were also subjected to frequent sport activities during winter months. Parthenium invasion was high at all sites with only a few Parthenium free patches. Other invasive weeds such as Lantana camara, Mesosphaerum suaveolens, Senna tora, among others, were also present in the study area.

Vegetation and soil seedbank studies

The vegetation sampling was carried out during late monsoon season (September) along 30 transects of about 10 m long with 10 transects at each of the three sites. Along each transect, three plots of 1 m × 1 m were sampled in such a way that they represented > 90%, > 40% to ≤ 60% and < 10% cover of Parthenium, corresponding to high, medium and low levels of Parthenium invasion, respectively. During the sampling, vascular plant species rooted in plots were recorded to determine the plant species richness and composition. Numbers of Parthenium individuals were also recorded to determine Parthenium density. Maximum height of Parthenium and coverage of other plant species inside each plot were also recorded. After vegetation sampling, each plot with high level of Parthenium invasion was marked with wooden pegs for collecting soil samples in the next month to study soil seedbank.

Soil samples were collected after one month of vegetation sampling (i.e., October) when most of the plants including Parthenium dispersed their seeds. They were collected from the plots having high (> 90%) Parthenium coverage. For this, each plot was divided into four sub-plots (50 cm × 50 cm) and soil sample was collected from one of these sub-plots selected randomly (lottery method). Prior to the soil collection, the aboveground vegetation and any plant debris (but not seeds) were removed. Soil samples were collected with the help of metallic soil core sampler (diameter 10 cm). Soil was collected in two layers: from surface to 5 cm (0–50 cm), and from 5 to 10 cm depths (5–10 cm). Altogether there were 60 soil samples collected from 30 plots.

Seedling emergence method was used to assess the germinable soil seedbank (Simpson 1989). This method has been also used previously for the study of soil seedbank of Parthenium weed (Ayele et al. 2013; Navie et al. 2004). Collected soil samples were observed for eight months in the greenhouse to determine the germinable soil seedbank. For this, earthen pots (15 cm diameter) were filled up to 2/3rd part with heat sterilized sand. Then soil sample was spread uniformly over the sand layer such that depth of the soil sample was ≤ 2 cm. Sand and soil sample was separated by a paper to prevent their mixing. There were altogether 60 pots. Soil was maintained moist by regular watering. Number of seedlings of each species emerged in each pot was recorded weekly during first one month and in two weeks period during next seven months. In each observation, once Parthenium seedlings and other species were identified they were removed to avoid crowding. Those species which could not be identified in vegetative stage were allowed to flower for proper identification.

Specimens of all plant species encountered during vegetation sampling as well as those grown in green house, were collected, tagged, dried and then mounted on the herbarium sheets. They were identified with the help of local flora (e.g., Siwakoti and Varma 1999) and by comparing with the specimens housed at Tribhuvan University Central Herbarium (TUCH), Kirtipur and National Herbarium and Plant Laboratories (KATH), Godawari. We followed Roskov et al. (2019) for the nomenclature.

Numerical analysis

Community and population attributes recorded in plots with different intensity of Parthenium invasion were compared either by using one-way analysis of variance (ANOVA) or Krushkal–Wallis test depending on the nature of the data. None of the original data met the assumption of the homogeneity of variance. Variance became equal after species richness (number of species per plot) and Parthenium density were log transformed. Therefore, mean values of these attributes were compared among plots with different intensity of Parthenium invasions using ANOVA (using log transformed values) followed by Duncan homogeneity test. Other attributes (e.g., maximum height of the Parthenium, Parthenium cover, and other plant species cover) did not meet the assumption of equal variance even after transformation. Therefore, they were compared using non-parametric Kruskal–Wallis test followed by pair-wise Mann–Whitney U-test. These non-parametric tests use median values for significance test of difference. These statistical analyses were done using Statistical Package for Social Science (SPSS Inc., Chicago, IL, USA).

Germinable soil seedbank density (seeds/m²) of Parthenium as well as other species were determined for two soil depths (0–5 cm and 5–10 cm). Percentage contribution of Parthenium to total germinable soil seedbank was also calculated. Similarity Indices (Jaccard’s and Sorenson’s Similarity Indices) were calculated to assess the similarity of the species composition between aboveground vegetation and the corresponding soil seedbank following Zobel et al. (1987).

Multivariate tests were used for plant species composition in plots with different levels of Parthenium invasions and also to compare above ground flora and below ground soil seedbank flora. For the plots with different levels of Parthenium invasions, the gradient length was 2.68 and for the above ground and belowground flora, gradient length was 4.02. So, we used Canonical Correspondence Analysis (CCA) in both analyses (Šmilauer and Lepš 2014) by using Canoco 5.12 (ter Braak and Šmilauer 2012). In CCA, model was built in the way to know associated plant species in low, medium and high invaded plots, and also above ground and below ground. The significance was tested by performing Monte Carlo permutation test (n = 4999) and we down weighted the rare species to reduce their effect on the results.

Species richness and composition at different levels of parthenium weed invasion

We recorded a total of 62 plant species including Parthenium in the plots of high, medium and low levels of Parthenium invasions (Additional file 1: Table S1). Among them, 59 species were present in the plots with high Parthenium invasion, 48 species in the medium, and 48 species in the plots with low Parthenium invasion. There was no significant difference in the species richness among various levels of Parthenium invasion though Parthenium density, maximum height, cover of Parthenium, and other vegetation cover were considerably diverse at different levels of Parthenium invasion (Table 1).

Table 1 . Community and population attributes measured in plots of the different levels of Parthenium invasion.

Attributes		Species richness (#species/m2)	Parthenium density (stem/m2)	Max. height of Parthenium (cm)	Parthenium cover (%)	Other vegetation cover (%)
Levels of Parthenium invasions	High	15a ± 5	410c ± 122	423c (201)	95c (10)	78b (30)
	Medium	15a ± 3	126b ± 23	147b (147)	45b (6)	98a (6)
	Low	14a ± 4	34a ± 6	8a (19)	4a (4)	100a (11)
Test statistics	F	1.68	106.89	-	-	-
	χ2	-	-	76.92	79.99	33.33
	Degree of freedom	2, 87	2, 87	2	2	2
	p-value	0.192	< 0.001	< 0.001	< 0.001	< 0.001

Main entries for values of species richness and density of Parthenium are mean ± standard deviation; values for other parameters are medians (interquartile range).

Attributes analyzed by ANOVA have their F values and those analyzed by Kruskal–Wallis test have chi-square (χ²) values. Values within each column with different alphabets (^a, ^b, ^c) in superscripts are significantly different at p < 0.05.

The species composition in plots with different levels of Parthenium invasion were significantly different (p = 0.004, Fig. 2). Most plant species were recorded across all the plots. Few species such as Ageratum conyzoides, Brachiaria ramosa, Murdannia sp., Bothriospermum tenellum, Hemarthria compressa, and Crotalaria prostrata were strictly found in highly invaded plots whereas Oxalis corniculata and Corchorus aestuans were found in medium invaded plots.

Figure 2. Species composition in plots with different levels of Parthenium invasion. First Canonical Correspondence Analysis (CCA) axis explained 3.46% of variation and second axis 0.85%. Full names of the species are in Additional file 1 (Table S1).

Comparison between above ground and belowground flora

We recorded a total of 46 species in soil seedbank (Additional file 1: Table S2); 42 species were from top layer (0–5 cm depth) and 25 species were from deeper layer (5–10 cm depth). Jaccard’s and Sorensen’s Similarity Indices between standing (above ground) flora in the plots having high levels of Parthenium invasion and the corresponding soil seedbank (below ground) flora were 19% and 32%, respectively.

The composition of associated plants varied significantly (p = 0.002) in between above ground flora and soil seedbank (below ground flora) with more plant species being reported from above ground (Fig. 3). Species such as Phyllanthus urinaria, Desmodium triflorum, Borreria alata, Polygala sp., Fimbristylis dichotoma, Chrysopogon aciculatus, Echinochloa colona, and Imperata cylindrica were found as above ground flora. Whereas, Carex sp., Phyllanthus sp., Oxalis corniculata, Bidens pilosa, Mecardonia procumbens, Ageratum sp., and Sonchus wightianus were strictly found as below ground flora.

Figure 3. Species composition of above ground flora and soil seedbank in plots with high Parthenium invasion. First Canonical Correspondence Analysis (CCA) axis explained 18.79% variation and the second axis 6.96%. Full names of the species are in Additional file 1 (Tables S1 and S2).

Contribution of parthenium weed on soil seedbank

Parthenium had the greatest germinable soil seedbank density in both soil depths. However, it was 14 times higher (p < 0.001) in the soil from the top layer (0–5 cm; 8624 ± 6509 seeds/m²) than in the deeper layer (5–10 cm; 620 ± 818 seeds/m²). The overall soil (0–10 cm depth) seedbank density of Parthenium was 9244 ± 6535 seeds/m², contributing 68% to the total germinable soil seedbank in the top layer (0–5 cm) and 38% in the deeper layer (5–10 cm). On considering the entire depth of the soil (0–10 cm), Parthenium contributed to 65% to the total soil seedbank.

Impacts of Parthenium on species composition and richness

Our study showed that Parthenium invasion has modified the species composition without significant impact on species richness. Plots with different levels of Parthenium invasion showed variation of species composition. Previous studies have also shown that invasive species including Parthenium may have little effect on species diversity but have significant effects on the species composition (Hejda and Pysek 2006; Martin 1999; Timsina et al. 2011). Five-year long study conducted in the same area of the present study also showed a significant shift in species composition with increase in Parthenium abundance over the time (Rokaya et al. 2020). Such compositional change in community due to Parthenium invasion may result in disappearance of native species (Ayele et al. 2013). Change in species composition may be driven by species and population traits of Parthenium which are responsible for its high invasiveness. These traits that contribute to high invasiveness of this plant are high canopy cover and density, prolific seed production, high nutrient use efficiency, and allelopathic potential (Bajwa et al. 2016; Rokaya et al. 2020). Some resident species growing in the invaded sites are more sensitive than others to the allelopathic potential of Parthenium (Belgeri and Adkins 2015; Paudel and Shrestha 2019). Sensitive species may be displaced by Parthenium easily while the resistant species may persist in or recolonize the Parthenium invaded micro-habitat. Additionally, change in species composition may also be driven by a change in disturbance regime at micro-habitat level due to Parthenium invasion which is not uniform spatially. For example, grazing intensity is likely to be low at highly invaded plots because livestock tends to avoid such micro-habitat due to low availability of forage (Rokaya et al. 2020; Shrestha et al. 2015; Timsina et al. 2011).

There was no significant difference in species richness among plots of different levels of Parthenium invasion. This result contradicts with other studies, which suggested that invasive species have negative effects on species richness (Bimová et al. 2004; Dunbar and Facelli 1999; Kohli et al. 2004). Kohli et al. (2004) reported a decline in species richness from 25 to 12 from Parthenium non-invaded site to high invaded site in lower Himalaya (India). Rokaya et al. (2020) found sharp decline in plant species richness from 2009 to 2012 suggesting adverse effect of increasing Parthenium invasion on many native species. They also reported a significant decline in species richness with increase in Parthenium height. However, in grazing land the species richness may increase at the early stage of invasion (i.e., before the Parthenium forms pure dense stand by replacing other species) due to grazing exclusion and shift of dominance (Timsina et al. 2011). Timsina et al. (2011) reported lower species diversity in the Parthenium non-invaded plots than in intermediately invaded plots. In our study, lack of significant difference in species richness between plots of high and low levels of Parthenium invasion might also be the outcome of positive impact of defoliation of Parthenium by beetle Zygogramma bicolorata, a biological control agent, or shorter duration of research time period than Rokaya et al. (2020). During the sampling, we observed > 60% of leaves of Parthenium damaged by the beetle, allowing adequate light to reach the ground surface. That environment might be less hostile than in the stands with intact leaves. In absence of such control measure, and under intense livestock grazing, Parthenium can competitively displace pasture species, reduce pasture vigor and seed set leading to habitat and ecosystem change (Evans 1997; O’Donnell and Adkins 2005; Shabbir and Bajwa 2006). The current and the previous studies discussed above have suggested that the ecological impacts of Parthenium invasion on species composition and diversity can be modified by disturbance regime shift (e.g., livestock grazing) and control measures such as the presence of biological control agents.

Comparison between above ground and belowground flora

There was very low species similarity (Jaccard’s Similarity Index being 19% and Sorensen’s similarity index being 32%) between standing flora (above ground) and soil seedbank (below ground) flora. Our result is in line with Looney and Gibson (1995) who showed the value for Jaccard’s Similarity Index to be 36% between standing flora and soil seedbank in coastal barrier island of Florida, the USA. We found significant difference in the species composition of above ground flora and soil seedbank. Such differences are also reported earlier (e.g., Ayele et al. 2013; Török et al. 2009). Ayele et al. (2013) compared the vegetation composition between the standing flora and soil seedbank flora in the plots with different levels of Parthenium invasion and found the highest similarity in the Parthenium non-invaded plots and the lowest similarity in the plots with high Parthenium invasion, indicating the impact of Parthenium invasion. High degree of allelopathy might have suppressed the germination and growth of the native species in highly invaded plots (Bajwa et al. 2016). Since Parthenium invasion has impacts on species composition, change in soil seedbank composition can be correlated with the change in species composition in the following year.

Beside the allelopathic effect of Parthenium, there are various reasons for affecting soil seedbank species composition. First, annual species tends to dominate soil seedbank, and perennial species to the aboveground vegetation in Parthenium invaded sites (Ayele et al. 2013). Second, loss of seeds due to predation and loss of seed viability also affect the plant species composition of soil seedbank (Schafer and Chilcote, 1970; Wilson et al., 1993). Third, mostly the annual forbs show high seed dormancy, low seed predation, and high persistence in soil seedbank which favor the germination of most of their seeds in the coming season. Whereas, in case of perennial grasses, most of the seeds get lost before germination due the low level of seed dormancy and high vulnerability to predators. Buried seeds might also fail to emerge due to death as a result of physiological aging and seed decomposition (Baskin and Baskin 2014). Fourth, non-dormant seeds of some species may reenter dormancy if environmental conditions are unfavourable for germination thereby affecting the seedbank dynamics (Baskin and Baskin 2014). Thus, the soil seedbank dynamics could also be the reason for the low similarity between aboveground and belowground flora in the present study area.

Contribution of Parthenium weed on soil seedbank

Parthenium had the highest value of germinable soil seedbank density among all the germinated species recorded in the soil samples collected from the present study site. Parthenium alone contributed about 65% to the total germinable soil seedbank density in soil samples from 0–10 cm depth. Dominance of Parthenium in soil seedbank has been also reported in other parts of the world. For example, in Moolayember Creek, Australia, Parthenium accounted for 65%–87% of the total soil seedbank (Navie et al. 2004). Similarly, in southeastern Ethiopian rangeland, Parthenium contributed up to 84% to total soil bank under the high weed invasion level (Ayele et al. 2013). Greater contribution of Parthenium in the soil seedbank might be due to its prolific seed production, persistence nature of soil seedbank, high viability of buried seeds, low seed predation, and innate dormancy mechanism of its seeds (Baskin and Baskin 2014; Haseler 1976; Joshi 1991). Parthenium is a short-lived weed possessing long lived seeds often with secondary dormancy mechanism (Rice 1989). Germination inhibitors (parthenin and phenolic acids) present in the accessory structures and seed coat of Parthenium seeds act as autotoxins that temporarily prevent the germination of nearby Parthenium seeds if their density is sufficiently high (Picman and Picman 1984) thereby decreasing the seed losses because of germination. Parthenin also increases seed survival by discouraging the decay (Ganeshan and Jayachandra 1993; Rice 1984) or predation (Ahmed and Bhattacharya 1991) of seeds.

A typical mature plant of Parthenium can produce about 15,000 seeds (Haseler 1976) forming enormous soil seedbank. Accumulation of seeds to the persistent soil seedbank of Parthenium leads to the increase in the size of its soil seedbank (Adkins and Shabbir 2014; Navie et al. 1998). The average seed density of Parthenium was 9244 seeds/m² for 0–10 cm soil depth in the present study site. Previous studies have reported a range of soil seedbank density of Parthenium: 20600 to 44600 seeds/m² at the Moolayember Creek, Australia (Navie et al. 2004); 300 seeds/m² at very low Parthenium invaded sites and 1880 seeds/m² at high Parthenium invaded sites in Ethiopia (Ayele et al. 2013). From the current study area, Rokaya et al. (2020) reported Parthenium soil seedbank at 5000 to 9000 seed/m² over the period of five years. Relatively smaller size of soil seedbank densities of Parthenium in our study site compared to Australian sites might be because of its recent invasion in Nepal which is not enough to contribute large soil seedbank as that of Australia. Low soil seedbank may also be attributed to the activities of biological control agents (e.g., Zygogramma bicolorata in the study site) (Dhileepan et al. 2018) but it requires further investigations in the present study site. The variation in soil seedbank is also attributed to the climatic factors and soil conditions which vary spatially as well as temporally (Baskin and Baskin 2014).

Most of the germinable seeds of Parthenium concentrated at the top layer (0–5 cm) of the soil which was supported by other seedbank studies (Fenner 2000; Török et al. 2009). More than 70% of the seeds buried below 5 cm soil depth remain viable for 2 years (Adkins and Shabbir 2014). Greater percentage of seedling emergence from the soil samples of top layer showed the higher accumulation of seeds to surface soil. The deep burial of seed is less likely in the area where there is less soil disturbance. The present study site being the grazing land, soil disturbance in this case might be just due to trampling effect of grazing animals which might lead to the shallow burial of the seeds (Baskin and Baskin 2014). Deep burial of seed in such case requires very long time. The lesser percentage of germinable seeds from the deeper layer could be also related with its invasion history which is not very long in the sampling site.

In a grassland subjected to livestock grazing, we showed that Parthenium invasion significantly altered species composition of the aboveground vegetation without significantly affecting the species richness. There was a large difference in species composition between aboveground vegetation and soil seedbank; and Parthenium accounted > 2/3rd of the total soil seedbank density. The present results suggest that multiple parameters have to be considered to fully understand the ecological impacts of the invasive alien plant species and to inform management and policy decisions. High dominance by Parthenium weed in soil seedbank is particularly challenging because of its ability to remain viable in soil for a prolonged period. The results suggest the need of monitoring the soil seedbank dynamics while managing Parthenium weed. Control measures such as the use of biological control agents can reduce seed output and deplete soil seedbank over the time which may eventually give way to native species for re-colonization of the habitats. Given that a biological control agent Zygogramma bicolorata is already present in the study area, impacts of this agent on soil seedbank dynamics and restoration of native plant communities can be an important area for future research.

Supplementary information accompanies this paper at https://doi.org/10.5141/jee.21.00083.

Additional file 1: Table S1. List of plant species recorded in above ground vegetation in the sample plots. Table S2. List of plant species at two different soil depths (0–5 cm and 5–10 cm) recorded during the soil seedbank study.

We are thankful to Ambika Paudel for help in field data collection and laboratory works, and Mahesh Bisht for preparing map of the study area. We are also thankful to Tribhuvan University Central Herbarium (TUCH) and National Herbarium and Plant Laboratories (KATH) for their permission to examine herbarium specimens for plant identification.

CCA: Canonical Correspondence Analysis

KATH: National Herbarium and Plant Laboratories, Nepal

SPSS: Statistical Package for Social Sciences

TUCH: Tribhuvan University Central Herbarium, Nepal

BBS conceptualized the research and did experimental design. JKC and BBS collected field and laboratory data. MBR and JKC analyzed data. JKC prepared first draft of the manuscript. BBS, MBR and JKC revised and finalized the manuscript.

JKC, MSc, is an Assistant Professor at Botany Department, Trichandra Multiple Campus, Tribhuvan University, Kathmandu, Nepal. She has completed her two years master’s degree in Botany (specialization is ecology). She has research interests in invasion ecology, soil seedbank study, plant allelopathy, and plant ecophysiology.

MBR, PhD, is a research fellow at Czech Academy of Sciences (Institute of Botany and Global Change Research Institute). His research interests are on biological invasions, economically important plants and spatial ecology mainly along elevational gradients in Himalayan region.

BBS, PhD, is an Associate Professor at the Central Department of Botany, Tribhuvan University, Kathmandu, Nepal, with research interests on biological invasions, forest ecology, and plant adaptation along environmental gradients in mountains. He is currently involved as Lead Author in thematic Assessment on Invasive Alien Species which is being commissioned by the Intergovernmental Science-Policy Platform for Biodiversity and Ecosystem Services (IPBES).

Field and laboratory works were supported by a grant from Nepal Academy of Science and Technology (NAST) to BBS. Partial financial support was also received by JKC from Cornell Nepal Student Programm (CNSP), Kirtipur, Nepal. MBR is partly supported by long-term research development project No. RVO 67985939 of the Czech Academy of Sciences and institutional support for science and research of the Ministry of Education, Youth and Sports of the Czech Republic.

The datasets supporting the conclusions of this article are included within the article (and its additional files).

Not applicable

Not applicable

The authors declare that they have no competing interests.