Introduction

Biological invasions are widely recognized as a major threat to global biodiversity and community structure (Ehrenfeld 2010; Simberloff et al. 2013); however, why a specific alien plant can successfully invade is still inconclusive. The enemy release hypothesis (ERH) predicts that the proliferation of an invasive plant is due to the fact that it escapes from the regulation by specialized herbivores and other natural enemies in the introduced range (Keane and Crawley 2002). Based on ERH, the evolution of increased competitive ability hypothesis (EICA) assumes that invasive plants may experience evolutionary shifts in resource allocation from defense to growth to increase their competitive ability due to release from coevolved enemies (Blossey and Notzold 1995). Although many studies have provided support for the EICA hypothesis that some invasive plants do have a competitive advantage over their native conspecifics (Beaton et al. 2011; Gruntman et al. 2014; Jakobs et al. 2004), some others have only partially supported this hypothesis (Meyer et al. 2005) or did not support it at all (Parker et al. 2013; Zheng et al. 2015). Thus, it is still unclear whether invasive plants are in general competitively superior to native plants.

In novel environments, invasive plants may confront both heterospecific and conspecific neighbors, and respond to their neighbors either in a positive (facilitation) or a negative (competition) way (Callaway and Walker 1997; Zheng et al. 2015). The pattern and the intensity of both inter- and intraspecific competition might differ greatly between invasive and native species (Huang et al. 2012a; Zheng et al. 2015). Classical competition theory predicts that intraspecific competition should be greater than interspecific competition because individuals within one species require similar resource conditions (Adler et al. 2018); however, some studies have suggested that the intensity of interspecific competition between native and invasive species could exceed the impact of intraspecific competition within species because of limited resources (Sheppard and Burns 2014). Both interspecific and intraspecific competition abilities are significant determinants of invasive success and population dynamics of alien plants in their introduced ranges (Mangla et al. 2011).

Another important factor that may affect the outcomes of competition of invasive and native plants is herbivory, which can potentially alter the magnitude and/or outcomes of competition either by causing greater damage to a dominant competitor or by causing a similar level of damage among species (Huang et al. 2012b; Wang et al. 2017; You et al. 2016). Herbivory can reduce plant fitness as well as the ability of plants to compete with neighbors, and its effects can be increased by competition (Backmann et al. 2019; Center 2005). Invasive plants are considered to be largely released from specialists, but may also be attacked by a diverse array of oligophagous and polyphagous herbivores from both aboveground (AG) and belowground (BG) compartments in their introduced ranges (Joshi and Vrieling 2005). Feeding by those herbivores may influence the outcomes of intraspecific and/or interspecific interactions of an invader with its native neighbors (Ibanez et al. 2013; Kim et al. 2013). Moreover, interactions between AG-BG herbivores may change the relative strengths of plant competition in additive or antagonistic ways through their effects on shared host plants (Bezemer and van Dam 2005; Wardle et al. 2004). Although an increasing number of studies have focusing on invasive plant-mediated interactions between AG-BG herbivores (Bezemer and van Dam 2005; Huang et al. 2012b; Li et al. 2020; Ohgushi et al. 2018; Wardle et al. 2004; Wei et al. 2016), few have tested effects of such interactions on the competition ability of invasive plants while interacting with different heterospecific and conspecific neighbors (Huang et al. 2012a; Jing et al. 2015).

Alternanthera philoxeroides is a perennial clonal herb that is native to South America and is aggressively invasive in temperate, tropical and subtropical areas of both aquatic and terrestrial environments of China (Pan et al. 2007) and elsewhere. In China, A. philoxeroides commonly reproduces via regeneration from clonal fragments. Alternanthera sessilis is the only native congener of A. philoxeroides and is often sympatric with A. philoxeroides in China (He et al. 2014; Lu and Ding 2010; Wei et al. 2016). However, A. philoxeroides always outperforms A. sessilis in the field. Previous studies have suggested that A. philoxeroides is more phenotypically plastic than A. sessilis across a wide range of growth, morphological, physiological and fitness-related traits (Geng et al. 2006; Zhang et al. 2021), and these advantages are believed to enhance the competitive capacity of A. philoxeroides and promote its invasion success in heterogeneous and adverse environments (Geng et al. 2006; Wang et al. 2018, 2021). In China, A. philoxeroides is damaged by a series of AG and BG herbivores, including the monophagous leaf beetle A. hygrophila (Coleoptera: Chrysomelidae) introduced from the native range, the indigenous oligophagous tortoise beetle C. piperata (Coleoptera: Chrysomelidae), and several generalists such as a root-knot nematode Meloidogyne incognita (He et al. 2014; Wei et al. 2016). Alternanthera philoxeroides exhibits higher tolerance to simulated and actual herbivory than A. sessilis in many environments (Lu and Ding 2012; Sun et al. 2010), and the invader can allocate resource flexibly from leaves to roots to tolerate AG herbivory or joint AG-BG herbivory (Sun et al. 2010; Wei et al. 2016). Invasive plants may have evolved in response to different monophagous, oligophagous, and generalist herbivores (Agrawal and Fishbein 2006), and thus, it is necessary to assess the impacts of individual herbivores and species combinations to provide new insights into the evolutionary changes of invasive plants. However, it is still unclear how the performance of A. philoxeroides and its functional counterpart A. sessilis would be affected by biotic interactions between multiple AG and BG herbivores under intra- and interspecific competition environments.

In this study, we conducted a common garden experiment by planting A. philoxeroides and its indigenous congener A. sessilis, either mixed with each other or planted individually, while subjecting both plants to AG herbivory by the leaf beetle A. hygrophila or the tortoise beetle C. piperata, and BG herbivory by the root-knot nematode M. incognita. We sought to compare how the growth and inter- and intraspecific competitive ability of A. philoxeroides and A. sessilis were differently affected by the interactions of AG and BG herbivores. Specifically, we asked the following questions: (1) How does AG-BG herbivory affect the outcomes of intra- and interspecific competition of the two plants? (2) How does joint herbivory by A. hygrophila (or C. piperata) and M. incognita differ in its effects on competitive outcome? (3) How do invasive and native plants vary in their responses to interactive effects between competition and damage by different types of herbivores? We predicted that: (1) both competition and AG-BG herbivory interactions would influence plant performance, but would be highly dependent on competition type and herbivore species; (2) AG-BG herbivore interactions would alter the outcomes of intra- and/or interspecific competition through differential impact on neighbour plants.

Material and methods

Rearing of plants and herbivores

Plants of each species (A. philoxeroides and A. sessilis) were collected from agricultural land in Nanning (108°24′E, 22°84′N), Guangxi Province, China, and propagated vegetatively at ambient temperature and light in a greenhouse at Guangxi University. Day temperatures in the greenhouse were about 28–33°C, night temperatures were about 24–28°C, relative humidity was about 75%. Stem fragments (2 cm) with a stem node centered around the node were cut from shoots of A. philoxeroides and A. sessilis and then planted in the seedling trays (54 × 28 cm, 50 holes) containing mixtures of loam and peat soil (1:1). Plants were grown under a natural photoperiod and watered regularly to maintain soil moisture. Three weeks later, the fragments of A. philoxeroides and A. sessilis that had similar growth states were used for the experiments.

Adults of the leaf beetle and the tortoise beetle were collected from alligator weed at a field adjacent to Guangxi University, Nanning, China, and each species was reared separately in containers (10 cm in diameter, 18 cm in depth, covered with gauze) with fresh ramets of A. philoxeroides as food. After the adult beetles (of each species) produced eggs, the leaves with eggs were transferred to new Petri dishes. Newly hatched larvae were reared with fresh ramets of A. philoxeroides until they pupated. Newly emerged adults were raised individually in Petri dishes to prevent mating and were used in the experiment. The root-knot nematode M. incognita was also collected from the roots of A. philoxeroides and A. sessilis at a farmland adjacent to Guangxi University, Nanning, China.

The common-garden experiment

We designed a 24-factorial common-garden experiment consisting of 2 plant origins (invasive, and native plants) × 2 competition types (intraspecific, and interspecific competition) × 2 BG herbivory treatments (with nematodes, and without nematodes) × 3 AG herbivory treatments (A. hygrophila, C. piperata, and without herbivory) (Fig. 1). Each treatment was replicated for 6 times (144 replicated combinations).

Fig. 1
figure 1

Experimental design. The experiment employed a four-way factorial design including two plant origins (invasive Alternanthera philoxeroides and native Alternanthera sessilis) × two competition types (intraspecific competition, or interspecific competition) × two belowground (BG) herbivory treatments (no nematodes, or nematodes) × three aboveground (AG) herbivory treatments (no herbivory [= “control”], herbivory by the leaf beetle Agasicles hygrophila [= “leaf beetle”], or herbivory by the tortoise beetle C. piperata [= “tortoise beetle”])

Full size image

For the interspecific competition treatment, one fragment of A. philoxeroides was cultivated in a pot (25 cm diameter × 16 cm height) along with one fragment of A. sessilis; but for the intraspecific competition treatment, two fragments of A. philoxeroides or A. sessilis were planted in the same pot. In addition, one fragment of A. philoxeroides or A. sessilis were planted in monoculture, which were set as the control for calculating relative neighbour effect (RNE) in the mixed culture experiment. Each pot was covered with a 0.15 mm nylon mesh to exclude herbivores other than the test species.

After one month of growth, each plant was randomly assigned to one of two groups of nematode treatment before nematode inoculation (0 or 10,000 M. incognita eggs per plant). The nematode suspension was prepared according to Liu (2000). In brief, the roots of tomatoes infested by M. incognita were rinsed with water and cut into about 1 cm length. These segments were blended in a juicer, and then the suspension was successively passed through sterilized 200 mesh and 500 mesh suspension screens, being washed with sterilized water. Finally, the filter on the 500 mesh was washed into a centrifuge tube, centrifuged, the concentration of the suspension was adjusted to 1000 eggs·mL-1 with distilled water. To inoculate plants with nematodes, five holes (3 mm deep) were created with a hole puncher near the plant’s roots, and 2 mL of nematode suspension were injected into each hole. The injection site was then covered with soil. For the control plants with no nematodes, similar holes were created and injected with 2 mL of sterilized water.

One month after nematode inoculation, one adult of the leaf beetle or the tortoise beetle was released individually on each plant under the corresponding AG herbivory treatment. After 10 days of feeding, all the beetles were removed. One week later, the stem diameter, the ramet number, the stolon length, and the root length of each plant were measured, respectively. Upon harvest, all the plants were cut, AG and BG plant parts were dried separately at 65°C until constant weight and weighed. The total biomass of each plant was then calculated.

Statistical analyses

Both interspecific and intraspecific interactions between A. philoxeroides or A. sessilis were quantified using the RNE index (Markham and Chanway 1996):

$${\text{RNE}} = (P_{-N}-{}P_{ + N})/x$$

where P-N is the total biomass of the species planted in monoculture and P+N is the total biomass of the same species in the presence of a neighbour species. The term x depends on which value of P is greater. If P-N>P+N, then x = P-N; however, if P-N < P+N, then x = P+N. RNE index values fall into a symmetrical range from +1 to −1 with negative values indicating facilitation and positive values indicating competition. An RNE value of 0 indicates that no interaction is occurring.

A series of two-way ANOVAs were conducted to test the effects of AG herbivory treatment (leaf beetle, tortoise beetle, or control), BG herbivory treatment (no nematodes, or nematodes), and their interactions on the growth traits (stem diameter, ramet number, stolon length, root length, and biomass allocation) and on the resultant RNE indexes of A. philoxeroides and A. sessilis under intra- or interspecific competition conditions, separately. AG herbivory treatment and BG herbivory treatment were taken as fixed factors. Tukey’s HSD tests were applied to test for pairwise differences among treatments. Significant level was set at P < 0.05. All data were statistically analyzed by the statistical software package SPSS 18.0 (SPSS Inc., Chicago, USA), and all figures were plotted by SigmaPlot 12.5.

Results

Growth of the two test plants

When the two plants were grown in monoculture (intraspecific competition), the stem diameter, ramet number, stolon length, AG biomass, and total biomass of A. philoxeroides and the BG biomass of A. sessilis were all significantly affected by the AG herbivory treatment (Table 1). The above mentioned growth parameters of A. philoxeroides were only significantly reduced by the leaf beetle A. hygrophila, except for the stem diameter of A. philoxeroides and the BG biomass of A. sessilis, which were also significantly reduced by the tortoise beetle C. piperata, compared to the plants with no AG herbivory (Figs. 2, 3). The ramet number and BG biomass of A. philoxeroides and the ramet number of A. sessilis were significantly affected by the BG herbivory treatment (Table 1): nematode-infested A. philoxeroides plants had more ramets and BG biomass, but nematode-infested A. sessilis plants conversely had fewer ramets, compared to conspecific plants with no nematodes (Figs. 2, 3). The root length of A. philoxeroides and the AG biomass and total biomass of A. sessilis were significantly affected by interactions between the AG herbivory treatment and the BG herbivory treatment (Table 1). The roots of A. philoxeroides were significantly longer under herbivory by the nematode alone compared to other AG-BG herbivory treatments (Fig. 2D). However, both the AG biomass and total biomass of A. sessilis were significantly reduced by all the AG-BG herbivory treatments compared to the control plants with no herbivory (Fig. 3).

Table 1 ANOVA results of effects of aboveground herbivory treatment (AG) and belowground herbivory treatment (BG) on the growth traits and the relative neighbour effect (RNE) indexes of Alternanthera philoxeroides and Alternanthera sessilis under intraspecific competition. Significant results with P < 0.05 are shown in bold
Full size table
Fig. 2
figure 2

Effects of aboveground (AG) and belowground (BG) herbivory treatments on the stem diameter (A), ramet number (B), stolon length (C), and root length (D) of Alternanthera philoxeroides and Alternanthera sessilis under intra- or interspecific competition. No nema = no nematodes; Nema = herbivory by nematodes. Different capital letters and lowercase letters indicate significant difference of the growth traits of A. philoxeroides and A. sessilis, respectively, among AG- and BG herbivory treatments at P < 0.05 level

Full size image
Fig. 3
figure 3

Effects of aboveground (AG) and belowground (BG) herbivory treatment on the biomass of Alternanthera philoxeroides and Alternanthera sessilis under intra- or interspecific competition. No nema = no nematodes; Nema = herbivory by nematodes. Different capital letters and lowercase letters indicate significant difference of the AG biomass and BG biomass of A. philoxeroides and A. sessilis, respectively, under AG- and BG herbivory treatments at P < 0.05 level

Full size image

When the two plant species were grown together, only the stem diameter and ramet number of A. philoxeroides were significantly affected by interactions between the AG herbivory treatment and the BG herbivory treatment (Table 2). Agasicles hygrophila herbivory alone significantly reduced the stem diameter, but the combinative herbivory by A. hygrophila and the nematode conversely increased the ramet number by 180% (Figs. 2B, C). The AG biomass and total biomass of A. philoxeroides were significantly affected by the AG herbivory treatment (Table 2). Both of these parameters were significantly reduced by A. hygrophila, compared to other AG herbivory treatment (Fig. 3). In contrast, none of the growth traits of A. sessilis were significantly affected by AG-BG herbivory treatments, except for the stolon length (Table 2), which was significantly reduced by C. piperata herbivory alone or the combinative herbivory by A. hygrophila and the nematode, compared to the control plants with no herbivory (Fig. 2C).

Table 2 ANOVA results of effects of aboveground herbivory treatment (AG) and belowground herbivory treatment (BG) on the growth traits and the relative neighbour effect (RNE) indexes of Alternanthera philoxeroides and Alternanthera sessilis under interspecific competition. Significant results with P < 0.05 are shown in bold
Full size table

The intra- and interspecific interactions between the two plants

For plants under intraspecific competition conditions, the RNE index values of A. philoxeroide were only significantly affected by the AG herbivory treatment, while the RNE index values of A. sessilis were significantly affected by interactions between the AG herbivory treatment and the BG herbivory treatment (Table 1). Specifically, neighboring conspecifics had weakly competitive effects (positive RNE index values) on A. philoxeroide but significant facilitative effects (negative RNE index values) on A. sessilis under intraspecific competition conditions with no herbivory (Fig. 4). Significant competitive effects from neighboring conspecifics were detected for A. philoxeroides under herbivory by A. hygrophila alone (Fig. 4). However, strong shifts from facilitative processes to significant competitive effects from conspecific neighbors were found for A. sessilis under attack by all the AG and/or BG herbivory treatments (Fig. 4).

Fig. 4
figure 4

Effects of aboveground (AG) and belowground (BG) herbivory treatment on the relative neighbour effect (RNE) values between Alternanthera philoxeroides and Alternanthera sessilis under intra- or interspecific competition. No nema = no nematodes; Nema = herbivory by nematodes. Asterisks indicate a significant positive or negative RNE index values (i.e., greater than one minimum significant difference from zero: Tukey HSD test, P < 0.05)

Full size image

In contrast, for plants under interspecific competition conditions, the RNE index values of A. philoxeroide were significantly affected by interactions between the AG herbivory treatment and the BG herbivory treatment, while the RNE index values of A. sessilis were only significantly affected by the AG herbivory treatment (Table 2). Specifically, interspecific relationships between A. philoxeroide and A. sessilis were facilitative when there was no herbivory, and the facilitative effect of A. sessilis on A. philoxeroides became stronger under combinative herbivory by C. piperata and the nematode (Fig. 4). However, no apparent competitive or facilitative effects from A. philoxeroides were detected for A. sessilis under all the AG-BG herbivory treatments (Fig. 4).

Discussion

An important factor affecting invasion success of plants is the species’ exact biology, including clonal traits (Cornelissen et al. 2014; Pan and Price 2002). It is well known that many important plant invaders, such as A. philoxeroides, grow and spread primarily or exclusively by vegetative growth and clonal propagation in their introduced regions (Liu et al. 2006; Wang et al. 2017). Stronger clonal potential of invasive clonal plants compared to their co-occurring, non-invasive clonal native plants may facilitate the establishment of invasive species in recipient habitats (Wang et al. 2017; You et al. 2016). Alternanthera philoxeroides generally had longer stolons and roots. Longer stolon length indicates a stronger capacity for dispersal and occupation in stoloniferous clonal plants (Oborny and Kun 2002). Wang et al. (2016) found that the cost of stolon elongation for A. philoxeroides was lower than that for A. sessilis, which may let A. philoxeroides more rapidly colonize novel environments. We also found that A. philoxeroides had higher BG biomass in comparison to A. sessilis, regardless of competition type. Higher total biomass reflects the better performance for A. philoxeroides, and its longer roots and higher root biomass may enable A. philoxeroides to use soil water and nutrients more efficiently and produce more storage and propagative organs compared to A. sessilis (Poorter and Nagel 2000; Wang et al. 2016). Some other studies have also demonstrated that many successful plant invaders likely benefit more from biomass allocation to BG parts compared to co-occurring native plants (Drenovsky et al. 2012; Ren et al. 2019).

Another important invasion mechanism of successful invasive plants is phenotypic plasticity (Davidson et al. 2011), which has been considered as a critical factor contributing to the invasiveness of A. philoxeroides, which occupies variable environments (Geng et al. 2006; Zhang et al. 2021). In this study, we found that A. philoxeroides produced a greater number of ramets, longer roots, and more biomass under AG and/or BG herbivory. The type of herbivory (A. hygrophila, C. piperata, or joint herbivory with the nematode) was found to affect the level of phenotypic plasticity. However, its native congener A. sessilis was greatly suppressed by herbivory under most conditions. Ramet number is a functional trait that is positively correlated with reproductive allocation because having more ramets can help plants occupy more space, leading to higher reproductive output. Increases in ramet number may lead clonal plants like A. philoxeroides to produce more reproductive parts to cope with biomass loss caused by AG-BG herbivores. More AG and/or BG biomass after herbivory indicates A. philoxeroides has both a higher compensatory growth ability and higher tolerance to herbivory (He et al. 2014; Lu and Ding 2012; Lu et al. 2014). Adaptive phenotypic plasticity may promote the optimal trait expression of invasive species and potentially facilitate successful invasion in a broad range of environments such as those that have multiple herbivores.

The invasive success of alien plants may reflect an evolutionary response to changes in natural enemy pressure in their invaded ranges (Müller-Schärer et al. 2004). Joshi and Vrieling (2005) proposed that invasive species will become better protected against polyphagous enemies but concurrently lose their defenses against monophagous or oligophagous enemies that are absent in the invaded range. Thus, coevolved specialists will have a greater impact on invasive plants, while indigenous generalist herbivores will have a greater impact on native plants in the introduced range (Joshi and Vrieling 2005; Müller-Schärer et al. 2004). Consistent with the above predictions, we also found that the specialist A. hygrophila reduced the growth of A. philoxeroides more seriously than did the native C. piperata, indicating that under the same population density, the co-evolved specialist beetle performed better on A. philoxeroides than those oligophagous herbivores from introduced ranges. Similarly, A. hygrophila could suppress the leaf biomass of A. philoxeroides more strongly than could two generalist insects (Atractomorpha sinensis and Hymenia recurvalis) (Yu and Fan 2018), while herbivory by C. piperata could not suppress the growth and expansion of A. philoxeroides (Wei et al. 2016). These findings support the hypothesis that release from co-evolved specialists (from the native range) may contribute to better performance of invasive plants in introduced regions, as proposed by the ERH (Keane and Crawley 2002).

Little is known about the effects of herbivory on the competitive ability of exotic plants (Huang et al. 2012a; Mangla et al. 2011). Stoloniferous clonal plants usually grow in diameter rather than height, and therefore intraspecific competition within clonal plants is likely to be very strong given that many invasive clonal plants form thick monospecific stands in the invade range (Wan et al. 2019). Yu et al. (2019) suggested that intraspecific competition can limit the growth and establishment of A. philoxeroides by suppressing its root sprouting and ramet growth. When resources are limited due to intraspecific competition, population growth of clonal plants may be easily affected by environmental disturbances such as herbivory (Huang et al. 2012a; Kim et al. 2013), which can influence the fitness and the ability of plants to compete with neighbors (Center 2005). In this study, we found AG-BG herbivory exerted different effects on intraspecific competition of A. philoxeroides and A. sessilis. Notably, negative effects from neighboring plants on A. philoxeroides were detected under herbivory by A. hygrophila. Thus, A. hygrophila could potentially prevent A. philoxeroides from aggressively expanding under competitive conditions. If so, native species such as A. sessilis would benefit from the feeding of A. hygrophila on the dominant A. philoxeroides. Similarly, following successful control of invasive leafy spurge (Euphorbia esula L.) by flea beetles (Aphthona spp.), several sod-forming grasses become dominant in northern Great Plains grasslands (Larson and Larson 2010). Moreover, intraspecific competition has the potential to generate feedbacks between plants and herbivores as intraspecific competition would result in a decrease in plant size and plant nutritional quality (Louda et al. 1990).

Interspecific competition and herbivory can both lead to reductions in plant growth, biomass, and reproduction (Adler et al. 2018; Ibanez et al. 2013; Jing et al. 2015). Herbivore damage may change the outcome of the interspecific competition based on the herbivores’ feeding preferences (Hambäck and Beckerman 2003; Ibanez et al. 2013; Jing et al. 2015). In this study, we found A. philoxeroides and A. sessilis differed in overall performance and their responses to interspecific competition and either AG and/or BG herbivory. The interspecific interactions between A. philoxeroides and A. sessilis was facilitative when there was no herbivory, suggesting that both A. philoxeroides and A. sessilis benefit from their interspecific competition. Many studies have also demonstrated that herbivory can affect the intensity of interspecific competition in plants (Center 2005; Huang et al. 2012a; Wei et al. 2016). In this study, we found that herbivory by A. hygrophila or A. hygrophila + nematodes tended to increase the intensity of interspecific competition between A. philoxeroides and A. sessilis. These findings suggest that we may be able to manage a specific invader by increasing interspecific competition intensity combined with manipulating of herbivory. Contrary to classical competition theory (Adler et al. 2018), interspecific competition should be more intense between native and alien plants occupying similar ecological habits or niches (Sheppard and Burns 2014). Indeed, A. philoxeroides always forms denser populations and grows larger when it occurs in sympatry with A. sessilis in the field. The high densities of invasive plants may monopolize resources and inhibit the growth of other plants (Silveira and Thiébaut 2020), and differences in aggressiveness and body size may lead to asymmetric competition between the two plants.

In contrast to competitive interactions among individuals, intra- or interspecific facilitation may enhance plant tolerances to herbivores through compensatory growth (Arsenault and Owen-Smith 2002; Rand 2004). In this study, the strong facilitative effect of A. sessilis on A. philoxeroides was detected under combinative herbivory by C. piperata and the nematode. However, the negative effects of the root-knot nematode and C. piperata appeared to be additive on some growth traits of A. philoxeroides grown in monoculture. This suggest that those indigenous natural enemies or their interactions, in some cases, may also be able to influence the growth of invasive plants like A. philoxeroides. This outcome, however, is highly dependent on the identity of neighboring plants. Similar to our results, Lu and Ding (2010) found that intraspecific interactions can increase the compensatory ability of A. philoxeroides to AG herbivory. Thus, combinative herbivory by indigenous AG-BG herbivores in the invaded range (which often occurs in the field), may potentially influence the further spread of some invasive plants like A. philoxeroides. Backmann et al. (2019) also found that timely elimination of herbivores can reduce competition from neighboring plants, and that herbivores may be a part of a plant’s strategy for reducing competition and increasing plant fitness. However, similar facilitative effects from A. philoxeroides were not detected for A. sessilis, suggesting that the competitive outcomes between invasive alien plants and their co-occurring native plants are highly dependent on AG and BG herbivores (He et al. 2014; Jing et al. 2015).

Conclusion

The results of our study indicate that the intra- or interspecific interactions of the invasive A. philoxeroides and its native congener A. sessilis were affected differently by AG-BG herbivore interactions, and significant competitive effects from conspecific plants were detected for A. philoxeroides under herbivory by the monophagous A. hygrophila alone. However, A. sessilis exerted strong facilitative effects on A. philoxeroides under combinative herbivory by the oligophagous C. piperata and the nematode. Our results suggest that intra- and interspecific competition of native and invasive plants can be greatly affected by AG-BG herbivory, and thus the interactive effects between AG-BG herbivory and plant competition will have the potential to greatly influence the establishment and expansion of invasive plants in the field.