Competition on a neutral playing field: invaders still win and size still matters… sometimes

Abstract

Exotic invasive plant species commonly outcompete native species. However, a great deal of the evidence for this comes from experiments conducted on an uneven playing field—in substrates containing soil biota from the non-native ranges of the exotics, which should give them a competitive advantage. In competition experiments with nine pairs of non-native invasive versus native species in neutral substrates composed of sterilized soil, we found that the competitive effect of invasive species on natives was approximately five times greater than the reverse, and gram-per-gram competitive effects of invasives on natives were almost two times that of the natives on invasives. The effect of plant size on competitive outcomes was complex. The size of invasive species correlated with their effects on natives but not with their tolerance to competition from natives. The size of natives was not correlated with either aspect of competitive ability. This is important since the tolerance of invaders to competition from natives is thought to be essential for successful invasion. Our results also suggest that assumptions about size-based evidence for the evolution of competitive ability in non-native ranges is reasonable, and that even without the advantage gained from escaping soil biota, invaders still win.

1. Introduction

Competition is an important part of exotic plant invasions, both in the context of native resistance [1–3] and the suppression of native species by exotics [4,5]. Exotic species, and especially those identified as invasive, are generally better competitors than native species in experimental matchups [6–8]. However, competitive outcomes are strongly dependent on environmental contexts [9–11] and the species identity of the matchups [7,12], leaving knowledge gaps in our understanding of exotic–native competition.

First and foremost, competitive outcomes depend on many different environmental contexts [9–11]. For example, if competition experiments are conducted in the field in the non-native ranges of exotic species, or in soils from the non-native ranges, exotic species may have a substantial advantage because they may have escaped from soil biota [13–15] and native herbivores [16], whereas the natives have not, potentially confounding the effects of competition. This could challenge our interpretations of reviews of competition between native and exotic species. For example, of the 35 publications we could access from Vila & Weiner’s classic review [6], 21 studies were conducted in the field in non-native ranges of exotic species. None were conducted in the native ranges of exotic species. Furthermore, of the 14 studies conducted under controlled conditions, 11 used live soil from the non-native ranges of the exotics, only 2 used neutral soils (sand or sterilized) and 1 did not report the source of the substrate used. Clearly, the majority of these studies did not separate competition between natives and exotics from the potential indirect effects of soil biota. This is not trivial, since soil biota can dramatically alter competitive outcomes [17], and because exotic plant species tend to experience much weaker inhibitory effects from soil biota than natives [13]. This suggests a knowledge gap in our understanding of direct competitive outcomes among native and exotic species.

The importance of this uncertainty emerges in a second knowledge gap, a substantial one, which is the need to understand how the size of invasive plant species determines their competitive impact on natives. Most common-garden experiments on the evolution of increased competitive ability (EICA) assume that the oft-found larger sizes of plants sourced from non-native ranges compared with conspecifics sourced from native ranges [18–20], bestows greater competitive ability. For example, in a recent review of EICA, data were available for size comparisons for 87 species but only 16 for actual competitive interactions [20]. Size should confer competitive advantages. In pairwise trials among three species, species that were larger when grown alone, generally exerted greater competitive effects and generated stronger competitive responses [21]. Other studies have shown that larger plants usually have an advantage relative to neighbouring smaller plants when competing together in experimental mixtures [22,23]. However, if a small individual in competition for resources with larger neighbours has greater rates of per-unit resource uptake (i.e. is a better per-gram competitor), it might have an advantage over the long term because it would acquire more resources relative to its size [24]. Thus, comparing these size-independent, per-unit competitive effects may lead to insights into direct competitive outcomes among species [7,12]. However, few studies compare the per-unit competitive effects among interacting exotics and natives.

A third knowledge gap in our understanding of competition among native and exotic species involves competitive ‘effects’ and ‘responses’ [22,25,26]. Both are aspects of competitive intensity, but competitive effects represent the degree to which a species suppresses its neighbours, whereas competitive responses represent the degree to which a species tolerates competition from its neighbours. Competitive effects and responses are separate components of competitive ability and may operate in different ecological and evolutionary ways [26–28]. For example, in pairwise competition, effects and responses appear to be of similar importance to overall competitive outcomes. In contrast, when pairwise experiments are integrated into models with multiple competitors interacting simultaneously, competitive responses (the ability to tolerate neighbours) appear to be more important than competitive effects for the ability of native species to resist invasion [26]. This was because the benefits of having a strong ability to suppress exotics are shared with all other natives in communities, regardless of their contribution to suppressing exotics. But native species, or genotypes, that are good at tolerating competitors (high responses) do not have to share this ability with other species. And on the flip side of this, a meta-analysis found a similar scenario, but for non-native species, which were more likely to outperform native species through competitive tolerance (response) of competition rather than competitive suppression (effect) [28]. Therefore, a better understanding of why and how invasive species outcompete natives is important; however, we have few explicit comparisons of the relative importance of these competitive responses and effects [20].

In a greenhouse experiment, we explored the competitive ability of nine pairs of coexisting invasive and native species from different sites across the western US, using neutral (sterilized) soil. We addressed the following questions: (i) do invasive species have stronger inherent competitive abilities than co-occurring native species in neutral soil? (ii) Does size confer stronger competitive effects and responses to invasive species? (iii) Do competitive effect, competitive response and plant size interact to confer greater competitive advantages to invasive species in neutral soil?

2. Material and methods

(a). Study site and species

We used nine pairs of coexisting native and invasive species that commonly occur in grasslands in the western US (electronic supplementary material, table S1). Our goal was to pair plants with similar growth forms from the same family. In the autumn of 2013, we collected seeds from one invasive species and one coexisting native species at nine invaded grasslands across the western US, all of which showed no evidence of recent human disturbance (electronic supplementary material, table S1). All native species were remnants, or survivors of invasion by the more abundant invasive species they were paired with. Different pairs of species were collected at each site, and we sought sites with low-to-moderate proportions of the invasive species in the communities (electronic supplementary material, table S1). For each species, we collected seeds from at least 30 maternal plants and stored the seeds in a refrigerator prior to the greenhouse experiments. We also collected the soil from the native species population that had not yet been invaded adjacent to the site of seed collection. At each site we collected a pooled soil sample that was a mixture of 10−20 samples taken across a site, at a depth of 5−20 cm. In total, we collected 8−10 l of soil from each site. Soil was placed in zip-lock bags and kept in a 4°C refrigerator prior to use. We used this natural soil because the species collected at a particular site might preferentially occur on soils with particular textures, mineral contents or nutrients, but we sterilized soils to eliminate biota.

(b). Experimental design

In October 2013, soils were sterilized by autoclaving once per day on 3 successive days. Seeds of pairs of native and invasive plant species, that co-occurred in nature, were germinated in separate Petri dishes containing sterilized sand and were kept moist with distilled water. Two weeks after germination, seedlings were transplanted into 250 ml pots, filled with sterilized soil that originated from the same site where the seeds were collected. The growth substrate was a 60:40 mixture of sterilized soil and sand. We used 250 ml pots, which were sufficient to study early establishment and competition. Each pair of co-existing native and invasive plant species was grown either together in the same pot (in competition) or alone (without competition). In the competition treatment, one individual of the invasive species and one of the native species were planted in the same pot. Without competition, only a single native or one invasive plant grew per pot. For each treatment combination (9 sites × 3 planting treatments, alone and competition), n = 10 resulting in a total n of 270 pots for the experiment. All the pots were randomly placed in a greenhouse of the University of Montana, watered daily and grown for 120 days. To ensure that space did not limit growth or competition, the experiment was terminated once the plants' root systems filled the pots. Plants were harvested, dried at 60°C for 3 days and their total biomass was weighed. We calculated the competitive effects of invasive species on native species and competitive responses of invasive species to native species using the relative interaction intensity (RII) index [29]

$${\displaystyle \mathrm{R}\mathrm{I}{\mathrm{I}}_{\mathrm{e}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t},\mathrm{i}}=\frac{\mathrm{B}{\mathrm{w}}_{\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e},\mathrm{i}}-\overline{\mathrm{B}{\mathrm{o}}_{\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}}}}{\mathrm{B}{\mathrm{w}}_{\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e},\mathrm{i}}+\overline{\phantom{20}\mathrm{B}{\mathrm{o}}_{\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}}}}}$$ (2.1)

$${\displaystyle \mathrm{R}\mathrm{I}{\mathrm{I}}_{\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{e},\mathrm{i}}=\frac{{\displaystyle \mathrm{B}{\mathrm{w}}_{\mathrm{i}\mathrm{n}\mathrm{v}\mathrm{a}\mathrm{s}\mathrm{i}\mathrm{v}\mathrm{e},\mathrm{i}}-\overline{\mathrm{B}{\mathrm{o}}_{\mathrm{i}\mathrm{n}\mathrm{v}\mathrm{a}\mathrm{s}\mathrm{i}\mathrm{v}\mathrm{e}}}}}{{\displaystyle \mathrm{B}{\mathrm{w}}_{\mathrm{i}\mathrm{n}\mathrm{v}\mathrm{a}\mathrm{s}\mathrm{i}\mathrm{v}\mathrm{e},\mathrm{i}}+\overline{\mathrm{B}{\mathrm{o}}_{\mathrm{i}\mathrm{n}\mathrm{v}\mathrm{a}\mathrm{s}\mathrm{i}\mathrm{v}\mathrm{e}}}}}}$$ (2.2)

where RII_effect,i and RII_response,i are the competitive effect and response of the i^th individual of an invasive species growing with a neighbour, respectively; ${\displaystyle \mathrm{B}{\mathrm{w}}_{\mathrm{i}\mathrm{n}\mathrm{v}\mathrm{a}\mathrm{s}\mathrm{i}\mathrm{v}\mathrm{e},\mathrm{i}}}$ and ${\displaystyle \mathrm{B}{\mathrm{w}}_{\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e},\mathrm{i}}}$ are the total biomasses of the i^th individual of an invasive species and its paired native competitor when grown together, respectively; ${\displaystyle \overline{\mathrm{B}{\mathrm{o}}_{\mathrm{i}\mathrm{n}\mathrm{v}\mathrm{a}\mathrm{s}\mathrm{i}\mathrm{v}\mathrm{e}}}}$ and ${\displaystyle \overline{\mathrm{B}{\mathrm{o}}_{\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}}}}$ are the mean total biomass of the focal invasive and native species when grown alone, respectively; values of this metric range from 1 to −1. A negative value indicates inhibition and a positive value indicates facilitation. We also calculated the gram-per-gram RII [7]

$${\displaystyle {\mathrm{G}\mathrm{r}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{r}\mathrm{a}\mathrm{m}\mathrm{R}\mathrm{I}\mathrm{I}}_{\mathrm{e}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t},\mathrm{i}}=\frac{{\mathrm{R}\mathrm{I}\mathrm{I}}_{\mathrm{e}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t},\mathrm{i}}}{\overline{\mathrm{B}{\mathrm{o}}_{\mathrm{i}\mathrm{n}\mathrm{v}\mathrm{a}\mathrm{s}\mathrm{i}\mathrm{v}\mathrm{e}}}}}$$ (2.3)

$${\displaystyle {\mathrm{G}\mathrm{r}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{r}\mathrm{a}\mathrm{m}\mathrm{R}\mathrm{I}\mathrm{I}}_{\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{e},\mathrm{i}}=\frac{\mathrm{R}\mathrm{I}{\mathrm{I}}_{\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{e}}}{\overline{\mathrm{B}{\mathrm{o}}_{\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}}}}}$$ (2.4)

Additionally, we quantified the competitive advantage of invasive over native species with competitive inequality in a similar manner as Golivets & Wallin [28]

$${\displaystyle \mathrm{I}\mathrm{n}\mathrm{e}\mathrm{q}\mathrm{u}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{t}{\mathrm{y}}_{\mathrm{R}\mathrm{I}\mathrm{I},\mathrm{i}}={\mathrm{R}\mathrm{I}\mathrm{I}}_{\mathrm{e}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t},\mathrm{i}}-{\mathrm{R}\mathrm{I}\mathrm{I}}_{\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{e},\mathrm{i}}}$$ (2.5)

$${\displaystyle \mathrm{I}\mathrm{n}\mathrm{e}\mathrm{q}\mathrm{u}\mathrm{a}\mathrm{l}\mathrm{i}\mathrm{t}{\mathrm{y}}_{\mathrm{G}\mathrm{r}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{r}\mathrm{a}\mathrm{m}\mathrm{R}\mathrm{I}\mathrm{I},\mathrm{i}}={\mathrm{G}\mathrm{r}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{r}\mathrm{a}\mathrm{m}\mathrm{R}\mathrm{I}\mathrm{I}}_{\mathrm{e}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{c}\mathrm{t},\mathrm{i}}-{\mathrm{G}\mathrm{r}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{r}\mathrm{a}\mathrm{m}\mathrm{R}\mathrm{I}\mathrm{I}}_{\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{o}\mathrm{n}\mathrm{s}\mathrm{e},\mathrm{i}}}$$ (2.6)

In the above equations, a negative value corresponds to competitive advantage of the invasive over the native species, whereas a positive value corresponds to competitive disadvantage for the invasive species.

(c). Statistical analysis

To explore whether invasive species have higher inherent competitive abilities than co-occurring native species, we first determined the differences from zero of competitive effects/responses using one-sample t-tests. Then, we compared the differences in the means of competitive effect and response. For all species combined, we constructed a general linear model (LM) using RII as the dependent variable with the components of competitive ability (effect versus response), species pair identity and their interaction as the independent variables. For individual invasive species, we constructed an LM model using RII as the dependent variable with the components of competitive ability as the independent variables. To explore whether size conferred stronger competitive effects and responses, we replaced RII with gram-per-gram RII in the above-mentioned LMs and reran the models to account for potential size-dependent effects.

To explore whether competitive effect, competitive response and plant size interactively conferred greater competitive advantage of invasive species, we used partial least squares (PLS) path models to detect any direct and indirect correlations between the magnitude of invader competitive advantage and plant size, as well as any direct correlations between competitive advantage and the magnitudes of competitive effect and response. Because more negative competitive effects and inequalities indicate greater magnitudes of competitive effect and advantage, we calibrated competitive effect/inequality with −1 before running the model. In this way, positive path coefficients indicate positive effects, whereas negative path coefficients indicate inhibitions. For plant size, in addition to invasive and native biomasses when grown alone, we supplemented a metric for size inequality by subtracting native biomass from invasive biomass. See the detailed PLS path model construction process in the electronic supplementary material, Note S1.

All statistics were conducted in R v.4.2.1 [30]. The R package ‘rstatix’ was used for conducting t-tests [31]. The R package ‘base’ was used for constructing the LM models [30]. The R package ‘plspm’ was used for constructing the PLS path models [32].

3. Results

(a). Competitive effects and competitive responses

As a group, the RIIs for the effects of non-native invasive species on natives were approximately five times higher than the responses of invasive species to native species (figure 1a, table 1). It is important to note that the response of invasives to natives, in our experiments, is equivalent to the effects of natives on invasives, so in other words, invasives suppressed natives much more than natives suppressed invasives. Of the nine pairs, five invasive species suppressed their native competitors more than they responded to the competitors, and in no case did a native outcompete their paired native (figure 1a, electronic supplementary material, table S2). Seven of the nine invasive species had competitive effects on their paired natives (the mean was different from zero), whereas only three invasive species were significantly affected by competition from natives (figure 1a, electronic supplementary material, table S3).

Figure 1.

Competitive effects of invasive species on natives and the competitive responses of invasive species to natives quantified by (a) unadjusted relative interaction intensity (RIIs) and (b) gram-per-gram RIIs. For all species combined, RIIs are shown as bars with stripes; for each species pair, RIIs are shown as bars with black outlines. White bars represent the competitive effect of invasive species on natives, while grey bars represent the competitive response of invasive species to natives. Error bars show 95% confidence intervals. Asterisks above brackets show significant differences for the pairs (*p < 0.050, **p < 0.010 and ***p < 0.001). Asterisks above bars show significant differences from 0. For each species pair, a more negative value of the white bar compared with the grey bar indicates a competitive advantage of invasive species over natives and vice versa. See electronic supplementary material, tables S2 and S3.

Table 1.

Effects of the components of competitive ability (effect versus response), species pair identity and their interaction on RII and gram-per-gram RII based on general linear models. Significant effects are shown in bold. d.f., degree of freedom; sum sq. sum of squares; mean sq, mean squares; LM, linear models, RII, relative interaction intensity.

dependent variable	source	d.f.	sum sq.	mean sq.	F	p	model R2
RII	component (C)	1	4.54	4.54	100.77	<0.001	R2 = 0.578, adjusted R2 = 0.532
	species pair (SP)	8	1.95	0.24	5.41	<0.001
	C × SP	8	3.25	0.41	9.00	<0.001
	residuals	158	7.12	0.05
gram-per-gram RII	c (C)	1	26.93	26.93	7.46	0.007	R2 = 0.250, adjusted R2 = 0.170
	species pair (SP)	8	50.77	6.35	1.76	0.089
	C × SP	8	112.67	14.08	3.90	<0.001
	residuals	158	569.96	3.61

(b). Gram-per-gram competitive effects and competitive response

As a group, invasives exerted stronger gram-per-gram competitive effects on native species than their response to natives (i.e. the competitive effect of natives), evaluated with our pairwise matchups (figure 1b, electronic supplementary material, table S2). Three of the nine invasive species exerted significant competitive effects on their paired native species, whereas no natives elicited a significant response in its paired invasive species (figure 1b). Gram-per-gram competitive effects had the same tendency as unadjusted RIIs, as seven of the nine non-native invasive species had significant competitive effects below zero, whereas only three invasives were affected by competition (figure 1b, electronic supplementary material, table S3).

(c). Size-dependent effects

Not accounting for size-dependent effects, the magnitudes of both competitive effect (quantified by −1 × RII_Effect) and response (quantified by RII_Response) were positively correlated with the overall competitive advantage of invasive species (quantified by −1 × competitive inequality) (figure 2a, electronic supplementary material, table S4). Competitive effect contributed slightly more to the overall competitive advantage than response (figure 2a). Plant size (quantified by invader size and invasive–native size inequality) was highly positively correlated with the overall competitive advantage of invasive species both through a direct correlation with competitive advantage and an indirect correlation with competitive effect (figure 2a, electronic supplementary material, table S4). After accounting for size dependencies of competitive effect and competitive advantage, the correlation between the strength of competitive effect/response and the strength of overall competitive advantage remained positive, but the relative contribution of competitive response surpassed that of competitive effect (figure 2b). The correlation between plant size and the strength of competitive effects became much weaker and changed from positive to negative (figure 2b).

Figure 2.

Predicted partial least squares path model for the direct and indirect correlations between (a) competitive advantage and plant size/competitive ability and (b) per unit competitive advantage and plant size/per unit competitive ability. Circles represent latent variables. Indicators for corresponding latent variables are shown in rectangles beside the circles. R² values with s.e. are shown in the circles for all endogenous latent variables (i.e. the ones that are pointed to by at least one arrow). Path coefficients are shown for each path. Goodness-of-fit for each path model is shown at the upper right corner of the corresponding panel. Blue lines and arrows denote positive correlations. Red lines and arrows denote negative correlations. Solid lines indicate significant correlations. Dashed lines indicate insignificant correlations. Because more negative values indicate greater magnitudes of competitive effect and advantages for the invasive species, we calibrated competitive effect/inequality with −1 before running the model. See indicator contributions to latent variables in electronic supplementary material, table S4.

4. Discussion

One of the key results of our experiment is that, in neutral substrates, the RIIs for the effects (suppression) of non-native invasive species on native species were over five times higher than the reverse. This mean effect was generally consistent across most species pairs, and there were no pairs in which the native species outcompeted the invasive. This overall competitive superiority was also reflected in gram-per-gram effects, which were almost twice as strong for invasives in comparison to natives. Gram-per-gram effects, which eliminated the plant-size-dependent effects, were significant and might reflect field abundance more closely than total competitive effects [24]. Gram-per-gram effects are also thought to derive from inherently proficient soil resource capture [12] through which smaller plants may efficiently deprive larger neighbours of resources, or allelopathic effects in which chemical effects are disproportionate to the size of the plant releasing them [7]. However, the relative contributions of these mechanisms are hard to quantify [33]. For example, woody plant species exerted greater total competitive effects on grasses than vice versa; however, grasses exerted 17 times stronger gram-per-gram competitive effects on woody species [12]. They attributed this to greater resource acquisition by grasses and their ability to deplete soil resources. Because our substrates were neutral, and considering gram-per-gram effects, our results indicate that, yes, per the first question, invasive species have stronger inherent competitive abilities than co-occurring native species, at least when sourced from the non-native range of invasives (see below). This is consistent with most reviews and analyses that do not account for non-neutral substrates [4,6,28].

Despite the stronger gram-per-gram effects of invasives, plant size still conferred greater competitive ability, per our second question. The size of invasive species was highly correlated with their ability to suppress native species, and competitive suppression in general has been closely associated with size-related traits [34–36]. However, the size of invasive species was not correlated with their tolerance to competition from natives, and the size of natives was not correlated with either their ability to suppress or tolerate invasives. This is interesting because some evidence suggests that the tolerance of invaders to competition from natives and the tolerance of natives to competition from exotics are key to the whole invasion process [26,28]. This is not to say competitive suppression exerted by invasives is irrelevant, but if tolerance is paramount, then size may not be as important as other less understood factors in invasion.

We are not sure why invasives exerted greater gram-per-gram effects, but this could be due to offensive allelopathic effects [37,38]. In a review, allelopathic effects were reported for 51.4% of the invasive species in the database [39]. In another review, 43% of the highly invasive non-native species they found in the literature produced at least one secondary compound not recorded for native North American species, in contrast to 11% for non-natives that were not invasive [40]. However, allelopathy could affect either competitive suppression or tolerance.

In a deeper comparison of our results to the large meta-analytics and model simulations that suggest a stronger role of competitive response than effect in the competitive advantage of invasive species [26,28], we integrated plant size, competitive effect and competitive response in models. Responses and effects were equally important in their effects on invasive–native competitive inequality. Plant size, especially the size of invasive species, may have a strong influence on invasive–native competitive inequality through direct effects and indirect influence on the competitive effect of invasive species on native species. It should be noted that our experiments did not provide the best set-up for exploring the relative importance of competitive effect and response for competitive advantage. Because competitive inequality, the metric used to quantify competitive advantage was essentially the subtraction of competitive response from effect, it is inherently correlated with at least one component of the subtraction. Nonetheless, our results suggest that the competitive suppression of invasives on natives may play a more substantial role in determining invasion success than previously found.

All nine of the exotic invasive species we tested were collected in their non-native ranges, and therefore one caveat is that we cannot evaluate whether the competitive advantages they displayed were inherent to the species or evolved after arrival in North America from Eurasia (see [20,41]). The ‘evolution of increased competitive ability’ appears to be common, or at least the evolution of increased size, however, as for our results, we do not know how much of the competitive advantage commonly reported in the literature for invasives is due to selection of certain genotypes during the invasion process or evolution once in the non-native range, which remains an important knowledge gap in general. Regardless, it appears that whatever advantages invasives gain in non-native ranges may not be expressed when reintroduced to native ranges. For example, Solidago gigantea collected in its non-native range performed better than S. gigantea collected in its native range, but only when grown in common gardens in the non-native range. When these accessions were grown in the native range, the outcome was reversed [42].

A second caveat relates to fecundity. Our experiments focused on the responses of plant total biomass to competition, which is a realistic metric. However, species can have threshold sizes for reproduction [43], and this might intrinsically (i.e. independently of competition) affect biomass production. If invasive and native species differ in these fecundity thresholds and biomass allocation, responses to competition could differ.

Another caveat for interpreting our results in the context of inherent differences, rather than the product of complex community-based and perhaps biogeographic conditionality, is that the natives in our experiments may also have recently undergone selection in response to invasions. Our natives were ‘remnants’. For example, remnant native Elymus multisetus plants surviving Bromus tectorum invasion, when transplanted to a greenhouse, were better competitors against the invader than E. multisetus plants from uninvaded areas [44]. Callaway et al. [45] extended this a generation and found that seed from several remnant species in Centaurea stoebe (a species used in our experiments) invasions produced better competitors than seed from uninvaded areas (also see [46–49]). We do not know how long our collection sites had been invaded, but the possibility exists that our experiments included this remnant effect. However, we intentionally collected from sites with low-to-moderate proportions of the invasive species (electronic supplementary material, table S1).

Is it reasonable to be concerned about comparing competitive ability among natives and invasives in non-neutral substrates and without natural enemies? Invasive species clearly benefit from lower negative effects of soil biota in their non-native ranges compared with either native species [13,50,51] or soils from their native ranges [52–57]. However, the interactions between competition and soil biota are complex. In a community-scale experiment with native species, plant size was reduced by interspecific competition, but this was not affected by plant–soil feedbacks (PSFs) [58]. Also, species were not as suppressed by competition when PSFs were generated by other species compared with when they were generated by conspecifics. Species exerting stronger competitive effects experienced more negative effects of soil biota [17]. Herbivory can also differ substantially among ranges or among natives and invasives in the non-native ranges of the latter [59–61]. All things considered; it appears that native plant species usually lose no matter what, with or without home field soil disadvantages.

Ethics

This work did not require ethical approval from a human subject or animal welfare committee.

Data accessibility

The data and statistical code for this study have been archived in the Dryad Data Repository [62].

Supplementary material is available online [63].

Declaration of AI use

We have not used AI-assisted technologies in creating this article.

Authors’ contributions

W.L.: investigation, writing — original draft, writing — review and editing; H.L.: formal analysis, writing — review and editing; R.C.: writing - review and editing; R.W.P.: data curation, investigation, writing — review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

We thank the Peoples Programme (Marie Curie Actions) of the European Union’s Seventh Framework Programme (FP7/2007-2013), REA grant number 300639 (R.W.P.), the National Natural Science Foundation of China, grant number 31000184 (W.B.L.).