Invasive herbaceous respond more negatively to elevated ozone concentration than native species

Many studies show that increase in ground‐level ozone (O3) has adverse effects on plant growth. Due to high phenotypic plasticity, invasive species is considered to be more adaptable to elevated O3 than native species. This idea is only tested by the very limited studies comparing invasive weeds with crops. However, whether it holds remains unclear when comparing invasive species with their co‐occurring native species in natural systems.

to reach 85 ppb by 2100 (IPCC, 2014). Consequently, research on the effects of O 3 pollution on ecosystems has garnered considerable interest in ecology.

Adverse effects of tropospheric O 3 on plants have been broadly
documented Burkey et al., 2020;Grulke & Heath, 2020;Sicard et al., 2017). As plant leaves are most exposed to elevated O 3 , they become the main places of phytotoxic effects of O 3 on plant. Ozone can enter into plant via the stomata and then react with unsaturated biomolecules to form reactive oxygen species initially causing programmed cell death and visible injury (Vainonen & Kangasjärvi, 2015). Such visible foliar O 3 injury is an unequivocal sign of phytotoxic O 3 levels (Paoletti et al., 2019), and is the only indicator of adverse effects of O 3 that can be used for routine field surveys (Sicard et al., 2016). Such injuries often appear as chlorosis, necrosis, spots and bronzing . Some studies suggested that visible foliar injury could be associated with not only plant susceptibility (Bergmann et al., 2017;Davison & Barnes, 1998;Li et al., 2016) but also negative impacts on fitness traits (Marzuoli et al., 2019). In agricultural ecosystems, the influence of elevated O 3 on crop species has received much attention and is well understood, especially the effect on yield reduction (Ainsworth, 2017;Tai et al., 2014;Wilkinson et al., 2012). However, few studies have focused on studying the potential influence of O 3 on other important ecological processes that occur frequently in non-management regions, such as plant invasion.
Plant invasions, similar to O 3 pollution and many other components of global change (Steffen et al., 2011), have started to rapidly increase from the last century, and the number of naturalized alien species is still increasing (Seebens et al., 2017(Seebens et al., ,2020. It is frequently assumed that plant invasion could interact with, and thus be affected by other components of environmental changes (Bradley et al., 2010;Dukes & Mooney, 1999;Liu et al., 2018). Indeed, a recent meta-analysis by Liu et al. (2017) showed that CO 2 enrichment and global warming increase the growth more strongly for invasive alien plants than for native species, and thus might promote plant invasion. On the other hand, many non-invasive alien plant species under current environmental conditions may become invasive with ongoing global changes Haeuser et al., 2017;Speißer et al., 2021;Walther et al., 2009). Therefore, given that the O 3 pollution varies across large scale (e.g. O 3 pollution in China; Figure S1), understanding its potential effects on plant invasion is important for the strategic management of increasing risks of such plant invasions in future.
Due to high phenotypic plasticity, invasive plants often exhibit broad environmental adaptation, and thus often show high fitness in altered environmental conditions (Davidson et al., 2011;Richards et al., 2006). Consequently, invasive species might be more adaptable to elevated O 3 than native species. Indeed, the very limited studies established in cropping systems showed that problematic weeds (i.e.

Amaranthus tuberculatus, Amaranthus palmeri, and Cyperus esculentus)
are more tolerant to elevated O 3 or can even benefit from elevated O 3 compared with co-occurring crop species. This suggests that increasing O 3 pollution can exacerbate these invasions in agricultural cropping systems (Grantz et al., 2019;Grantz & Shrestha, 2006;Paudel et al., 2016;Shrestha & Grantz, 2005). Nevertheless, based on the weed-crop experiments limited to agricultural systems, it is still difficult to determine the general pattern of elevated O 3 effects on alien plant invasion because agricultural systems are not natural plant communities and crop species are usually selected by high productivity. Moreover, some studies have found that O 3 has a greater effect on wild plants than on cultivated plants (Biswas et al., 2008).
Thus, to rigorously test the responses of invasive and native species to elevated O 3 , invasive species must be compared with their closely related and co-occurring native species from non-management systems. However, to the best of our knowledge, no such studies have been conducted.
Competition is considered one of the most important processes that determine the likelihood of alien plant invasion (Baker, 1965;Bishop & Cook, 1981;Roy, 1990). Competition can often interact with other abiotic factors (Larson et al., 2018;Low-Décarie et al., 2011) and is affected by elevated O 3 concentration (McDonald et al., 2002;Scebba et al., 2006). Moreover, invasive weeds in weedcrop experiments are more tolerant to O 3 , thereby resulting in greater competitive success of specific crop species (Grantz et al., ,2008(Grantz et al., ,2019Paudel et al., 2016;Shrestha & Grantz, 2005). Therefore, it is important to consider the potential interaction between competition and O 3 when assessing plant invasiveness.
In this study, we performed an open-top chamber (OTC) experiment with multiple species to test whether responses to elevated O 3 differ between invasive and native species. This is of particular interest to understand whether O 3 pollution can promote or suppress plant invasion. As the degree of susceptibility to O 3 stress can differ among species (Van Goethem et al., 2013), we selected six congeneric pairs of invasive and native species and grew them with and without competition under ambient and elevated O 3 conditions. We measured the aboveground biomass and the number of leaves damaged by O 3 , and addressed (1) whether invasive species are more tolerant to O 3 than their co-occurring native species and (2) whether there is an interactive effect between O 3 and plant competition, and if so, whether it differs between invasive and native species.

| Study species and cultivation
To test for differences in plant responses to elevated O 3 between invasive and native species with and without competition, we selected 12 herbaceous species. To control for phylogenetic relatedness, we used congeneric pairs, that is six congeneric pairs of invasive and native species (Table S1). In each pair, one species is alien invasive in China, and the other is native co-occurring with the invader in natural habitats. We classified the species as invasive or native to China based on the information available in the databases of Invasive Alien Species in China (www.china ias.cn) and Flora of China (www.eflor as.org). The plant materials used in this experiment were derived from ramets that were collected in natural habitats (Dai et al., 2016;Wang et al., 2017) or acquired from commercial seed companies (Table S1).
From 13 to 26 May 2020, we used seeds or stolon/rhizome fragments from maternal plants to obtain seedlings/plantlets for the experiment. For species with clonal reproduction, we selected healthy, strong stolons (for the stoloniferous species) and rhizomes

| Experimental setup
To assess the performance of invasive and native species in response to elevated O 3 concentration, we conducted an OTC experi-  Figure S2b).  (Table S2).

| Measurements
We concluded the experiment on 6 August 2020, 43 days after the

| Statistical analyses
To test for differences in biomass production between invasive and native species in response to elevated O 3 , we fitted a linear mixed-effects model in R 3.0.2 (R Core Team, 2013) using the lme function in the nlme package (Pinheiro et al., 2013). As the aboveground biomass analysis model had a Gaussian error distribution, we square-root-transformed the data prior to analysis to improve the normality and homogeneity of the residuals. The fixed part of the model included status (invasive vs. native), O 3 (elevated vs. ambient O 3 concentration), competition (with vs. without) and all their interactions. To account for heterogeneity of variance (i.e. differences in variance between the different species), the model also included a variance structure using the varIdent function in the nlme package, which allowed each level of the factor "species" to have a different variance. To test for differences in the number of total leaves and the number of damaged leaves between the invasive and native species under elevated O 3 treatment, we fitted generalized linear mixed-effects models using the glmer function in the lme4 package (Bates et al., 2015). For both models with negative binomial error terms, we used the bobyqa optimizer with a maximum of 10,000 iterations to fit these models. The fixed part of the model included status (invasive vs. native), competition (with vs. without) and their two-way interaction. To account for non-independence of plants from the same species, non-independence of species from the same genus, and non-independence of plants from the same OTC, all three models included genus, species (nested within genus) and OTC as random effects. In the (generalized) linear mixed-effect models described above, we assessed the significance of the main fixed-effect terms (i.e. status, competition and O 3 ) and their interactions using likelihood-ratio tests (Zuur et al., 2009).

| RE SULTS
Elevated O 3 concentration had a significant negative effect on aboveground biomass (6.9 g vs. 5.1 g; Table 1; Figure 1). In particular, elevated O 3 treatment reduced the aboveground biomass of invasive species (−33.2%) significantly more than that of native species (−17.6%) (Significant interaction of status × O 3 , Table 1;

| DISCUSS ION
In this study, we found although there was no significant differ on the production of aboveground biomass between invasive and native species, invasive species reduced significantly more aboveground biomass than native species in response to elevated O 3 .  Figure S3). Our study provides more evidence to support the previous findings that exposure to elevated O 3 can result in suppressed photosynthesis, accelerated senescence and decreased growth in plants (Ashmore, 2005;Burkey et al., 2020;Grulke & Heath, 2020). However, in weed-crop study systems, it was found that invasive weeds often suffer less from O 3 stress than specific crop species (Grantz & Shrestha, 2006;Shrestha & Grantz, 2005). Contrarily, we found that O 3 stress reduced the aboveground biomass of invasive species significantly more than that of native species Moreover, similar pattern is found in most native-invasive species pairs ( Figure S4). This indicates that in general, elevated O 3 has a stronger adverse effect on invasive species than on native species.

Moreover
Ozone can damage plant leaves, and visible foliar injury is an accurate indicator of plant susceptibility to O 3 (Li et al., 2016). We compared the numbers of total leaves and O 3 -induced damaged leaves between invasive and native plants and found that invasive plants had significantly more damaged leaves than native plants, whereas the number of total leaves did not differ. Moreover, similar pattern is found in most native-invasive species pairs ( Figure S5). That indicates in general invasive species had higher foliar injury rates under O 3 stress. Given that the degree of foliar injury is closely related with photosynthesis (Marzuoli et al., 2019;Vainonen & Kangasjärvi, 2015), the most likely explanation for the higher growth suppression of invasive species than of native species is the greater photosynthesis suppression due to higher foliar injury in invasive species.
However, some studies on invasive weeds versus crops in agricultural systems found that invasive species are more tolerant to elevated O 3 concentration than crops (Grantz et al., 2008(Grantz et al., ,2019Grantz & Shrestha, 2006;Paudel et al., 2016;Shrestha & Grantz, 2005). Therefore, our results highlight that the effects of elevated O 3 concentration on alien plant invasions may differ between natural and agricultural systems. A plausible explanation is that crop species are not really equivalent to native species from non-management systems. The strong selection for crop species to grow fast and be  (Cooper et al., 2014). The populations of some species have been demonstrated to differ in O 3 sensitivity/tolerance, and these differences were statistically related primarily to the "O 3 climate" of their site of origin (Lyons et al., 1997;Pearson et al., 1996;Reiling & Davison, 1992). Therefore, more studies are needed to test the evolution of O 3 sensitivity/tolerance of invasive species using different populations originating from varied O 3 climates.
In some weed-crop studies, there was a significant interactive effect between O 3 and competition on plant performance. In general, invasive weeds are more tolerant to O 3 , which results in greater competitive ability against crops (Grantz et al., 2008(Grantz et al., ,2019Grantz & Shrestha, 2006;Paudel et al., 2016;Shrestha & Grantz, 2005).
However, there was no significant interactive effect between O 3 and competition on species biomass production and the degree of foliar injury in this study. The situation in which no substantial competition occurs between species due to the use of a single competitor or the relatively short duration of the experiment can also lead to failure in detecting the interaction between competition and O 3 . Nevertheless, this could not be the case in this study. Although the overall competition effect was not significant, a separate analysis of the data subset excluding elevated O 3 treatment revealed that competition had a significant effect on aboveground biomass (p = .003; Figure S6). Thus, competition could decrease plant biomass production under ambient O 3 conditions. However, there was a marginally significant effect of the interaction between species status and O 3 on the competition outcome (p = .077; Figure S7), thereby inferring a mild interaction be-

F I G U R E 2
Mean values of the total leaves (a) and damaged leaves (b) averaged across six invasive and six native species grown without (Alone) and with competition (Comp.) under elevated O 3 condition. Error bars represent standard errors ground-level O 3 concentration is often higher for Northern China than for Southern China ( Figure S1). In other words, invasive species might be more competitive relative to their co-occurring native species when they invade to the area with lower O 3 pollution. Second, it indicates that species distribution models might also consider the role of ground-level O 3 concentrations to improve its accuracy when predicting future distributions of invasive species.
In conclusion, we found that the biomass reduction and leaf

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest.

PE E R R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1111/ddi.13452.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data available from the Dryad Digital Repository https://doi. org/10.5061/dryad.vt4b8 gtt5 (Wang et al., 2021).