Growth‐form and spatiality driving the functional difference of native and alien aquatic plants in Europe

Abstract Trait‐based approaches are widely used in community ecology and invasion biology to unravel underlying mechanisms of vegetation dynamics. Although fundamental trade‐offs between specific traits and invasibility are well described among terrestrial plants, little is known about their role and function in aquatic plant species. In this study, we examine the functional differences of aquatic alien and native plants stating that alien and native species differ in selected leaf traits. Our investigation is based on 60 taxa (21 alien and 39 native) collected from 22 freshwater units of Hungarian and Italian lowlands and highlands. Linear mixed models were used to investigate the effects of nativeness on four fundamental traits (leaf area, leaf dry matter content, specific leaf area, and leaf nitrogen content), while the influence of growth‐form, altitude, and site were employed simultaneously. We found significantly higher values of leaf areas and significantly lower values of specific leaf areas for alien species if growth‐form was included in the model as an additional predictor.We showed that the trait‐based approach of autochthony can apply to aquatic environments similar to terrestrial ones, and leaf traits have relevance in explaining aquatic plant ecology whether traits are combined with growth‐forms as a fixed factor. Our results confirm the importance of traits related to competitive ability in the process of aquatic plant invasions. Alien aquatic plants can be characterized as species producing soft leaves faster. We argue that the functional traits of alien aquatic plants are strongly growth‐form dependent. Using the trait‐based approach, we found reliable characteristics of aquatic plants related to species invasions, which might be used, for example, in conservation management.

are most likely to be invaded (Pyšek & Richardson, 2006). All of these questions are related to the invasion paradox (Fridley et al., 2007), providing an explanation for how alien species can be more successful than natives in a more or less natural environment. Native species are generally described as being indigenous to an area since the last Ice Age, whereas aliens have been established due to human activities since then (Pyšek, 1995). Invasive species are those which have been able to overcome a series of geographical, environmental, and dispersal barriers and reproduce successfully in a new environment (Richardson et al., 2000). Both alien and native species are able to become invasive; the latter are also called "expanding natives" (Pyšek, 1995). To date, papers on alien species success mostly focused on terrestrial species; however, aquatic ecosystems are also seriously invaded by alien species (Lukács, Mesterházy, Vidéki, & Király, 2016) and there are papers that aim to study the success and biological attributes of alien aquatic plants (Kliber & Eckert, 2005;Riis et al., 2010).
An increasing number of studies are being published in invasion biology identifying species which are potentially invasive (Pyšek et al., 2012) and attempting to determine which traits enable them to be successful (Fenesi & Botta-Dukát, 2010;Shea & Chesson, 2002). For this purpose, a trait-based approach is frequently used, where predominantly whole plant or leaf traits are investigated (e.g., Grotkopp & Rejmanek, 2007;van Kleunen, Weber, & Fischer, 2010;Leishman, Haslehurst, Ares, & Baruch, 2007). Leaf traits are extensively used in plant ecology, being relatively easy to measure and strongly related to plant functions and fitness parameters (Pérez-Harguindeguy et al. 2013). Green leaves are strongly linked to primary production and carbon accumulation along the "leaf economics spectrum" (LES) (Wright et al., 2004), which describes the trade-off between the acquisitive and conservative strategies through leaf traits. This trade-off was found to be globally universal (i.e., independent from growth-forms and only moderately depending on climate) (Wright et al., 2004).
The applicability of traits in predicting and analyzing biological invasions is nowadays often debated (van Kleunen, Dawson, & Dostal, 2011;Thompson & Davies, 2011), emphasizing that invasive aliens exhibit the same set of traits as successful expanding natives (Leishman, Thomson, & Cooke, 2010). However, other studies revealed that only a handful of traits were universally linked to invasiveness, such as plant height, vegetative spatial growth, specific leaf area (SLA), and other traits related to performance (Pyšek &Richardson, 2007 andliterature therein, van Kleunen et al., 2010). In addition, Pyšek and Richardson (2007) concluded that invasiveness is strongly related to leaf traits associated with rapid C capture (high SLA, high leaf area ratio (LAR), and fast relative growth rate), while van Kleunen et al. (2010) found that invasive species have higher trait values for performance-related traits (reflecting physiology, leaf-area allocation, shoot allocation, growth rate, size, and fitness).
Hydrophytes are usually neglected from large-scale comparative trait-based studies; Poorter, Niinemets, Poorter, Wright, and Villar (2009) provided the only work which applied this group, classifying hydrophytes into a single life-form category on the one hand and splitting terrestrial species into numerous categories. In this review, the authors found that hydrophytes exhibited the lowest LMA values (i.e., leaf mass area, the reciprocal of SLA, indicating highly acquisitive strategies) compared to a range of terrestrial plant life-forms. In fact, hydrophytes represent a wide range of life-history strategies (hereafter "growth-forms") (Wiegleb, 1991;Wiegleb et al., 2015). The adaptive strategy of hydrophytes can be directly compared to those of terrestrial species by combining leaf economics and size traits. Besides, their adaptive strategy variation reflects the fundamental trade-offs in economics that govern all terrestrial plants (Pierce, Brusa, Sartori, & Cerabolini, 2012) so that they could be included in the global spectrum of plant form and function (Díaz et al., 2015). Due to the various economics in contrasting hydrophyte growth-forms, we suggest that the general models of plant traits comparing alien and native species should also be applied to aquatic species.
Two alternative hypotheses exist to explain the probability of success of alien species: "phenotypic convergence" (Daehler, 2003;Smith & Knapp, 2001) and "phenotypic divergence" (van Kleunen et al., 2010;Lake & Leishman, 2004) depending on whether they found phenotypic similarities or differences between the studied traits in native and alien species. "Phenotypic convergence" is based on the concept of habitat filtering (Weiher, Clarke, & Keddy, 1998), which refers to environmental (abiotic) factors that prevent the establishment or persistence of certain species in a given location; that is, they are "filtered out" based on their traits. This suggests that alien species can only be successful if they are similar to natives. Alternatively, "phenotypic divergence" is related to the concept of limiting similarity (MacArthur & Levins 1967) meaning that competition is strongest between the most similar species. Therefore, by having different traits, alien species can be more successful than natives in the introduced community.
The first aim of the study was to compare alien and native aquatic plant species in terms of four key leaf traits (leaf area, leaf dry matter content, specific leaf area, and leaf nitrogen content) to determine whether we can identify specific traits that might explain the success of alien species over natives. Secondly, we aimed to investigate which hypotheses ("phenotypic convergence" or "phenotypic divergence") explain the trait composition of co-occurring native and alien aquatic plant species. Studies aiming at comparing native and invasive plant species by traits usually use key traits that represent independent axes of ecological strategy or niche dimension such as leaf, seed and height traits (Westoby 1998, Ordonez, Wright, & Han, 2010. Height is measured as the difference between the elevation of the highest photosynthetic tissue in the canopy and the base of the plant (Weiher et al., 1999). For hydrophytes, canopy height is difficult to measure where different growth-forms position leaves equally at water-air interface, but may be free floating or anchored to the substrate. Among seed traits, the average individual seed weight (SWT) is predicted to be the most adequate trait; however, among aquatic plants, vegetative reproduction usually predominates over sexual reproduction (Grace, 1993).
In contrast, leaf economics and size traits can reflect adaptive strategy variations among hydrophytes (Pierce et al., 2012).
It is important to note that our alien species are invasive aliens according to the definition of Richardson et al. (2000) (i.e., naturalized plants that produce reproductive offspring, often in very large numbers, at considerable distances from parent plants; <50 years for taxa spreading by seeds and other propagules; >6 m/3 years for taxa spreading by roots, rhizomes, stolons, or creeping stems and thus have the potential to spread over a considerable area). Thus, we conducted invasive-native comparisons. This question focuses only on the ability of species to become invasive and do not consider community invasibility (Pyšek & Richardson, 2006).

| Databasecompilation
Trait values of aquatic plant species from 20 water bodies of North Italy were extracted from the TRY database (Kattge et al., 2011).
Additional data were obtained from field sampling from two creeks (Hévíz-creek and Tapolca-creek) of West Hungary. As a result, our database contains 50 species from Italy and 19 species from Hungary, and nine species were common between the countries. There were three sites contain only alien species and 13 sites have only native species; alien and native species co-occurred in six sites. At each site, we measured pH, conductivity, water depth, altitude, and latitude as these are the most relevant environmental variables of macrophytes (see Barendregt & Bio, 2003;Lacoul & Freedman, 2006;O'Hare, Gunn, Chapman, Dudley, & Purse, 2012). Among the environmental variables, altitude has been proven to have an effect on trait variation via temperature (Reich, Walters, & Ellsworth, 1997), whereas water depth significantly promotes intraspecific trait variability in macrophytes (Fu et al., 2014). In the sampled watercourses, SECCHI transparency of the water was higher than the average water depth (>2 m); creeks were not shaded by buildings, trees, or shrubs. Thus, we assumed that light availability and photosynthetically active radiation (PAR) were constant and had no effect on plant morphology in any of the sampling sites. In this way, we make sure that leaf traits reflect the spectrum of chemical, structural, and physiological properties of species (Shipley, Lechowicz, Wright, & Reich, 2006). We generally considered it is the best metric to evaluate the differences in species ecological behaviors (Wilson, Thompson, & Hodgson, 1999).
Native species names follow Tutin et al., 2001; and alien species names follow USDA 2016 database, while names of Nymphaea cultivars follow Slocum, 2005.

| Selectionoftraits
We followed the same standardized protocol (Cornelissen et al., 2003) as recommended by Pierce et al. (2012): 10 fully expanded, intact leaves of each species were collected from separate individuals. We measured four key traits on all aquatic plant species: (i) Leaf area (LA or leaf size) is strongly related to the energy and water balance of leaves (Cornelissen et al., 2003); (ii) specific leaf area (SLA) is part of the leaf economics spectrum (LES) and strongly correlated with photosynthetic capacity, nitrogen content per leaf mass, and leaf life span (Reich et al., 1999;Wright et al., 2004); (iii) leaf dry matter content (LDMC) reflects the average density of leaf tissues and a trade-off between the investments in structural tissues versus liquid-phase processes. Leaf dry matter content is a key variable that governs the correlations among the traits in the leaf economics spectrum (LES), which is considered a robust trait (Roche, Diaz-Burlinson, & Gachet, 2004) and usually negatively correlated with relative growth rate (Weiher et al., 1999); (iv) leaf nitrogen content (LNC) is calculated as the total amount of nitrogen per unit of dry leaf mass. High values of LNC are associated with high nutritional quality (Cornelissen et al., 2003), which is a predictor of photosynthetic capacity in terms of leaf economics, similar to SLA (Nijs, Behaeghet, & Impens, 1995). This indicates the nitrogen-use efficiency of plants, which varies with nitrogen availability in the environment, however.
Out of every sampled individual, the youngest, fresh, and healthy but fully developed leaf was collected and scanned using a flatbed desktop scanner and leaf area (LA, mm 2 ) was measured using ImageJ (http://imagej.nih.gov/ij) open source image analysis software. The same leaves were weighted in fresh conditions (fresh mass, g) and weighted again after 48 hr of oven-drying at 80°C (dry mass, mg); then, leaf dry matter content (LDMC, dry mass/fresh mass, mg/g) and specific leaf area (SLA, mm 2 /g = leaf area/dry mass) were calculated.
LNC values were measured in three oven dry leaves per species using ICP-MS (Agilent 8800 triple quad). Individual measurements were averaged for each species.
Species were classified into "native" and "alien" types based on their native/alien status following the DAISIE 2009 list and Lukács et al., 2016 at the corresponding sampling site. All plants were grouped into growth-form categories according to Wiegleb (1991)

| Dataanalysis
We applied individual trait comparisons to reveal the differences of native and alien species. We used linear mixed models to test whether native and alien species differed significantly in individual traits (LA, SLA, and LDMC). We specified the model in a hierarchical form: The evaluated trait was treated as a response variable; plant type (native or alien), growth-form, and altitude were treated as a fixed factor, while country, site (nested within country), and species identity were treated as a random factor. The use of site and country as a random factor allowed us to compare native and alien communities cooccurring under the same environmental conditions, while the use of species identity as a random factor controlled for the possible relatedness of native and alien species. Taxonomic similarity (i.e., congeneric or confamiliar) cannot be considered due to the lack of native-alien species pairs within genera. To improve normality of predictors, all traits were log 10 -transformed for all analysis.
During model fitting, we entered and excluded all effects sequentially until only variables explaining significant variation remained.
Significance of fixed terms was accepted if t > 2.00 (Crawley, 2007).
All dropped variables were included again in the model to obtain levels of nonsignificance. We applied the same method to test whether significant effects had not been wrongly excluded. The minimal model was derived by removing terms from the maximal model and adding effects to the simplest model (Pinheiro & Bates, 2000).

All analyses was performed in R environment (R Development
Core Team 2009) using the lme4 package (Bates, Mächler, Bolker, & Walker, 2015).  Table 3). Alien species exhibited higher LA values within Myriophyllid, Peplid (just as across all species), Pleustophyte, and Potamid species (opposite to all species), whereas no differences were seen within Herbids, Nymphaeids, and Vallisnerids. Substantial differences between leaf areas can also be attributed to pH within Peplids, whereas it can be attributed to latitude, pH, and water depth within Pleustophytes.

| RESULTS
LDMC of alien and native species substantially differed among Herbids and Potamids with shifts to the opposite direction which is seen in the all-species comparison. Differences between LDMC can also be attributed to altitude, latitude, and pH within Myriophyllids. For LNC values, the group difference was found to be substantial between alien and natives in Herbid, Nymphaeid, and Vallisnerid species (just as across all species). Latitude can be attributed to the substantial differences of LNC within Pleustophytes.

T A B L E 2 (Continued)
T A B L E 3 Differences of leaf traits between alien and native aquatic vascular plants using final linear mixed models  (Continues) nutrients in the leaves to remain compact, whereas at the same time, aliens tend to produce larger leaves and have faster growth rates (via high SLA) and therefore are able to outcompete co-existing natives.
However, the comparisons of individual traits revealed that growth-form also have a substantial effect on traits variation (LA, SLA, and LNC). Hydrophytes incorporate several morphologically distinct growth-forms; thus, species (via growth-forms) represent different plant strategies as well. We assumed that the large differences between growth-forms might mask differences between alien and native species; therefore, in order to control for this effect, we compared native and alien species within growth-forms and obtained significant differences in the case of all four traits (Figure 2). This finding is in line with studies demonstrating that the response of different growthform of aquatic plants to local environmental variables (Akasaka & Takamura, 2011;Alahuhta et al., 2013) and the response of plant invasiveness (Hamilton et al., 2005) varied significantly.
LDMC is a key trait of the leaf economics spectrum representing the average density of leaf tissues, positively correlated with leaf life span and negatively correlated with relative growth rate and SLA (Cornelissen et al., 2003). Country, site (nested within country), and species identity were treated as a random factor. Peplid and Vallisnerid growth-form was omitted from the analyses due to the lack of alien-native species pairs. Traits with |t| > 2.00 mean substantial differences and indicated with an asterisk (*). Traits with |t| < 2.00 means non-substantial differences and indicated with NS dense leaf tissues of alien species might be an advantage in the competition for light. In fact, lower average density of leaf tissues enables aliens to build up their photosynthetic organs faster and more easily and invest less into structural tissue elements. Overall, these taxa can reach faster growth rates (Weiher et al., 1999), which is certainly a competitive advantage.
Our results indicate that leaf area is an important trait in the separation of alien and native aquatic plant species, as it was also proved in the case of terrestrial species (Daehler, 2003;Pyšek & Richardson, 2007). Interspecific variation in LA (and leaf area index-LAI) has been usually related to climatic, geological, altitudinal, and latitudinal factors. Within climatic zones, LA may be linked to ecological strategies (Westoby & Wright, 2003). In our study, differences in LA were paramount in the Myriophyllid, Peplid, Pleustophyte, and Potamid growth-forms. Apparently, a large-leaved Potamid or a dense-leaved Myriophyllid alien can achieve dominance over native species, forming a monolayer or dens canopy and can directly inhibit other species in the competition for light. Interestingly, traits related to growth rate (SLA) were found to be substantial also in the case of these growthforms (except Peplids), which suggest that these species produce larger leaves in a faster and easier way.
Former studies have shown that SLA is the most influential trait in the leaf economics spectrum (Saverimuttu & Westoby, 1996), which also reflects relative growth rate (RGR = assimilation rate × leaf mass ratio × specific leaf area). There is also a trade-off between SLA and leaf life span (leaf longevity) and have a strong relationship with net photosynthesis (i.e., growth rate) (Osnas, Lichstein, Reich, & Pacala, 2013). Therefore, species with high SLA are associated with a strategy where only a small amount of biomass is invested into building shortlasting structures. Our results suggest that SLA of alien and native aquatic plants differs substantially in almost all growth-forms (except Nymphaeid and Pepild), which are in line with the results of Lake and Leishman (2004) and Hamilton et al. (2005)  Nitrogen (together with phosphorous) is generally considered to be one of the most limiting elements in terrestrial and aquatic environments. LNC, similar to SLA, also reflects photosynthetic activity (i.e., growth rate) in an alternative way (Cornelissen et al., 2003). In contrast to SLA, LNC represents differences in photosynthetic activity, considering the effectiveness of nutrient recovery. We supposed that alien aquatic plants have an enhanced nitrogen-use efficiency which was reflected by higher LNC values. Global patterns of leaf N content showed a decline toward the Equator, which indicates a strong relationship with latitude and temperature (Reich & Oleksyn, 2004).
Contrary to our hypotheses, our results indicate that nativeness is not related to the photosynthetic activity of aquatic plants in general.
However, alien Nymphaeid and Vallisnerid species exhibit higher nitrogen concentrations, which indicate enhanced and more effective photosynthesis and growth rates therein. Moreover, we also justified the latitude and temperature response of leaf N content.

| Theeffectofenvironmentalvariablesontrait variationandnativeness-traitsrelationship
The plasticity of certain traits and the diversity of traits are known to depend on the ambient environment (Capers, Selsky, & Bugbee, 2010;Hodgson et al., 2011;Richards, Bossdorf, Muth, Gurevitch, & Pigliucci, 2006). Daehler (2003) found that differences between alien and native species strongly depend on the environment, and the performance of alien species might be better under high resource availability in benign conditions (Richards et al., 2006). Contrary to van Kleunen et al. (2010), who found that the trait relationship of terrestrial species did not depend on the quality of the environment, and the differences between native and alien terrestrial species were robust across environments, our analysis showed the relative importance of abiotic factors in some of the trait-nativeness relationship within some aquatic plant growth-forms.
The effect of altitude and latitude on aquatic plant diversity and distribution is well known in the literature (e.g., Heegaard, Birks, Gibson, Smith, & Wolfe-Murphy, 2001;Jones, Li, & Maberly, 2003;Lukács et al., 2015). Temperature and light availability varies on an elevational and latitudinal gradient, and there is a clear trade-off between altitude and diversity (Jones et al., 2003). Moreover, it has been proven that altitude and latitude (via temperature and precipitation) have an effect on trait variation of terrestrial plant species (Hulshof et al., 2013;Reich et al., 1997). Riis et al. (2012) also pointed out that temperature can affect the competitive ability of the alien Lagarosiphon major via phenotypic plasticity. Contrary to that we found, little effect of latitude on the trait variability was found between native and alien aquatic plants, which is presumably due to the short latitudinal gradient. Latitude substantially affected the trait-nativeness relationship only in case of LA within the Pleustophyte growth-form, but also had a substantial effect on LDMC variation of Myriophyllids.
Aquatic ecosystems at high altitudes are considered as extreme environments in which physical stressors and severe climate may limit the distribution of aquatic plants (Lacoul & Freedman, 2006 Potamids. This is partly in line with the results of Hulshof et al. (2013) who pointed out intra-and interspecific variation of SLA along elevational gradients.
The importance of pH has been well documented in aquatic plant ecology; it is related to physiological differences (i.e., the ability to use bicarbonate as carbon source) among species (Madsen & Sand-Jensen, 1991 Hutchinson, 1975;Nielsen & Sand-Jensen, 1989).
Previous studies indicated that water depth has a significant influence on individual trait variation (e.g., shoot height, stem dry mass, see Maberly, 1993;Fu et al., 2012) . We also pointed out that environmental variables such as altitude, pH, and water depth are important factors in studied response of nativeness and leaf traits.
The question of identifying traits promoting plant invasiveness is important for understanding plant success in general and also in planning risk assessment protocols and management and preventive actions. We believe that our results provide new insights into what makes an aquatic alien plant to be successful in temperate climate which is expected to inform conservation management strategies or as a base to make an inventory of alien species whose import and placement on the market will be prohibited/permitted (black/white lists).
We emphasize that the trait approach we applied here can only partly contribute to the understanding of the mechanism of aquatic plant invasions. Further research is needed to clarify the trait dependence of this issue, especially (i) to explore the phylogenetical dependence of trait differences, (ii) to explore the intraspecific variation of aquatic plant's trait values to obtain finer conclusions, (iii) to collect more functional trait data from aquatic plants to make multitrait comparisons possible, and (iv) to explore the differences in functional community assembly between native and alien aquatic plant communities.