Is seed availability enough to ensure colonization success?: An experimental study in road embankments

We tested the hypothesis that seed availability is a limiting factor for plant colonization of road embankments under Mediterranean climate conditions. Experimental sowing on 10 road embankments was carried out to compare the colonization success of plants that successfully colonize the road embankment and species that appear only occasionally in the road embankments. After sowing, we measured plant establishment, biomass production, and reproductive capacity of the species.The species that appear only occasionally in the road embankments had lower emergence rates (l.l ± 0.3%) than species that were successful colonizers (18.8 ± 2.9%). None of the species of the former group survived or reproduced. The results did not support the hypothesis that seed availability was the main factor limiting plant colonization in the road embankments. We concluded that the arrival of seeds to road embankments under Mediterranean climate conditions was not enough to ensure colonization success of plants. Other factors, like hydric stress, appeared to affect seedling establishment and plant growth. Reclamation measures such as species selection should be taken in account to ensure revegetation success of road embankments.     

Spatial variation in seed limitation of plant species richness and population sizes in floodplain tallgrass prairie

The relative importance of seed availability versus biotic interactions that affect early life stages in limiting plant population sizes and determining composition of plant communities is a central debate in plant ecology. We conducted a seed addition experiment in restored tallgrass prairie in central Kansas to determine (1) whether addition of seed of 18 native forb species produced persistent (three growing seasons) increases in the species' population sizes and plant species richness, (2) what properties of recipient communities best explained spatial variation in added species' establishment, and (3) whether seed size explained interspecific patterns in establishment success. Adding seed led to persistent increases in the number of added species present and in plant species richness at one of three sites. Increased species richness at the one site where community composition was structured by seed availability largely resulted from greater densities of four species. Seed size did not predict species' establishment success. Pre-existing plant species richness was correlated with added species' establishment success, but the direction of the relationship (positive vs. negative) varied among sites. Living aboveground plant biomass in experimental plots in the year of seed addition was negatively correlated with the number of added species established three years later. Our results provide further evidence for large spatial variation in seed limitation of plant community composition. Surprisingly, mean light availability and heterogeneity in light, both important parameters in conceptual models of grassland plant coexistence, did not predict the response of the recipient plant community to seed addition as well as pre-existing plant species richness and living aboveground biomass.     

Priority Effects of Time of Arrival of Plant Functional Groups Override Sowing Interval or Density Effects: A Grassland Experiment

Priority effects occur when species that arrive first in a habitat significantly affect the establishment, growth, or reproduction of species arriving later and thus affect functioning of communities. However, we know little about how the timing of arrival of functionally different species may alter structure and function during assembly. Even less is known about how plant density might interact with initial assembly. In a greenhouse experiment legumes, grasses or forbs were sown a number of weeks before the other two plant functional types were sown (PFT) in combination with a sowing density treatment. Legumes, grasses or non-legume forbs were sown first at three different density levels followed by sowing of the remaining PFTs after three or six-weeks. We found that the order of arrival of different plant functional types had a much stronger influence on aboveground productivity than sowing density or interval between the sowing events. The sowing of legumes before the other PFTs produced the highest aboveground biomass. The larger sowing interval led to higher asymmetric competition, with highest dominance of the PFT sown first. It seems that legumes were better able to get a head-start and be productive before the later groups arrived, but that their traits allowed for better subsequent establishment of non-legume PFTs. Our study indicates that the manipulation of the order of arrival can create priority effects which favour functional groups of plants differently and thus induce different assembly routes and affect community composition and functioning.     

Early Dynamics of Plant Communities on Revegetated Motorway Slopes from Southern Spain: Is Hydroseeding Always Needed?

The increasing global rate of road construction is leading to a parallel increase of areas of degraded soil conditions and steep slopes that need revegetation. Hydroseeding with commercial seeds of fast‐growing grasses and legumes is a common practice in revegetation of motorway slopes. We carried out 3 years of monitoring of vegetation dynamics on hydroseeded and nonhydroseeded motorway slopes (48 slopes) in a maritime Mediterranean zone in Málaga (southern Spain). Our main objectives were to test whether hydroseeding significantly increases species richness and plant cover and whether hydroseeded species act as starters, facilitating the establishment of the vegetation and quickly disappearing once the communities are established. A hydroseeding success index (HSI, ranging from 0 to 1) was used to assess the relative abundance over time of the 14 species from the hydroseeding mixture. Species richness and cover was significantly higher on embankments (50–70 species per embankment, 80–90% cover) than on roadcuts (6–10 species per roadcut, 18–30% cover). Performance of hydroseeded species was poor from the very beginning (HSI, 0.2–0.3). On embankments, either presence or abundance of hydroseeded species did not significantly vary throughout the study. Both hydroseeded and nonhydroseeded communities exhibited a significant decrease in species richness, a significant increase in plant cover, and a highly dynamic species composition over time, with Sorensen index of 0.3–0.5 between years. There were no significant differences in plant cover, species richness, and aboveground biomass between hydroseeded and nonhydroseeded plots on embankments throughout the study. Our results demonstrate that there are situations in which the use of hydroseeding for revegetation is not needed. Further research should focus on understanding the establishment of autochthonous species and identifying environmental conditions under which the addition of commercial seeds may not be needed, or indeed situations where it may be harmful in suppressing autochthonous species.     

Testing Germination of Species for Hydroseeding Degraded Mediterranean Areas

The germination performance of native species and their suitability for a rapid erosion control are uncertain. Together with their relatively low commercial availability and high costs, these are still strong reasons preventing their common use in hydroseeding for restoration of Mediterranean degraded slopes, despite the increasing number of studies recommending it. In this study, 14 non‐native (commercial) and native herb and woody species were tested. Their germination performance was evaluated under laboratory (Petri dishes) and greenhouse conditions (seeds sown in target substrate). The results obtained were compared with the seedling densities in a Mediterranean quarry slope hydroseeded with the same species. In the laboratory, commercial species had a better germination performance than most native species, but this trend was not maintained in the greenhouse. Greenhouse tests were extended beyond spring and showed that many native species germinated better, or exclusively in autumn. Germination performance and success decreased, from laboratory to greenhouse and field conditions, for many species, but not for all. Relative to field performance, the predictive value of laboratory and greenhouse tests was poor, yet sowing on the target substrate under greenhouse conditions may be a better approach for certain native species. The main drawbacks revealed by native species in the present study included: (1) relatively slow germination; (2) seasonality; and (3) seed dormancy‐breaking requirements. The results suggest that these problems may be overcome through species selection, seed pre‐treatments, hydroseeding scheduling, and/or manipulation of seeding density and relative species proportion.     

The influence of seed addition and cutting regime on the success of grassland restoration on former arable land

Questions: Can seed addition enhance the success of establishing species‐rich grassland on former arable land? Are sowing date and cutting regime important in determining success? Location: Aberdeen and Elgin, northeast Scotland, United Kingdom. Methods: A field experiment was conducted at two sites to assess the effect of seed addition, sowing date and cutting regime on the vegetation developing on former arable land, the aim being to compare the success of different treatments at producing a species‐rich grassland. Results: Sowing a seed mix resulted in the establishment of vegetation very distinct from the species‐poor vegetation dominated by perennial grasses which otherwise developed, though establishment success of the sown grassland species was highly variable between sites. Where establishment of the sown species was poor, sowing date had no significant effect on species composition, whereas the cutting regime was very important. Cutting the vegetation significantly increased both the number and abundance of sown species compared with the uncut control. Conversely, where establishment had been good, the cutting regime in the first year had little effect on species composition. Cutting the vegetation at least twice a year appeared to be the most effective management over the length of the experiment. Conclusions: Sowing a seed mixture significantly reduced the abundance and number of naturally colonising species, effectively controlling problem weed species such as Senecio jacobaea and Cirsium vulgare, highlighting the agronomic value of sowing seed mixtures on fallow farmland. The sowing of a seed mix on former arable land has demonstrated that it is feasible to create vegetation similar in character to that of species‐rich grasslands.     

Regional adaptation improves the performance of grassland plant communities

Many plant species are adapted locally or regionally. Whether such individual species performance translates into effects at community and ecosystem levels has rarely been tested. Such tests are crucial, however, to predict ecosystem consequences of sowing seed mixtures for grassland restoration or hay production. We compared the performance of replicated sown plant communities of regional origin with the performance of four foreign communities consisting of the same grassland species but originating from distances up to 890 km from our experimental site. The regional communities performed better than foreign communities in plant cover and diversity but not in aboveground biomass production. Additionally, in communities based on regional seeds fewer unsown species occurred and less bare ground was left open for erosion. Variation in community performance among source regions was related to climatic differences rather than to geographic distance to source regions. Individual species performance only partly explained community patterns, highlighting the importance of community level experiments. Our results suggest that the use of regional seeds represents an important approach to improve sown managed grasslands.     

Restoration of Prairie Community Structure and Ecosystem Function in an Abandoned Hayfield: A Sowing Experiment

Using a multispecies seed sowing experiment, we investigated the roles of seed and microsite limitation in constraining the restoration of native prairie diversity and ecosystem function in an abandoned upland hayfield in northeastern Kansas. Seeds of 32 native and naturalized plant species from the regional pool were sown into undisturbed and experimentally disturbed field plots. After six growing seasons, experimental sowing led to major shifts in species and functional group composition, increases in native species abundance and floristic quality, declines in abundance of non‐native species, and increases in plant diversity. These changes in community structure led to significant changes at the ecosystem level including increases in light capture, peak biomass, primary production, litter biomass, root biomass, and C storage in roots. Our findings reveal the importance of seed limitations in constraining the natural recovery of prairie vegetation, biodiversity, and ecosystem functioning in this grassland and confirm broadcast sowing as a useful tool for the restoration of upland hayfield sites.     

Analysis of interactions between the invasive tree-of-heaven (<Emphasis Type="Italic">Ailanthus altissima</Emphasis>) and the native black locust (<Emphasis Type="Italic">Robinia pseudoacacia</Emphasis>)

Invasive exotic plants can persist and successfully spread within ecosystems and negatively affect the recruitment of native species. The exotic invasive Ailanthus altissima and the native Robinia pseudoacacia are frequently found in disturbed sites and exhibit similar growth and reproductive characteristics, yet each has distinct functional roles such as allelopathy and nitrogen fixation, respectively. A four-month full additive series in the greenhouse was used to analyze the intraspecific and interspecific interference between these two species. In the greenhouse experiment, the inverse of the mean total biomass (g) response per plant for each species was regressed on the density of each species and revealed that the performance of the plants was significantly reduced by interspecific interference and not by intraspecific interference (p < 0.05). Other biomass traits such as root dry weight, shoot dry weight, stem dry weight, and leaf dry weight were also negatively affected by interspecific interference. Competition indices such as Relative Yield Total and Relative Crowding Coefficient suggested that A. altissima was the better competitor in mixed plantings. Ailanthus altissima consistently produced a larger above ground and below ground relative yield while R. pseudoacacia generated a larger aboveground relative yield in high density mixed species pots.     

Assessment of Cu,Pb and Zn content in selected species of grasses and in the soil of the roadside embankment

It was assumed in the study that heavy metals occurring in soils and the air accumulate in grasses constituting the main species used in the turfing of soil in road verges and embankments along traffic routes and in other parts of urbanized areas. The aim of the present study was to assess the bioaccumulation of Cu, Pb, and Zn in three selected lawn cultivars of five grass species and in the soil of the roadside green belt in terms of soil properties and heavy metal uptake by plants in the aspect of determining their usefulness in protecting the soils from contamination caused by motor vehicle traffic. Samples of the plant material and soil were collected for chemical analysis in the autumn of 2018 (October) on the embankment along National Road No. 17 between Piaski and ?opiennik (Poland), where 15 lawn cultivars of five grass species had been sown 2 years earlier. During the study, Cu, Pb, and Zn levels were determined in the aboveground biomass of the grasses under study and in the soil beneath these grasses (the 0–20 cm layer). All the grass species under study can thus be regarded as accumulators of Cu and Zn because the levels of these elements in the aboveground biomass of the grasses were higher than in the soil beneath these grasses. The present study demonstrates that the grasses can accumulate a large amount of Cu and Zn from soils and transfer it to the aboveground biomass. Tested species of grasses are not a higher bioaccumulators for Pb. The best grass species for the sowing of roadsides embankment, with the highest BCF values for the studied metals, is Lolium perenne (Taya variety).     

Community assembly and biomass production in regularly and never weeded experimental grasslands

We studied the natural colonisation of new species in experimental grasslands varying in plant species richness (from 1 to 60) and plant functional group richness (from 1 to 4) in either regularly or never weeded subplots during the first 3 years after establishment. Sown species established successfully, with no differences in species richness or their relative abundances between the regularly and never weeded subplots during the study period. Aboveground biomass of sown species increased with increasing sown species richness in both treatments. While a positive relationship between sown species richness and total aboveground biomass (including colonising species) existed in the 2nd year after sowing in the regularly and never weeded subplots, this positive relationship decayed in the 3rd year in the never weeded subplots because of a higher biomass of colonising species in species-poor mixtures. Total aboveground biomass varied independently of total species richness 3 years after sowing in both treatments. Jaccard similarity of coloniser species composition between regularly and never weeded subplots decreased from the 2nd to the 3rd year, indicating a divergence in coloniser species composition. Coloniser immigration and turnover rates were higher in regularly weeded subplots, confirming that weeding counteracts species saturation and increases the chance that new colonisers would establish. Although our study shows that low diversity plant communities are unstable and converge to higher levels of biodiversity, the effects of initially sown species on community composition persisted 3 years after sowing even when allowing for succession, suggesting that colonising species mainly filled empty niche space.     

Effects of seed traits variation on seedling performance of the invasive weed,Ambrosia artemisiifolia L.

Seedling performance can determine the survival of a juvenile plant and impact adult plant performance. Understanding the factors that may impact seedling performance is thus critical, especially for annuals, opportunists or invasive plant species. Seedling performance can vary among mothers or populations in response to environmental conditions or under the influence of seed traits. However, very few studies have investigated seed traits variations and their consequences on seedling performance. Specifically, the following questions have been addressed by this work: 1) How the seed traits of the invasive Ambrosia artemisiifolia L. vary among mothers and populations, as well as along the latitude; 2) How do seed traits influence seedling performance; 3) Is the influence on seedlings temperature dependent. With seeds from nine Western Europe ruderal populations, seed traits that can influence seedling development were measured. The seeds were sown into growth chambers with warmer or colder temperature treatments. During seedling growth, performance-related traits were measured. A high variability in seed traits was highlighted. Variation was determined by the mother identity and population, but not latitude. Together, the temperature, population and the identity of the mother had an effect on seedling performance. Seed traits had a relative impact on seedling performance, but this did not appear to be temperature dependent. Seedling performance exhibited a strong plastic response to the temperature, was shaped by the identity of the mother and the population, and was influenced by a number of seed traits.     

Altering native community assembly history influences the performance of an annual invader

Understanding the determinants of early invasion resistance is a major challenge for designing plant communities that efficiently repel invaders. Recent evidence highlighted the significant role of priority effects in early community assembly as they affect species composition, structure and functional properties, but the consequences of native community assembly history on the success of subsequent invasions has not been elucidated yet. In a greenhouse experiment, we investigated how (1) the identity of the first native colonizing species (one of two grasses: Dactylis glomerata and Lolium perenne, or two legumes: Onobrychis viciifolia and Trifolium repens), each introduced four weeks before the rest of the native community, and (2) timing of species establishment (synchronous vs. sequential sowing), influenced early establishment success of Ambrosia artemisiifolia, an annual noxious weed in Europe. First colonizer identity and establishment timing both affected early biomass production and composition of the community, and had implications for A. artemisiifolia early invasion success. Invasion success decreased when all native individuals were sown simultaneously, quickly generating a high biomass production, while it increased when the productive N-fixing legume T. repens was sown first. These findings support that native species assembly history matters to invasion resistance in the early growth stages, thus opening the way to more effective invasive species management strategies in restoration.     

More plant biomass results in more offspring production in annuals,or does it?

Competitive ability in plants has been previously measured almost exclusively in terms of traits related to growth (biomass) or plant size. In this study, however, we used a multi‐species competition experiment with six annuals to measure relative competitive ability in terms of reproductive output, i.e. the number of offspring produced for the next generation. Under greenhouse conditions, plants of each species were started in pots from germinating seeds and were grown singly (free of competition) and at high density in both monocultures and in mixtures with all study species. Several traits traditionally regarded as determinants of competitive ability in plants were recorded for each species grown singly, including: seed mass, germination time, early growth rate and potential plant size (biomass and height). Under competition, several traits were recorded as indicators of relative performance in both monocultures and mixtures, including: biomass of survivors, total number of survivors, number of reproductive survivors, and reproductive output (total seed production) of the survivors. As expected, species that grew to a larger biomass in isolation had higher seed production in isolation. However, none of the traditional plant growth/size‐related traits, measured either in isolation or under competition, could predict between species variation in reproductive output under competition in either monocultures or mixtures. In mixtures, 97% of this variation in reproductive output could be explained by between‐species variation in the number of reproductive survivors. The results indicate that traits measured on plants grown singly may be poor predictors of reproductive output under competition, and that species’ rank order of competitive ability in terms of the biomass of survivors may bear no relationship to their rank order in terms of the number of offspring produced by these survivors. This has important implications for the interpretation of mechanisms of species coexistence and community assembly within vegetation.     

Density may alter diversity–productivity relationships in experimental plant communities

Contemporary biodiversity experiments, in which plant species richness is manipulated and aboveground productivity of the system measured, generally demonstrate that lowering plant species richness reduces productivity. However, we propose that community density may in part compensate for this reduction of productivity at low diversity. We conducted a factorial experiment in which plant functional group richness was held constant at three, while plant species richness increased from three to six to 12 species and community density from 440 to 1050 to 2525 seedlings m⁻². Response variables included density, evenness and above- and belowground biomass at harvest. The density gradient converged slightly during the course of the experiment due to about 10% mortality at the highest sowing density. Evenness measured in terms of aboveground biomass at harvest significantly declined with density, but the effect was weak. Overall, aboveground, belowground and total biomass increased significantly with species richness and community density. However, a significant interaction between species richness and community density occurred for both total and aboveground biomass, indicating that the diversity–productivity relationship was flatter at higher than at lower density. Thus, high species richness enabled low-density communities to reach productivity levels otherwise seen only at high density. The relative contributions of the three functional groups C₃, C₄ and nitrogen-fixers to aboveground biomass were less influenced by community density at high than at low species richness. We interpret the interaction effects between community density and species richness on community biomass by expanding findings about constant yield and size variation from monocultures to plant mixtures.     

The effects of nutrient addition on floral characters and pollination in two subalpine plants, <Emphasis Type="Italic">Ipomopsis aggregata</Emphasis> and <Emphasis Type="Italic">Linum lewisii</Emphasis>

The availability of soil and pollination resources are main determinants of fitness in many flowering plants, but the degree to which each is limiting and how they interact to affect plant fitness is unknown for many species. We performed resource (water and nutrients) and pollination (open and supplemental) treatments on two species of flowering plants, Ipomopsis aggregata and Linum lewisii, that differed in life-history, and we measured how resource addition affected floral characters, pollination, and reproduction (both male and female function). We separated the direct effects of resources versus indirect effects on female function via changes in pollination using a factorial experiment and path analysis. Resource addition affected I. aggregata and L. lewisii differently. Ipomopsis aggregata, a monocarp, responded to fertilization in the year of treatment application, increasing flower production, bloom duration, corolla width, nectar production, aboveground biomass, and pollen receipt relative to control plants. Fertilization also increased total seed production per plant, and hand-pollination increased seeds per fruit in I. aggregata, indicating some degree of pollen limitation of seed production. In contrast, fertilization had no effect on growth or reproductive output in the year of treatment on L. lewisii, a perennial, except that fertilization lengthened bloom duration. However, delayed effects of fertilization were seen in the year following treatment, with fertilized plants having greater aboveground biomass, seeds per fruit, and seeds per plant than control plants. In both species, there were no effects of resource addition on male function, and the direct effects of fertilization on female function were relatively stronger than the indirect effects via changes in pollination. Although we studied only two plant species, our results suggest that life-history traits may play an important role in determining the reproductive responses of plants to soil nutrient and pollen additions.     

Seedling recruitment on agriculturally improved mesic grassland: the influence of disturbance and management schemes

Question: Are the recruitment patterns of deliberately introduced wildflower species influenced by cutting frequencies and disturbance treatments? To what extent do these different treatments affect productivity and sward structure of an agriculturally improved grassland? Location: A mesic lowland grassland near Göttingen, Lower Saxony, Germany. Methods: Recruitment success of eight sown wildflower species was studied in a permanent grassland treated by a factorial combination of different pre‐sowing cutting intervals (1, 3 or 9 wk), post‐sowing cutting intervals (1,3 or 9 wk) and disturbance (control, harrowing, removal of sward). Seedling emergence and survival, biomass production and sward structure were followed over two years. Results: For most species seedling emergence was highest in the harrowing treatment. The complete sward removal did not further increase seedling emergence. Seedling survival was strongly influenced by the post‐sowing cutting frequency with highest mortality in the 9 wk cutting interval compared to one and 3 wk cutting intervals. Annual dry matter yield varied between 4.4, 5.9 and 9.4 t.ha^‐1 in the 1,3 and 9 wk pre‐sowing cutting treatment, respectively. In June, when the seeds were sown, the tiller number of the 1 wk cut plots was twice as high as for the 9 wk cut plots and five times higher than in the harrowing treatment. Conclusions: Disturbance by harrowing provided the optimal environmental cues to trigger germination, whereas seedling survival was facilitated by increased light penetration due to frequent cutting. The investigation revealed the overriding importance of frequent standing crop removal in the early phase of seedling establishment on agriculturally improved grassland.     

Aggregated seed arrival alters plant diversity in grassland communities

Aims Species aggregation is commonly seen in plant communities and may increase diversity by causing intraspecific competition to exceed interspecific competition. One potential source of this spatial aggregation is seed dispersal but it is unclear to what extent aggregated seed distributions affect plant diversity in real communities. Using a field experiment, I tested whether uniform or aggregated seed arrival alters community structure and whether these effects vary with sowing density.Methods The experiment consisted of two spatial seeding treatments (uniform and aggregated) that were fully crossed with three seed density treatments. Sixty, 3 × 4-m plots were arrayed in a low-diversity grassland located in Kansas, USA. Each plot was divided into forty-eight, 0.5 × 0.5-m patches. For aggregated seeding treatments, each of the 15 species was sown into three randomly selected patches within the plot (3×15 = 45). To create a uniform species arrival but control for the seed addition method, all 15 species were sown into 45 individual patches (with three patches remaining unsown) within each plot. Seed mass for each species was held constant at the plot scale between uniform or aggregated treatments within a given level of the sowing density treatment. After two growing seasons, plant density was quantified for all sown species in 15 randomly selected patches from each plot.Important findings I found evidence for shifts in community structure in response to the different spatial seeding patterns. The evenness of added species was higher under aggregated than uniform sowing patterns. There was no detectable effect of aggregated seed sowing on species richness at 3.75 m 2 scale. However, when species richness was extrapolated to larger scales (11.25 m 2), aggregated sowing was predicted to have greater richness than uniform sowing. Effects of seed aggregation on community structure were apparent only at moderate to high sowing rates, yet the latter are within the range of measured seed dispersal in similar grasslands. Additionally, as sowing density increased, seed mass became an increasingly effective predictor of relative abundances for added species, but only under uniform sowing patterns supporting the idea that aggregated dispersal may buffer weaker (smaller seeded) species from competition during colonization. This is the first experiment to show that aggregated seed dispersal patterns can increase at least some components of plant diversity in undisturbed grasslands and suggests that previous seed dispersal experiments, which utilize uniform seed sowing, may underestimate the potential effect of dispersal on plant community structure.     

Emergent interactions influence functional traits and success of dune building ecosystem engineers

Aims Dune building processes are affected by interactions between the growth of ecosystem engineering dune grasses and environmental factors associated with disturbance such as sand burial and sea spray. Research investigating how species interactions influence dune community structure and functional trait responses in high abiotic stress environments is minimal. We investigated how species interactions influence the functional trait responses of three dominant dune grasses to common abiotic stressors.Methods We performed a multi-factorial greenhouse experiment by planting three common dune grasses (Ammophila breviligulata Fern., Uniola paniculata L. and Spartina patens Muhl.) in different interspecific combinations, using sand burial and sea spray as abiotic stressors. Sand burial was applied once at the beginning of the study. Sea spray was applied three times per week using a calibrated spray bottle. Morphological functional trait measurements (leaf elongation, maximum root length, aboveground biomass and belowground biomass) were collected at the end of the study. The experiment continued from May 2015 to August 2015.Important findings Species interactions between A. breviligulata and U. paniculata negatively affected dune building function traits of A. breviligulata, indicating that interactions with U. paniculata could alter dune community structure. Furthermore, A. breviligulata had a negative interaction with S. patens, which decreased S. patens functional trait responses to abiotic stress. When all species occurred together, the interactions among species brought about coexistence of all three species. Our data suggest that species interactions can change traditional functional trait responses of dominant species to abiotic stress.