Growth and competitive interaction between seedlings of an invasive Rumex confertus and of co-occurring two native Rumex species in relation to nutrient availability

Rumex confertus is an alien invasive perennial plant that has increased its range rapidly within central Europe in the last 100 years. This study examined the effects of a commercial fertilizer on the competition between the invasive Rumex confertus and two non-invasive native species R. acetosa or R. conglomeratus in terms of morphological and physiological traits and relative yield. All three Rumex species were grown in the open field with two levels of nutrient availability in field plots. Competition and fertilizer had significant effects on height, relative growth rate (RGR), specific leaf area (SLA) as well as shoot and root biomass of all three species. The fertilized plants had high macronutrient and nitrate contents in leaf tissue. Relative yield of R. confertus was <1, indicating that for this species the effects of interspecific competition were greater than those of intraspecific competition. The results of this experiment indicate that there is interaction between the nutrient status of the soil and the competition between species. Competitive superiority of R. confertus could explain its dominance in grasslands and in disturbed areas, and might explain its great influence on the occurrence of native species because competition intensity was high in fertilized plots.

(1) R. confertus would be the strongest competitor in both unfertilized and fertilized plots and would have higher seedling height and shoot biomass than non-invasive species especially at high nutrient availabilities (2) plants from nutrient fertilized plots would have higher specific leaf area (SLA), leaf macronutrient content than unfertilized ones and (3) under nutrient limitation all three species would have greater biomass allocation to roots.

Soil solution.
Increasing amounts of the fertilizer added to the soil, resulted in enhanced concentrations of N, P and K in the soil solutions (Table 1).
Seedling height. In unfertilized and fertilized plots, the height of both native species was significantly decreased by the invasive species (each P < 0.01). In addition the height of native species in the fertilized monoculture and in the mixture was significantly higher than without the fertilizer (Fig. 1). Analysis of variance showed that the seedling heights of R. confertus, R. conglomeratus and R. acetosa were influenced by competition (for all P < 0.001; Sypplementary Table S1). This analysis also revealed a significant interaction between competition and fertilization (for all P < 0.001; Table 2), i.e. fertilization mediated the effect of competition on the seedling height.
Shoot biomass. Application of the fertilizer significantly increased (for all P < 0.01) the shoot biomass R.
conglomeratus and R. acetosa in both monocultures and mixtures. For example, in monocultures the fertilized seedlings of R. confertus, R. conglomeratus and R. acetosa had respectively 18%, 17.6% and 22% higher shoot biomass compared to the unfertilized ones (for each P < 0.01). In both unfertilized or fertilized monocultures the shoot biomass of R. confertus significantly exceeded that of R. conglomeratus (P = 0.0023), which in turn significantly exceeded that of R. acetosa (P = 0.0121). In the mixture, R. confertus caused a marked reduction in the shoot biomass of R. conglomeratus or R. acetosa (Supplementary Table S1). For example, on the unfertilized plots, R. confertus reduced R. acetosa shoot biomass by 25%, a proportion significantly higher than that caused by R.

Root biomass.
In the unfertilized monoculture the root biomass of R. confertus significantly exceeded that of R. conglomeratus (P = 0.0013), which in turn significantly exceeded that of R. acetosa (P = 0.0191). In all three species root biomass decreased with increasing nutrient supply (R. confertus: P = 0.0019; R. conglomeratus: P = 0.0078; R. acetosa: P = 0.0035). When R. confertus competed with R. acetosa or R. conglomeratus root biomass of the later two was significantly reduced both with and without the fertilizer. However, the effect was more pronounced for R. acetosa than for R. conglomeratus (each P < 0.001; Fig. 1; Supplementary Table S1). A two-way ANOVA showed that fertilization had significant effect on root biomass of the invasive species (P < 0.001) and two native species (each P < 0.001; Table 2).
Root: shoot ratio. All three Rumex species increased biomass partitioning to roots as nutrient supply decreased (Fig. 1). A two-way ANOVA showed that root: shoot ratio was influenced by competition (Supplementary Table S1), fertilization and the interaction between these factors (for all P < 0.01; Table 2).
Specific leaf area. The increased nutrient supply increased SLA. Within mixtures competition with R. confertus significantly reduced SLA of two native species (Fig. 1). Analysis of variance (ANOVA) showed a significant effect of competition and fertilization on SLA of R. confertus, R. conglomeratus and R. acetosa (for all P < 0.01; Supplementary Table S1) with significant interaction between these factors (for all P < 0.01; Table 2).
Relative growth rate (RGR). RGR was influenced by competition and fertilization, with interaction between these factors (for all P < 0.01). Addition of the fertilizer increased RGR of all three species in the monoculture and the mixed treatments. However, in the monoculture R. confertus showed the greatest RGR among all species under both nutrient treatments. Compared with the monoculture, R. confertus reduced RGR of R. conglomertaus and R. acetosa under both unfertilized and fertilized conditions ( Fig. 1; Supplementary Table S1). To summarize, RGR of all species increased with nutrient availability (ANOVA, for all post hoc P < 0.01; Table 2).
Leaf macronutrient (N, P, K) and nitrate (NO 3 − -N) contents. Analysis of variance (ANOVA) of N, P, K or NO 3 − -N contents indicated significant effects of fertilization. In addition, ANOVA showed that in all mixtures, competition had no effect on leaf nutrient content indicating that N, P, K or NO 3 − -N content in R. conglomeratus or R. acetosa when grown with R. confertus were not significantly different from those in monocultures (Supplementary Table S2. However, in the fertilized plots the leaf macronutrient or NO 3 − -N contents of R. confertus plants increased to a greater extent than those of R. acetosa and R. conglomerates plants. For example, in monoculture, leaf NO 3 − -N content of R. confertus was 46% higher in the fertilized plots as compared to the unfertilized ones, whereas in R. acetosa the increase was only 27% ( Fig. 2; Table 3).

Relative yields.
Based on the observed changes in the total biomass of the harvested plants under both nutrient treatments the relative yields of R. acetosa in relation to R. confertus (RY acetosa+confertus ) and of R. conglomeratus to R. confertus (RY conglomeratus+confertus ) were significantly (in both P < 0.01) smaller than 1. So R. confertus outperformed both R. acetosa and R. conglomeratus in competition. Moreover, under unfertilized and fertilized conditions the relative yield of R. conglomeratus in relation to R. acetosa (RY conglomeratus+acetosa ) was significantly higher (P < 0.01) than 1 which implies that R. conglomeratus outperformed R. acetosa in competition (Table 4).

Discussion
The competitive effects of invasive R. confertus on two native species, R. conglomeratus and R. acetosa were examined at different nutrient levels. Competitive ability can be measured in various ways 45 . In the present study, it was evaluated by calculating the relative yield (RY) per plant for each species. As predicted (hypothesis 1) the invasive R. confertus outperformed the native R. conglomeratus and R. acetosa in competition at both low (control treatment) and high nutrient availability. Plant size (plant height) was shown to be highly correlated with competitive ability 46 . For example, Gaudet and Keddy 47 comparing the short-term effects of 44 wetland species on the biomass of a phytometer (a plant on which the competitive effect of a test plant is measured), showed that the competitive ability of the test plants relative to the phytometer was highly correlated with their above-ground www.nature.com/scientificreports www.nature.com/scientificreports/ biomass and associated traits, such as plant height. Although R. confertus was not the tallest at the beginning of the experiment, it overtopped both native species rather early. Thus R. confertus is a strong competitor against its shorter native species, especially R. acetosa, which was suppressed by R. confertus in terms of all traits measured   48 .
In all experiments, RY of R. confertus was <1, indicating that for this species the effects of interspecific competition were greater than those of intraspecific competition 38 . The relative yields of R. acetosa in relation to R. confertus (RY acetosa+confertus ) or of R. conglomeratus in relation to R. confertus (RY conglomeratus+confertus ) on the unfertilized plots were smaller than on the fertilized ones. So, interspecific competition was lower on fertilized plots because RY was closer to 1. According to Grime 2 competition is more intensive at high nutrient levels, which might explain low species diversity in nutrient-rich grassland ecosystems. But, on the other hand Tilman 49 stated that the intensity of competition was either independent of or decreases with soil fertility.
In this study nutrient supply increased growth of all three Rumex species but the effect was most pronounced for the invasive R. confertus. A number of other examples were reported in which the competitive ability of a component species was altered by nutrient availability, for example between invasive grass Phragmites australis and a native competitor Spartina pectinata 36 , between Daucus carota and Chenopodium album 50 , between invasive Hydrocotyle vulgaris and terrestrial species 51 and among nutrient-limited heachland plants 38 .
The shoot biomass of all three species was strongly influenced by competition especially in the high-nutrient treatment. The shoot biomass of native R. acetosa was reduced in competition with both congeners which shows that R. acetosa is the weakest competitor of the three species. The shoot biomass of R. conglomeratus decreased in two nutrient treatments in competition with R. confertus and increased in competition with R. acetosa. This indicates an intermediate position of native R. conglomeratus in the competitive hierarchy within the members of genus Rumex. The shoot biomass of R. confertus was not significantly affected by the congeners, which also indicates this species competitive superiority.
Three species investigated in this study exhibited significant differences in growth attributes. As predicted (hypothesis 2) SLA of the studied species increased with N, P, K supply. In addition invasive R. confertus had higher SLA than its non-invasive congeners, R. conglomeratus and R. acetosa under the same conditions. These results are consistent with those of other studies that reported that high SLA might be associated with invasive species [52][53][54] . Leaves with higher SLA have higher N concentrations 55 , leading to higher respiration 55 , carbon assimilation and RGR 56 . Therefore, high SLA values are characteristic of faster-growing species 57 . Differences in SLA between species might be caused by differences in leaf thickness or leaf tissue composition. For example leaf tissue density will be greater in the leaves with higher concentration of phenolic compounds (lignin and tannins), or secondary metabolites, resulting in lower SLA 57 . Experimental studies showed that RGR may be a better indicator of invasiveness than SLA. For example, SLA was more closely correlated with invasiveness than RGR for Acacia and Acer, while RGR was more tightly associated with invasiveness for the Rosaceae 53 . Invasive species achieved higher RGR than natives primarily by having a higher net rate of dry matter production and/or lower rates of respiration (high NAR, net assimilation ratio) by allocating more biomass to leaves (high LMR, leaf mass ratio), or producing thinner or less dense leaves resulting in more leaf area per unit leaf biomass (high SLA) 58 . As Grime and Hunt 59 showed RGRs of 130 herbaceous species and tree seedlings in the local flora of Sheffield, England, ranged from 0.031 to 0.314 g g −1 d −1 (mean = 0.152). In the present study, mean RGR of R. confertus seedlings in the high-nutrient treatment was 0.35 g g −1 d −1 , thus, it was above of the range of RGRs of the 130 species studied by Grime and Hunt 59 . High seedling RGR under non-limiting resources was found to be the most important trait of six species of dayflowers (Commelinaceae) 60 , four species of Senecio (Asteraceae) 61 and of pine species 53 . A few other comparisons of RGRs of invasive and non-invasive congeners also generally showed that invasive plants had higher RGR than the native ones 29,53,60 . In contrast, Bellingham et al. 62 did not find any correlation between invasiveness and RGR for seedlings of 33 woody species of gymnosperms and angiosperms. In this www.nature.com/scientificreports www.nature.com/scientificreports/ study, at both low and high level of fertility I expected R. confertus to be the superior competitor, with R. conglomeratus second and R. acetosa third. The results were as expected in all the species combinations. This suggests that R. confertus has a higher output per unit input and such a species is able to dominate an area quickly. Higher RGR gives R. confertus competitive advantage allowing it to pre-empt resources at an early stage of the growing season.

Source of variation
Phenotypic plasticity can be broadly defined as the ability of plants to alter their morphology and/or physiology in response to varying environmental conditions it may increase the competitive ability of a plant over a range of different resource availabilities 63,64 . The two common native species, R. acetosa and R. conglomeratus, as well as the invasive species R. confertus allocated relatively more biomass to the roots at low nutrient supply (Fig. 1), possibly increasing their competitive ability for belowground resources. However, both in monocultures or in mixtures the percentage decrease in biomass allocation to the roots in R. confertus exceeded those in both native species (Fig. 1) thus pointing to its higher phenotypic plasticity regarding the partitioning of biomass between shoots and roots. Similar pattern was found by Aerts et al. 38 in evergreen shrubs, Erica tetralix and Calluna vulgaris and perennial grass, Molinia caerulea. High plasticity of R. confertus may have significant effects on seedling establishment and competitive ability, and hence on species distribution 65 .
Rumex confertus showed significantly higher values of N, P, K or NO 3 − -N content, than the other Species. Nazaryuk et al. 66 showed that the nitrate accumulation in plants depended on three major groups of factors: the amount and kind of a fertilizer applied, treatment with physiologically active substances, and natural and www.nature.com/scientificreports www.nature.com/scientificreports/ anthropogenic changes in the soil environment. All these factors may be arranged in the following order: fertilizers > physiologically active substances > soil. Nitrate accumulation in plants increased with nitrogen supply, whereas limiting the nitrogen availability reduced nitrate content significantly (see Umar and Iqbar 67 for review). Both net absorption and assimilation rates during the growing season are genetically determined, explaining a large variability of plant nitrate content among plant species, and even among cultivars of the same species (see Cárdenas-Navarro et al. 68 and references therein). This study showed that the nutrient content levels in the leaves were within the range of the reference values known for R. crispus 69 and for R. alpinus 70 . Changes in the biomass of shoots and roots caused by species competition, did not translate into changes of the amounts of N, P, K or NO 3 − -N stored in the leaves. This is consistent with Zaller 71 results, who tested the competitive ability of R. obtusifolius against grassland species.
Differences between species in root biomass are associated with environmental conditions. Boot and Mensink 72 found that species from fertile sites had higher root: shoot ratios at both low and high N supply than did species from infertile sites. Others authors reported that fast-growing species from fertile sites generally had higher capacity to adjust mass allocation (see Hill et al. 73 for the list and references). There are reports concerning many higher plant species showing decreased root: shoot ratio correlated with increased growth due to higher nutrient supply 74 . These results are consistent with the results presented here because the tested species were generally able to modify their root systems in response to nutrient. The hypothesis 3 in the studied species, the relative amount of biomass allocated to the roots was lower in the fertilized variant was confirmed. These findings are similar to those reported by McConnaughay and Coleman 74 for the weed species Abutilon theophrasti, Chenopodium album and Polygonum pensylvanicum.
In central Poland, damp meadows on which R. confertus grows are often located below the fields and mineral fertilizers from them flow with water to these neighboring meadows. Thus its success may result from positive response to high nutrient availability. This is consistent with the pattern found in P. australis by Rickey and Anderson 36 .

Conclusion
The high invasiveness of R. confertus seems to result from its competitive dominance over the other congeners across a very wide range of nutrient availability. The high competitive ability of R. confertus with respect to both R. acetosa and R. conglomeratus was caused by (a) higher SLA, (b) great plant height, and (c) higher potential growth rate. This combination of plant traits reduces light absorption or photosynthate production of R. acetosa and R. conglomeratus thereby limiting their growth.   www.nature.com/scientificreports www.nature.com/scientificreports/ With strengthened competitive ability of R. confertus, weakened competitive ability in the other species (R. acetosa and R. conglomeratus) and consequent changes in competitive hierarchy, I expect possible changes in species dominance and structure of meadow plant community in central Europe in the future.

Materials and Methods
Study species. The studied Rumex (Polygonaceae) species are biennials with similar life-histories and reproductive characteristics, they coexist in some habitats but with different origin and invasion status in Poland. Rumex acetosa, R. conglomeratus and R. confertus were chosen to test interspecific competition because these species are the most common neighbors in moist to wet meadows and pastures in Poland. All three studied Rumex species reproduce clonally (ramets produced from root and rhizome buds) and sexually from seeds.
According to their ecological indicator values 75 , R. acetosa (Ellenberg's N-value = 6) and R. conglomeratus (Ellenberg's N-value = 8) show similar requirements for soil nutrients. Generally, both species are indicators of high soil nitrogen concentrations 75 .
Rumex confertus is native to temperate Eastern Europe and Asia where it thrives on meadow-steppes and glades in forest-steppe. Its height varies from 60 to 120 cm. In Poland it occurs predominantly along rivers and invades seminatural vegetation (e.g. meadows, wet ditches, riparian-scrub), but has been recently colonizing disturbed habitats (such as roadsides, railway tracks and embankments), forest clearings and margins 76,77 . The aggressiveness of R. confertus results from its ability to establish quickly from seed, to flower in its first year and to its fast growth and high seed production some of which can remain viable for very long periods in the soil 77 .
Rumex conglomeratus grows in wet meadows, stream and river banks, ditches, field margins and gateways, often in places flooded or waterlogged in winter 78 . This species is characterized by a height similar to that of R. confertus and a broad ecological amplitude. This Eurosiberian Southern-temperate element became widely naturalised so that distribution is now Circumpolar Southern-temperate 78 .
The third species, R. acetosa occurs in a wide range of habitats, but particularly at waste ground, road sides, disturbed areas, grasslands (mainly pastures), and arable land 78  Experimental design and growth conditions. An experiment was conducted in the experimental garden on a private property in the vicinity of Uniejów, 60 km west of Lodz, where the mean annual temperature was 8.8 °C and mean annual precipitation (rain and snow) was 587.2 mm (meteorological data, Lodz station). The seeds were stratified in a refrigerator at 4 °C for a 16 weeks and then germinated on moist filter paper in a Petri dish in growth cabinet (16 h light, 24 °C, 20 μmol m −2 s −1 PPFD; 8 h dark, 10 °C). After three weeks, uniform seedlings with two leaves were selected and transplanted into plastic pots 25 cm in diameter, 22 cm deep (1962 cm 2 ) filled with a mixture (1: 4, v/v) of sterilized soil and sterilized sand and cultured in a growth room (23 °C; 14 light/10 h dark; 80 μmol m −2 s −1 PPFD, Philips TL 94,) for three months.
Next the same size seedlings of individual species were transferred to field plots, 22 days after germination ([T 1 ], 17 April 2017). Seedling biomass (±SD) at the start of the experiment (T 1 ) was equal for R. confertus (0.41 ± 0.05 g) and R. conglomeratus (0.39 ± 0.04 g) while R. acetosa seedlings had smaller mass (0.28 ± 0.02 g). After planting, the seedlings were grown in the open air during May−August a period of moderate temperature ( Table 5).
The plants were planted in monocultures and mixtures using a replacement design: each species in monoculture and also in a 1:1 mixture of each of two species. All treatments were carried out with four replicate plots in a randomized block design. A total number of 48 plots (2 nutrient treatments × 6 species-competition combinations × 4 replicates) were used for the experiment.
To avoid border effects, only the inner 30 individuals of each plot (6 rows of 5 plants in) in both monocultures or in mixtures were analyzed (Fig. 3).
Two nutrient treatments for each species-competition combination were imposed after transplanting the seedlings into plots: unfertilized (control) and fertilized. The commercial (Inco VERITAS S.A., Poland) fertilizer "Azofoska" (40 g m −2 yr −1 ) containing (%): 5  www.nature.com/scientificreports www.nature.com/scientificreports/ The plots (1 × 1 m) were separated by equally sized (0.5 m wide) buffer zones. There were 8 rows of 9 seedlings in both monocultures and mixture plots. However, plant density of each species in the mixture was half of the density of that species in monoculture. In the plots containing mixtures of seedlings, 36 seedlings of each of the two species were placed alternately in rows at the spacing of 10 cm. Thus each seedling (except at the margin) was surrounded by four seedlings of the other species. To avoid border effects, only the inner 30 individuals of each plot (6 rows of 5 plants: both in monocultures and mixtures) were analyzed (Fig. 3). Watering to field capacity, using tap-water, was carried out every three days.
Soil physical and chemical analysis. In each plot one soil core (3 cm diameter × 5 cm deep) was collected at random and put in a plastic bag. Soil cores were mixed by hand up into a single bulk sample for each plot. Then the soil samples were air-dried (20-25 °C for 3 days) to a constant weight and sieved through a 2-mm mesh screen. The pH was determined in water (ratio 1:1, soil: water). The soil samples (100 g dry soil) were analysed for selected macronutrients: total nitrogen (N), phosphorus (P) and potassium (K). Phosphorus was measured by colorimetric molybdenum blue method in 0.5 M sodium bicarbonate (NaHCO 3 ) extract of soil. Potassium was extracted from soil with calcium lactate (C 6 H 10 CaO 6 ) and determined by atomic absorption spectrometer (AAS) method. Total N was determined by the micro-Kjeldahl procedures by the wet oxidation of organic matter using sulfuric acid (H 2 SO 4 ) and a digestion catalyst. The amount of soil organic matter was estimated as loss on ignition (550 °C, 3 h) and calculated as % dry weight ( Table 2). The analyses were performed at the Chemical-Agricultural Station in Lodz.
Harvest and biomass sampling. At the end of the experiment ([T 2 ]; 17 August 2017, 16 weeks following transplantation) height (cm) of the main shoot of each seedling was recorded before they were harvested. In general, ten whole plants of each species-competition combination were randomly collected within both unfertilized and fertilized plots. A total number of 240 seedlings per species (2 nutrient treatments × 3 species combination × 4 replicates × 10 seedlings) were collected for the experiment. To measure shoot biomass, each species shoot (leaves + stem) was cut off at the soil surface. Roots and shoot were carefully washed in the lab. Leaf area was measured on the same day with a CI-202 portable laser leaf area meter (CID Bio-Science, USA). After determination of the area, the leaves were dried for at least 48 h at 70 °C and dry mass was determined. Specific leaf area (SLA; cm 2 g −1 ) was calculated as area per unit mass. Shoots and roots of a single plant were dried to a constant weight for at least 72 h at 70 °C in an oven, and weighed (g). The relative growth rate (RGR) was estimated using the following formula: where, W t2 W t1 are plant dry weight (DW in mg) of seedlings at time T 1 and time T 2 , respectively 81 . www.nature.com/scientificreports www.nature.com/scientificreports/ Relative yield. The effect of competitive response (neighbours) was quantified using relative yield per plant for each species 82 . The relative yield (RY ab ) was calculated as: where, Y ab is the dry biomass yield of species a in mixture with species b and Y a is the dry biomass yield of species a in monoculture. If RY = 1, there is absolutely no competition. If RY ab > 1 species a outperformed species b in competition, when RY ab < 1 the reverse is true 38 .
Foliar macronutrient and nitrate analyses. Samples of three (youngest fully expanded) leaves were collected from each of 10 different plants separately for each competition treatment. Pooling of three leaves on a plant was necessary to obtain enough tissue for the macronutrient (N, P and K) and nitrate (NO 3 − -N) analyses. Afterwards, the leaf samples were dried (105 °C for 5 h) and ground with a plant-sample mill and siewed with a 2-mm mesh screen before the sub-samples were weighed with a precision balance. Prior to chemical analyses, the leaf samples of each harvest were combined into four independent sub-samples (100 mg of each). Two thirds being used for macronutrient determination and one third − for nitrate determination. The leaf material was mineralized in a boiling mixture of 10 ml of HNO 3 and 5 ml of H 2 SO 4 . Phosphorus was determined colorimetrically (see soil properties), potassium by atomic absorption spectrometry (as above), and the total nitrogen content by the micro-Kjeldahl method (as above). Nitrate concentration in plant tissue was determined with a Corning NO 3 − ion selective electrode (Orion Analyzer Application Bull. No. 7, Determination of nitrate in plant tissue). Leaf macronutrients and tissue NO 3 − -N contents were as calculated as the total amount of macronutrients and nitrate per unit of dry leaf mass (mg/g dry wt −1 ).

Data analysis.
Normality was verified with the Shapiro-Wilk test. All data were log transformed prior to the analysis if necessary to meet the assumption of normality. Student's t-test was performed to test for statistical differences in relative yields (RY ab ) between unfertilized and fertilized plots. For unfertilized and fertilized plots, differences between species in seedling height, shoot biomass, root biomass, SLA and RGR were analysed by one-way ANOVA with post-hoc Tukey's range tests. The effect of competition (monoculture vs mixture plots), fertilization (control vs fertilized) and their interaction on seedlings traits (plant height, shoot and root biomass) or relative growth rate (RGR) for each species was analyzed using a two-way nested ANOVA. Two-way ANOVA was also performed to test the effect of competition (monoculture vs mixture plots), fertilizatiom (control vs fertilized) and their interaction on concerning macronutrients (N, P, K) or nitrate (NO 3 − -N) concentrations in leaves of three species. Statistical analyses were performed using the Statistica 13.0 package 83 .

Data Availability
All data generated or analysed during this study are included in this published article (and its Supplementary Information File).