Copper Uptake and Its Effects on Two Riparian Plant Species, the Native Urtica dioica, and the Invasive Fallopia japonica

Copper accumulating in stream sediments can be transported to adjacent riparian habitats by flooding. Although being an essential element for plants, copper is toxic at high concentrations and restricts, among other things, plant growth. Besides copper, invasive plants, such as Fallopia japonica, which are known to be tolerant toward heavy metals, modify riparian habitats. If the tolerance of F. japonica is higher compared to native plants, this could accelerate invasion under high heavy metal stress. Therefore, we aimed to compare the effect of copper on two common riparian plants, the invasive F. japonica and the native Urtica dioica. We performed a pot experiment with a gradient from 0 to 2430 mg kg−1 of soil copper. We hypothesized that (i) negative effects on plant growth increase with increasing soil copper concentrations with F. japonica being less affected and (ii) accumulating higher amounts of copper in plant tissues compared to U. dioica. In support of our first hypothesis, growth (height, leaf number) and biomass (above- and belowground) of F. japonica were impacted at the 810 mg kg−1 treatment, while the growth of U. dioica was already impacted at 270 mg kg−1. Due to 100% mortality of plants, the 2430 mg kg−1 treatment was omitted from the analysis. In contrast, chlorophyll content slightly increased with increasing copper treatment for both species. While U. dioica accumulated more copper in total, the copper uptake by F. japonica increased more strongly after exposure compared to the control. In the 810 mg kg−1 treatment, copper concentrations in F. japonica were up to 2238% higher than in the control but only up to 634% higher in U. dioica. Our results indicate that F. japonica might be able to more efficiently detoxify internal copper concentrations controlling heavy metal effects compared to the native species. This could give F. japonica a competitive advantage particularly in polluted areas, facilitating its invasion success.


Introduction
Since the 19th century, the use of copper as a fungicide has been an inherent part of agriculture [1]. Copper does not degrade but accumulates in soils leading to elevated concentrations [2]. In fact, the long-term application of copper resulted in high levelsreaching up to 3000 mg kg −1 -in agricultural areas [3]. In European vineyards, for example, soil copper levels of up to 600 mg kg −1 have been observed, while corresponding uncontaminated sites ranged between 20 and 100 mg kg −1 [4]. From agricultural soils, copper can be transported to adjacent non-target aquatic ecosystems through spray drift, groundwater drainage, and surface run-off [5]. In rivers and streams, copper tends to accumulate in sediments [6] where it can be remobilized and transferred to riparian ecosystems during

Results
Since no plants survived in the highest copper treatment (2430 mg copper kg −1 soil) (see Table S1), this treatment was disregarded for further analysis. In the text, all mentioned values will be mean and standard deviation (see Section 4.4 for more details).

Seasonal Plant Growth
Plant height and leaf number significantly decreased with increasing copper concentrations in both Fallopia japonica and Urtica dioica. In the control treatment, F. japonica (26.8 ± 12.9 cm) was about 25% higher than in the treatment with 810 mg kg −1 (21.4 ± 12.6 cm), which was statistically significant (Table 1). Similarly, in U. dioica the average height was significantly higher in the control (35.7 ± 22.9 cm) than at 810 mg copper kg −1 (30.0 ± 22.3), a reduction by about 20%. Regarding leaf number, F. japonica had an average of 56.2 ± 37.4 leaves in the control, a value significantly lower by 25% (43.0 ± 26.0) in the 810 mg copper kg −1 treatment. For U. dioica, a significant reduction in leaf number from 140.0 ± 131.9 to 73.6 ± 72.2 (by approximately 50%) in the same treatments was also observed (Table 1). Overall, the decrease in plant height and leaf number became apparent at 810 mg copper kg −1 for F. japonica while U. dioica showed significant responses already at 270 mg copper kg −1 (Figure 1).
In contrast, the chlorophyll content increased statistically significantly with increasing copper concentration (Table 1). For F. japonica, chlorophyll increased by about 10% from 21.1 ± 5.7 in the control to 23.1 ± 6.0 in the 810 mg copper kg −1 treatment. With an increase in chlorophyll from 27.3 ± 7.2 to 30.6 ± 13.1, the effect was comparable for U. dioica. Table 1. Results of the lmer models for both Fallopia japonica and Urtica dioica performance parameters across the growing season of 2021. Chi-squared and corresponding p-values for all analyzed parameters are given. Treatment represents the effect of the copper treatment; time represents the effect of the time over the season when the parameter was recorded; and treatment * time represents the interaction between both parameters. Significant values below the level of 0.05 are printed in bold. Apart from treatment effects, the factor time also significantly influenced all plant parameters. Height and leaf number of both species increased during spring but remained rather constant thereafter ( Figure 1). In contrast, the chlorophyll content significantly decreased over time independent of the plant species (Table 1). Both copper treatment and time significantly interacted for several variables in both species (Table 1). At lower copper concentrations, leaf number of F. japonica increased more strongly over time while higher treatment levels showed a decrease in the later part of the season. Height, a parameter describing growth over time, of U. dioica was lower at higher copper concentrations while leaf numbers decreased more strongly later in the growing season ( Figure 1).

Biomass after a Two-Year Experiment
The final biomass of F. japonica for both above-and belowground plant parts significantly decreased with higher copper concentrations, which became statistically significant at 810 mg copper kg −1 ( Table 2). Above-and belowground biomass significantly decreased by a factor of two from 72.1 ± 24 g to 32.5 ± 28 g and 228.2 ± 98.5 g to 114.6 ± 126 g, respectively. At lower copper concentrations, biomass remained largely constant relative to the control (Figure 2a,c). Table 2. Results of the lmer models for both Fallopia japonica and Urtica dioica above-and belowground plant biomass collected at the end of the second growing season. Chi-squared and corresponding p-value for each analyzed parameter are given. Treatment represents the effect of the copper treatment. Significant values below the level of 0.05 are printed in bold.  For U. dioica, there was a tendency for a decrease in above-and belowground biomass with increasing copper concentrations but no significant effect overall ( Table 2). Above-and belowground biomass decreased by a factor of two and three respectively from 29.9 ± 19.1 g to 12.6 ± 11.3 g and 54.3 ± 30.1 g to 17 ± 18.8 g. The statistical non-significance of the copper treatments in U. dioica may be explained by a rather low biomass in the control, with the lowest copper treatment of 90 mg kg −1 having higher biomass than the control (Figure 2b,d).

Copper Content of Plant Tissues
The copper content of almost all plant tissues was higher in U. dioica compared to F. japonica across the experimental soil copper gradient. The only exception was the root copper content, which was 46% higher in F. japonica at the 810 mg kg −1 level (17.1 ± 9.5 mg g −1 in F. japonica and 11.7 ± 2.1 mg g −1 in U. dioica). Especially for the aboveground and rhizome material, the samples from U. dioica showed much higher copper contents compared to F. japonica. In the 810 mg kg −1 treatment, U. dioica aboveground material contained about 2.5 times more copper compared to F. japonica (9.7 ± 5.9 mg g −1 in F. japonica and 24 ± 11.9 mg g −1 in U. dioica) but aboveground copper content was higher in U. dioica across all treatment levels (Figure 3a,b). Copper content of the three sample types (aboveground, roots, and rhizome) significantly increased with increasing soil copper concentration in both F. japonica and U. dioica (Table 3). Like the absolute concentrations, the strongest increase was found in the rhizome where the mean copper content increased from 1.3 ± 1.0 mg g −1 in the control to 30.4 ± 31.9 mg g −1 in the 810 mg kg −1 level for F. japonica and from 9.4 ± 13.5 mg g −1 to 69.0 ± 33.9 mg g −1 in U. dioica (Figure 3e,f). Unlike the absolute concentrations, F. japonica showed a much higher accumulation compared to the control, increasing copper concentrations in rhizomes by more than 23.3 times compared to 7.3 times in U. dioica. Accumulation in the 810 mg kg −1 treatment compared to the control for the two other plant parts was also higher in F. japonica compared to U. dioica. For aboveground samples, copper content increased by 19 times in F. japonica vs. 17 times in U. dioica and for the roots, the difference was 3.5 times compared to 2 times. Table 3. Results of the lmer models for both Fallopia japonica and Urtica dioica plant copper content separated by aboveground, root and rhizome samples collected at the end of the second growing season. Chi-squared and corresponding p-value for each analyzed parameter are given. Treatment represents the effect of the copper treatment. Significant values below the level of 0.05 are printed in bold. In F. japonica, the highest copper content in tissues was found in the rhizome, with a two-fold higher copper concentration at the highest treatment level compared to root samples (30.4 ± 31.9 mg g −1 compared to 17.1 ± 9.5 mg g −1 ) and a three-fold higher concentration compared to aboveground samples (30.4 ± 31.9 mg g −1 compared to 9.7 ± 5.9 mg g −1 ). The rhizome copper concentration increased by 2238% from the control (1.3 ± 1 mg g −1 ) to the 810 mg kg −1 treatment. For U. dioica, the average copper content was also highest in rhizomes (six times higher than roots and three times higher than aboveground material at the highest treatment; Figure 3). For U. dioica, the increase from control (9.4 ± 13.5 mg g −1 ) to the highest treatment was 634%.

Discussion
We investigated the effect of copper contamination (treatments with 0, 90, 270, and 810 mg kg −1 ) on the invasive plant Fallopia japonica and the native Urtica dioica. Higher soil copper concentrations negatively impacted the growth of both the invasive plant species Fallopia japonica and the native Urtica dioica. However, the native species was more sensitive relative to the invasive species. Uptake of copper by plant tissue increased with soil copper content for both plants. Unexpectedly, U. dioica showed higher absolute content of copper in plant material compared to F. japonica. However, the accumulation of copper compared to the control was higher in F. japonica than in U. dioica, which corresponds to the hypothesis that F. japonica can accumulate copper more effectively. As expected, increasing copper concentrations impacted all observed plant parameters (above-and belowground biomass, plant height, leaf number and chlorophyll content) in both plant species. We also found that the growth of F. japonica and U. dioica depended on the time of measurement. Plant height and leaf number increased during the growing season due to seasonal progression in plant growth. Regarding the copper treatment, we found that biomass, plant height, and leaf number decreased with increasing copper concentration. This reduction in plant growth is in line with the previous studies on various species (Buxus sempervirens, Cotoneaster divaricatus and Rhododendron obtusum [13]; Phaseolus vulgaris [14]; Prunus cerasifera [12]) and reflects the general known responses of plants to copper [9]. The impact of copper on plant growth can be explained by oxidative damages caused by excess copper in cells, which damage cell organelles and DNA [9][10][11]. This leads to reduced growth and necrosis in affected plants [12][13][14].
Importantly, we found that the reduction in growth occurred at different threshold concentrations: For F. japonica, growth was only lower at the highest analyzed concentration (810 mg kg −1 ), while U. dioica already showed significant loss of growth at 270 mg kg −1 .
Our results therefore only partly confirm the few existing studies on Fallopia species, where negative effects of copper started at lower concentrations than in our study. For example, Sołtysiak [23] found impact on F. japonica at 300 mg copper kg −1 and Pedersen et al. [20] estimated a threshold for the onset of adverse effects on Fallopia convolvulus growth at slightly lower concentrations (i.e., 200 mg kg −1 ). In our study, plants grown at 270 mg copper kg −1 had similar biomass compared to the lower treatments. Although we observed the highest above-and belowground biomass in U. dioica at 90 mg copper kg −1 and a subsequent decrease toward the higher concentrations, the overall effect of copper was not significant, likely due to the high variability in our data. The stimulation of above-and belowground biomass in U. dioica corresponds well to the higher leaf numbers and height in the same treatment. In earlier studies, U. dioica has shown the ability to tolerate higher concentrations of copper with phytostabilisation of contaminated soil, meaning a reduction in the bioavailability of copper in the soil [26]. The pattern observed in our study might be due to hormesis, where a mild stress, here induced by copper, can stimulate plant growth, which might be linked to the production of antioxidants in combination with a stimulation of chlorophyll levels [29]. This effect is often observed in hyperaccumulator plants [30]. A possible reason for the stronger growth at high copper concentrations in F. japonica might be its higher tolerance to copper [19], which is thought to be connected to the plant's ability to sequester copper into the cell wall [31] and produce specific binding proteins even at high metal concentrations [32]. This also allows F. japonica to hyperaccumulate metals [33,34]. Our findings therefore support the hypothesis that invasive plants benefit from disturbances by heavy metal contamination due to their higher tolerance to stressors such as copper [18].
In contrast, the chlorophyll content increased with copper concentration in both species and decreased along the time of measurement, since chlorophyll production is reduced in order to focus on production of flowers and seeds later in the year. According to Kalaji et al. [35], the effect of heavy metal contamination on the number of photosynthetic pigments is not uniform. For example, lead caused a decrease in the amount of pigment in Pisum sativum [36]. Similarly, chromium contamination decreased total chlorophyll by up to 61% in Brassica juncea [37] and by 54% in Solanum lycopersicum [38]. On the other hand, cadmium had no impact on the amount of photosynthetic pigments in Brassica napus but still disrupted photosynthetic processes [35]. However, in mosses (Ptychanthus striatus, Thuidium delicatulum, and T. sparsifolium) as well as in Empetrum nigrum and Camellia sinensis, copper showed a stronger inhibitory effect on total chlorophyll than other heavy metals, due to the disruption of chlorophyll production [39][40][41]. The contrary pattern observed in our study could be explained by substantial differences in plant traits. Both F. japonica and U. dioica are known to accumulate heavy metals [20,42] and might therefore be efficient in detoxifying copper before it can damage chlorophyll production. This might be due to chelating agents such as oxalic acid which can bind heavy metals and reduce toxicity. This has been shown for chromium toxicity [37,43]. Another factor might be a higher concentration of the plant hormone salicylic acid in these species, which can increase the stress tolerance of plants. Application of salicylic acid can decrease the toxicity of heavy metals such as chromium and aluminum [38,44]. A further explanation might be that the observed loss of leaves in higher copper treatments caused an increase in chlorophyll production in the remaining leaves to keep the photosynthetic activity of the organism stable. This form of compensation has previously been found as a response to herbivory [45,46].

Copper Content in Plant Tissue Is Higher in U. dioica Than in F. japonica
It could be verified that both F. japonica and U. dioica accumulate copper, which is in line with the studies dealing with roots and rhizomes [21,28,31,42]. We also found that the copper content was highest in the belowground plant parts, especially rhizomes. This observation contrasts with literature highlighting higher heavy metal concentrations in aboveground plant parts (mainly leaves) for F. japonica [23] and U. dioica [26,28]. Moreover, our study points to a lower copper concentration in the plant relative to the surrounding environment (i.e., soil) for both species, suggesting they are not hyperaccumulators of copper, which was postulated for F. japonica [47] and Fallopia × bohemica [48]. However, copper hyper accumulation was reported for roots of F. convolvulus with a factor 2:1 for root: soil [20].
Unexpectedly, the absolute copper concentrations in F. japonica were lower in nearly all copper treatments and plant parts than in U. dioica. This difference might be explained by F. japonica being able to regulate the copper uptake compared to U. dioica. Lerch et al. [49] showed in a pot experiment that the transfer of copper from soil to plant does not increase strongly under copper contamination. While copper in roots of F. convolvulus increased linearly with increasing soil copper concentration, a saturation in aboveground plant parts was reported for 200 to 300 mg kg −1 [20]. Only in our treatment with a copper concentration of 810 mg kg −1 , less copper was detectable in roots of U. dioica compared to F. japonica (Figure 3). Since saturation in copper uptake was observed for F. convolvulus only for aboveground plant parts [20], this could also provide an explanation for the observation of increased copper content in roots of F. japonica. Furthermore Lu et al. [50] demonstrated a higher concentration of Fe, Al, and Cu in roots compared to rhizomes, stems, and leaves in F. sachalinensis. This accumulation in roots of Fallopia spp. is consistent with both our hypothesis and literature where F. japonica was found to have a high potential to accumulate heavy metals under field conditions and elevated heavy metal concentrations [21][22][23]50]. Metal accumulation is linked to metal tolerance and is common in plants found in metalliferous habitats [51]. The higher metal tolerance in F. japonica found in this study might therefore be linked to this ability.

Experimental Setup
The mesocosm experiment was set up on an open field located in Siebeldingen at the Julius Kühn-Institut (Federal Research Centre for Cultivated Plants), Rhineland-Palatinate, Germany (49.219643, 8.049299). Pots were established in April 2020 and the experiment lasted until September 2021 for two full growing seasons. In 2020, the average temperature was 11.9 • C and the total precipitation was 630 mm, while in 2021, the average temperature was 10.4 • C and the total precipitation was 814 mm [52].
In April 2020, 100 pots with a diameter of 63 cm were filled with 5 cm of gravel at the bottom as a root drainage and a mixture of 70 liters of sand originating from locally sourced Triassic sandstone (BVG, Albersweiler, Germany) and 20 liters of natural riparian soil. The riparian soil (topsoil 0-25 cm) including its natural microbial community was collected in March 2020 from a nearby unpolluted riparian area within the biosphere reserve Palatinate Forest (Dürrentalbach, 49.255048, 7.939021). Before mixing, the soil was sieved and visible plant parts were removed. While filling the pots with soil, 10 g of nitrogen fertilizer, N-P-K: 6-6-9 (ORGASAN, Cuxin DCM, Telgte, Germany) and 9 g of trace element fertilizer (Micromax Premium, Everris International, de Heerlen, The Netherlands) were added to each pot. The fertilization was repeated in the second growing season in April 2021. The final average soil pH was 7.5.
The pots were arranged in a two-factorial randomized block design (Figure 4). Plant species (F. japonica and U. dioica) and soil copper concentration were used as treatment factors. Five levels of CuSO 4 (Centrum Metal Odczynniki Chemiczne, Falenty, Poland) in 5 liters of tap water solution were applied and mixed with the soil. This compound was chosen since it is a common fungicide utilized in agriculture. The four nominal concentrations were 0, 90, 270, 810, and 2430 mg Cu kg −1 soil. Each treatment combination was replicated ten times.
In April 2020, 100 pots with a diameter of 63 cm were filled with 5 cm of gravel at the bottom as a root drainage and a mixture of 70 liters of sand originating from locally sourced Triassic sandstone (BVG, Albersweiler, Germany) and 20 liters of natural riparian soil. The riparian soil (topsoil 0-25 cm) including its natural microbial community was collected in March 2020 from a nearby unpolluted riparian area within the biosphere reserve Palatinate Forest (Dürrentalbach, 49.255048, 7.939021). Before mixing, the soil was sieved and visible plant parts were removed. While filling the pots with soil, 10 g of nitrogen fertilizer, N-P-K: 6-6-9 (ORGASAN, Cuxin DCM, Telgte, Germany) and 9 g of trace element fertilizer (Micromax Premium, Everris International, de Heerlen, The Netherlands) were added to each pot. The fertilization was repeated in the second growing season in April 2021. The final average soil pH was 7.5.
The pots were arranged in a two-factorial randomized block design (Figure 4). Plant species (F. japonica and U. dioica) and soil copper concentration were used as treatment factors. Five levels of CuSO4 (Centrum Metal Odczynniki Chemiczne, Falenty, Poland) in 5 liters of tap water solution were applied and mixed with the soil. This compound was chosen since it is a common fungicide utilized in agriculture. The four nominal concentrations were 0, 90, 270, 810, and 2430 mg Cu kg −1 soil. Each treatment combination was replicated ten times.  Nominal copper concentrations were verified using inductively coupled plasmaoptical emission spectrometry (ICP-OES) (Agilent 720 ICP OES, Agilent Technologies, Santa Clara, CA, USA), following digestions with aqua regia [53,54] directly after test initiation (April 2020) and at its termination (August 2021). Measurements suggest adequate spiking with deviations relative to nominal concentrations not exceeding 15%, justifying the use of the latter throughout the present document (see Table 4). The full results of the ICP-OES measurements can be found in Table S2. Table 4. Total copper found in the soil in mg kg −1 in relation to the desired treatment. See Table S2 for the corresponding data. Plants were introduced into the pots in June 2020 as rhizomes with one aboveground shoot. Rhizomes of both species originated from a riparian zone at the river Queich in Landau (49.198357, 8.096627). This stand of F. japonica was selected since it is genetically pure without the hybrid influence of F. sachalinensis (F. Schmidt ex Maxim.) [55]. Rhizomes were washed and cut to approximately 5 cm, making sure that at least one node was included for regeneration. Four rhizomes of the invasive or the native species, respectively, were evenly distributed in each pot.
During the growing season, the plants were watered depending on the weather conditions. Under hot and dry conditions, each pot received four liters of tap water at a maximum rate of twice a week. After precipitation events no water was provided. Randomly growing other plant species were weeded every week to minimize their impact. During the first season, the test system equilibrated and plant-related measurements started in spring of the second season. Pots without plant growth were excluded from analyses (see Table S1 for an overview of plant mortality in the experiment).

Plant Parameter Measurements
Plant parameters were measured bi-weekly covering ten timepoints between April and August 2021 (Table S3). These parameters included plant height, leaf number, and chlorophyll content. Plant height (cm) was defined as size from the soil surface to the extended tip of the highest stem using a measuring stick. The number of leaves were counted. Chlorophyll content, a proxy for photosynthetic activity [56], was measured using a SPAD-502 device (Konica Minolta, Tokyo, Japan) on three randomly chosen healthy leaves per plant. Measurements with the SPAD-502 produce relative values that are proportional to the amount of chlorophyll present in the leaf [57]. The average of these measurements for each pot entered further statistical analyses. In September 2021, plants were harvested individually, separating above-and belowground biomass (Table S4). All plant material was dried for 4 days at 60 • C. The dry biomass was then quantified using a KERN ADJ 200-4 scale (linearity ± 0.4 mg, Kern & Sohn, Balingen, Germany). In the subsequent statistical analyses, total biomass per pot separated into above-and belowground biomass was used.

Plant Copper Content
Dried plant material of aboveground parts (stems and leaves), roots, and rhizomes was ground into a powder separately. For each plant species and the three plant parts selected, five samples of each copper treatment were prepared, resulting in a total of 120 samples. Plant copper content was measured using ICP-OES (Agilent 720 ICP OES, Agilent Technologies, Santa Clara, CA, USA). Measurement of copper was done at a wavelength of 327.395 nm using the same method as the soil copper measurement [53,54]. Prior to analyses, samples were digested in 10 mL aqua regia overnight and afterwards heated to 175 • C for 15 min in a microwave (Mars 5 System, CEM Corporation, Matthews, NC, USA). The samples remained in the microwave at 175 • C for another hour. Aliquots were subsequently diluted to 1:10 or 1:20 with MiliQ water depending on the limits of quantification supporting a reliable quantification of copper. The measured copper content in the plant parts can be found in Table S5.

Data Analysis
R version 4.0.5 [58] was used for statistical analyses. Linear mixed-effects models (lmer) were created separately for the two plant species (F. japonica and U. dioica) using the package "lme4" [59]. The highest copper treatment (2430 mg kg −1 soil) was disregarded for further analysis because no plants survived (Table S1). The models for each dependent variable (above-and belowground final biomass, plant height, leaf number, chlorophyll content, plant copper content) used the copper levels (factor with the four levels "0", "90", "270", "810" mg kg −1 ) as a fixed effect. In the models for plant height, leaf number, and chlorophyll content, time (factor with 10 time points over the growing season) and the interaction of time and copper treatment were included as additional fixed effects. In all models, the random effect "block" was used to account for the block design and possible variability introduced by the position on the experimental site. Additionally, the random effect "stem number", representing the total number of plant individuals found in each pot, was used to account for differences in growth caused by the uneven number of plant individuals that survived until the start of measurement in 2021. Significance levels were calculated using the ANOVA command from the "car" package [60].

Conclusions
Overall, we found that soil copper contamination inhibited the growth in terms of leaf number, height, and biomass in both Fallopia japonica and Urtica dioica pointing to a higher sensitivity of U. dioica (significant effects starting with 270 mg copper kg −1 ) compared to F. japonica (significant effects starting at 810 mg copper kg −1 ). On the other hand, F. japonica contained far lower absolute concentrations of copper in its above-and belowground parts compared to U. dioica but was able to accumulate copper more effectively when exposed to it. In this way, F. japonica may gain a selective advantage against this commonly occurring native species in its invasive range and especially in polluted areas because of its higher tolerance to copper. Our study demonstrated that disturbances by agricultural activity, one major source of copper pollution, could have an impact across much wider distances than expected by facilitating plant invasion success of F. japonica in previously undisturbed areas.