Three perennial, submerged, and invasive aquatic plant species threatening Central Europe were selected to assess their growth responses to low light and low temperature conditions: Cabomba caroliniana A. Gray (carolina fanwort), Elodea nuttallii (Planch.) H. St. John (Nuttall’s waterweed), and Vallisneria spiralis L. (eelweed). E. nuttallii is native to North America, while C. caroliniana occurs naturally in both Americas. V. spiralis originates from the Mediterranean region, with native habitats in Northern Africa, Western Asia, and Southern Europe. C. caroliniana and V. spiralis are considered thermophilic, whereas E. nuttallii is adapted to temperate climates. Since their introduction to Europe and parts of Asia, these species have shown high vegetative growth, strong competitiveness, and the ability to form dense monospecific stands, making them difficult to eradicate. As a result, they are classified as invasive. C. caroliniana and E. nuttallii are included on the EU List of Invasive Alien Species of Union Concern (Commission Implementing Regulation (EU) 2016/1141), requiring member states to eradicate them. These two species are commonly found in both natural and anthropogenically altered, nutrient-rich water bodies, while V. spiralis typically proliferates in thermally polluted waters, often associated with mining or power plants.

For the purpose of this study, several hundred individuals of each species were collected and transported to Adam Mickiewicz University in Poznań. E. nuttallii specimens were gathered from Lake Skoki (Kujawsko-Pomorskie, Poland; 52°36'18"N, 19°23'32"E), C. caroliniana from a fish pond in Krążek (Małopolska, Poland; 50°17'25"N, 19°27'08"E), and V. spiralis ramets from a thermally altered water canal near Lake Licheńskie (Greater Poland, Poland; 52°18'31"N, 18°20'31"E), part of the thermally polluted Konin Lakes. These species were collected from separate locations, as they do not co-occur naturally in Poland. Although no genetic analyses were conducted, it is likely that all specimens were clonal, considering the absence of viable seed production and the tendency of non-native aquatic plants to reproduce mainly vegetatively in their introduced ranges (McCracken et al. 2013).

Laboratory cultivation was conducted in summer using identical methods and conditions. Due to structural differences, preparation and morphometric measurements varied. E. nuttallii and C. caroliniana shoots were trimmed to 11 cm and 13 cm, respectively and included a top shoot. V. spiralis leaves were trimmed to a 15 cm length from shoot base. After one-week acclimation, 190 healthy and morphologically similar individuals per species were selected and randomly divided into three groups: 80 for the light experiment, 80 for the temperature experiment, and 30 as controls.

The initial morphometric parameters were recorded. For E. nuttallii and C. caroliniana, these included length of the main shoot, number of offshoots, length of individual offshoots and total plant length (sum of the length of main shoot and its offshoots). For V. spiralis, number of leaves, as well as their length and width, were measured. Dry weight (whole plants) was determined from control specimens. Experimental plants were randomly assigned to one of ten treatment groups (five treatments for the temperature experiment and five treatments for the light experiment). In the temperature experiment, plant were cultivated at 7, 10, 14, 17, and 21 °C under constant low light (46 μmol m^–2s^–1). In the light experiment, plants were kept at 21 °C and exposed to 3, 10, 25, 50, or 100% of maximum light (2.9, 9.7, 22.0, 46.4 and 91.1 μmol m^–2s^–1). A 12-hour day-night cycle was applied. Light intensities matched conditions in eutrophic lakes at 1.5–5 m depth (Sobczyński et al. 2012; Rybak et al. 2024), while the tested temperature range reflects conditions commonly observed in Central European lakes during the vegetative season (Zhu et al. 2023; Ptak et al. 2024).

Each treatment group contained 16 plants divided into four replications. Four plants from each replication were placed within 2-liter glass tank (water column height: 19 cm), filled with filtered site water (0.45 μm GF/C filters) and were planted in neutral substrate (stream quartz gravel, 2–5 mm) in plastic pots. Four tanks from the same treatment were placed in the same water bath consisted of a 32 L aquarium, half-filled with distilled water, and enclosed in polystyrene containers with lids (Suppl. material 1), to stabilize temperature and light conditions. In total, 10 polystyrene containers with water baths were used for each species (five for the temperature experiment and five for the light experiment), one for each treatment group. For V. spiralis, pots with plants were individually numbered due to higher variability among individuals, allowing for the tracking of each specimen’s fate.

Light within polysteryne containers was provided by adjustable LED lights (1.7–91.1 μmol m^–2s^–1 PAR), mounted on the underside of the container lids. Light intensity was verified using a LI-1400 meter with LI-193SA spherical sensor. Water temperatures were controlled using thermostated aquarium heaters placed within the water baths and monitored by DIVER-type submersible data loggers. Plants cultivated at 21 °C were kept in the cultivation room (maintained at a constant temperature of 20 °C), while those grown at lower temperatures were placed in a cold room maintained at 5 °C. In the latter case, aquarium heaters were used to raise the water bath temperature to the desired level. Plants were cultivated under these conditions for seven weeks. Afterward, they were remeasured, including chlorophyll a content via CCM-300 Chlorophyll Content Meter. For V. spiralis, newly formed daughter ramets were also measured as well as their number. Finally, all plants were dried, and their final dry weight was recorded.

Most statistical analyses were conducted in R (R Core Team 2023) using the RStudio environment (Posit Team 2023). To assess differences in species responses to treatments, one-way ANOVA followed by Tukey’s post hoc test was used. Prior to ANOVA, assumptions of normality and homogeneity of variances were tested using the Shapiro-Wilk and Levene’s tests, respectively. When variances were unequal, Welch’s ANOVA was applied; in cases of strong non-normality, the Kruskal-Wallis test was used. Species response curves to temperature and light gradients were modeled using Generalized Additive Models (GAM; Hastie and Trevor 1992), with Poisson distribution and smoothing complexity determined by the Akaike Information Criterion (AIC; Lepš and Šmilauer 2003; (Suppl. material 2). Data visualization was performed with the ggplot2 package (Wickham 2016), and dplyr (Hadley et al. 2023) was used for data wrangling. ChatGPT was occasionally used to improve text grammar and fluency.

Tested temperature levels significantly influenced elongation (total length of the main shoot and offshoots, main shoot length) and mean chlorophyll a content of Cabomba caroliniana (Fig. 1a−d, Suppl. material 3), elongation (total length of the main shoot and offshoots, main shoot length) of Elodea nuttallii (Fig. 2a−d, Suppl. material 3) and all the tested attributes for Vallisneria spiralis (Fig. 3a−d, Suppl. material 3) (Table 1, Suppl. material 4). The parameters of C. caroliniana decreased with decreasing temperature, with the lowest dry mass recorded in treatments with the coldest water. Temperature of 14 °C or colder almost completely inhibited the plants elongation in case of this species (p < 0.0001 for both total and main shoot length). No offshoots were present in the treatments with water temperature of 14 °C or colder (p = 0.0601). Chlorophyll a content was highest in the individuals kept in 21 °C (p = 0.0002). The elongation of E. nuttallii decreased alongside the drop of temperature gradient with the shortest specimens found in the water temperature of 7 to 14 °C (p < 0.0001 for both total and main shoot length). However, even under the coldest treatments, the plants exhibited, on average, growth compared to their initial values. In general, all the parameters of V. spiralis decreased alongside with the decrease in the temperature gradient, including total leaf length and the mean length of the three longest leaves (p < 0.0001 for both parameters). Leaf elongation of the plants kept at 7 or 10 °C was minimal. Dry mass growth was observed across all treatments, with the lowest values recorded in plants grown at temperatures of 10 °C or lower (p = 0.0026). Plants cultivated at the temperature of 7 or 10 °C on average produced almost no leaves (p < 0.0001), had the lowest chlorophyll a content (p < 0.0001) and did not produce any daughter ramets (p = 0.0001).

Figure 1. 

Trait comparison of Cabomba caroliniana cultivated under different water temperatures (a–d) and light levels (e–h). The name of the performed statistical test is provided on each plot. Total length was calculated as the combined length of the main shoot and its offshoots. Black dots on the boxplots represent mean values, empty dots represent outliers, while dashed lines indicate the initial parameter values of the control group.

Figure 2. 

Trait comparison of Elodea nuttallii cultivated under different water temperatures (a–d) and light levels (e–h). The name of the performed statistical test is provided on each plot. Total length was calculated as the combined length of the main shoot and its offshoots. Black dots on the boxplots represent mean values, empty dots represent outliers, while dashed lines indicate the initial parameter values of the control group.

Figure 3. 

Trait comparison of Vallisneria spiralis cultivated under different water temperatures (a–d) and light levels (e–h). Change in total length was calculated as the difference in the sum of leaf lengths from the main shoot and its ramets before and after the experiment. Black dots on the boxplots represent mean values, empty dots represent outliers, while dashed lines indicate the initial parameter values of the control group.

Significant differences between light intensity treatments were observed across all species. In C. caroliniana, elongation (total and main shoot length) and mean chlorophyll a content varied between groups (Fig. 1e–h, Suppl. material 3). For E. nuttallii (Fig. 2e–h, Suppl. material 3) and V. spiralis (Fig. 3e–h, Suppl. material 3), all measured parameters differed significantly (Table 2, Suppl. material 4). In C. caroliniana, elongation was strongly reduced at 25% light intensity or lower (p < 0.0001 for both total and main shoot length). The highest chlorophyll a content occurred at 25% light (p = 0.0112), with lower values at both ends of the light spectrum. In E. nuttallii, both extreme light levels (3%, 10% and 100%) resulted in reduced values for most parameters. The highest shoot length (p < 0.0001) and dry mass (p < 0.0001) were observed at 25% and 50% light. At 3% light, elongation ceased and dry mass declined. Additionally, necrosis was observed in 7 of the 16 specimens under the 3% light treatment, affecting up to 1.5 cm (0.86 cm on average) of the basal part of the shoot. Before measurements were taken, these parts were removed, and only the green portions of the plants were analyzed. Offshoots developed in all treatments, but their number declined under the lowest light (p = 0.0138). Mean chlorophyll a content increased with decreasing light (p < 0.0001). For V. spiralis, reduced light negatively affected all growth parameters. The lowest dry mass and smallest increase in leaf number were recorded at the lowest light level (p < 0.0001 for both), with values falling below initial levels at 3% intensity. Daughter ramet formation was completely inhibited at 3% and 10% (p < 0.0001). Total leaf length and mean length of the three longest leaves differed significantly across treatments, with plants at 3% showing marked reductions (p < 0.0001 for both). Although mean chlorophyll a generally increased as light decreased (p < 0.0001), values at 3% were significantly lower than at 10% and 25%.

Our findings show that three of the studied species exhibit unique responses to thermal and light conditions, occupying distinct ecological niches. Using Generalized Additive Models (GAM), we analyzed the response of key morphological traits to light and temperature (Figs 4, 5, 6, Suppl. material 5). All species exhibited negative growth responses to decreasing water temperatures. For C. caroliniana, a pronounced decline in total length of the main shoot and offshoots as well as in the number of offshoots occurred between approximately 14 and 21 °C, while a stable decrease in main shoot length and dry mass was observed along the entire temperature gradient (Fig. 4, Suppl. material 5). For this species, a strong negative growth response to low light intensities was also observed, notably a decrease in the total length, as well as in main shoot length. In the case of E. nuttallii, an especially strong decline in main shoot and total length was recorded between 14 and 21 °C, while the reduction in dry mass remained stable along the temperature gradient (Fig. 5, Suppl. material 5). In contrast, the response of E. nuttallii to decreasing light levels differed from that to temperature, as the species exhibited a largely unimodal pattern with an optimum around 50–55 μmol m^–2 s^–1, where most parameters such as main shoot length, total length, and dry mass peaked, followed by a significant decline at higher and lower light intensities. In turn, the number of offshoots decreased steadily with decreasing light, with the rate of loss accelerating along the gradient. For V. spiralis, both elongation and new leaf formation decreased stably as temperature decreased, while dry mass loss accelerated at temperatures between 7 and 14 °C (Fig. 6, Suppl. material 5). For this species, a strong negative response to low light intensities was also observed, specifically an accelerating decrease in total length of the leaves and dry mass.

Figure 4. 

Generalized Additive Models (GAMs) illustrating the responses of selected, statistically significant growth traits of Cabomba caroliniana to varying water temperature (a–c) and light intensity (d, e).

Figure 5. 

Generalized Additive Models (GAMs) illustrating the responses of selected, statistically significant growth traits of Elodea nuttallii to varying water temperature (a, b) and light intensity (c–e).

Figure 6. 

Generalized Additive Models (GAMs) illustrating the responses of selected, statistically significant growth traits of Vallisneria spiralis to varying water temperature (a, b) and light intensity (c, d).

Our results show that the studied aquatic macrophytes tolerate a broad range of water temperatures and persist under extremely low light conditions. C. caroliniana is often described as light-demanding (Ørgaard 1991; Hiscock 2003), with optimal growth reported at ≥ 300 light μmol m^–2s^–1 (Koleszár et al. 2022; Huang et al. 2023). However, in our experiment, where maximum light intensity was 91.1 light μmol m^–2s^–1, it still elongated and accumulated biomass in all but the darkest treatment (2.9 light μmol m^–2s^–1). These results align with recent findings suggesting C. caroliniana can tolerate moderate shade and outcompete other macrophytes at ≤ 150 light μmol m^–2s^–1 (Koleszár et al. 2022; Huang et al., 2023). Though considered thermophilic, C. caroliniana has been documented in diverse habitats, including near-freezing conditions (Ørgaard 1991). Our results support this ecological plasticity: the species survived at 7–10 °C but showed minimal growth below 14 °C. Vegetative reproduction was completely suppressed below 17 °C, suggesting that although survival in colder water is possible, growth and spread are constrained until temperatures exceed this threshold.

E. nuttallii was the only species that exhibited increased growth under lower light. Its low-light preference is well-documented (Angelstein and Schubert 2009), and our results reinforce its adaptation to such environments. Optimal growth occurred between 22.0 and 46.4 light μmol m^–2s^–1 - slightly lower but consistent with prior reports: 40–48 (Barrat-Segretain 2004), 28–80 (Szabó et al. 2019), 51–94 (Angelstein and Schubert 2009), and even 80–180 in a short-term (12-day) study (Szabó et al. 2020). While direct comparisons are complicated by differing methodologies, the species consistently shows strong shade tolerance. This may provide it with a competitive advantage in early spring, facilitating its dominance in plant communities at the start of the growing season. Notably, basal shoot necrosis under low light may indicate a detachment strategy, enabling shoots to float toward better-lit zones (Kunii 1984). In nature, these detached fragments may form floating mats (Zehnsdorf et al. 2015), especially in autumn, reducing light penetration and oxygen levels in water. Native to temperate North America, E. nuttallii was expected to exhibit the highest cold tolerance, which was confirmed by our findings. While shoot elongation decreased at lower temperatures, individuals grown at 7–10 °C still elongated, increased in biomass, and produced offshoots, unlike the other species. This observation aligns with findings by Kunii (1981), who observed winter growth in water as cold as 4 °C. Although growth occurred in all treatments, elongation was significantly reduced at the lowest temperatures. A three-year long field study in Slovenia (Grudnik et al. 2014) showed rapid expansion once winter/spring temperatures exceeded 10 °C. Our results, in agreement with these studies, underscore E. nuttallii’s broad thermal tolerance and the critical role of temperature in its growth and invasive success.

Data on V. spiralis’ light and temperature preferences are limited. A greenhouse study (Zhao et al. 2013) showed growth across a light range of 90–500 light μmol m^–2s^–1. Similarly, in our experiment, the species grew and produced biomass even under 9.7 light μmol m^–2s^–1, though ramet formation occurred only above 22.0. Our findings suggest that V. spiralis is more temperature- than light-limited. Growth and vegetative reproduction were strongly inhibited below 14 °C, with optimal performance at higher temperatures. This aligns with its distribution in northern Europe, where it mainly occurs in naturally or artificially heated waters exceeding 10 °C in winter (Hussner and Lösch 2005; Hutorowicz 2006; Wasowicz et al. 2014). Field observations by Hutorowicz (2006) reported population dominance at water temperatures above 20 °C and a decline below 15 °C, consistent with our findings and highlighting the strong impact of low temperatures on the development of this species.

In summary, although the tested invasive macrophytes can survive suboptimal conditions, temperature clearly emerges as the dominant environmental factor shaping their growth dynamics and potential for spread under future climate scenarios. This confirms that temperature remains a key limiting factor for the expansion of the tested invasive macrophytes, especially for C. caroliniana and V. spiralis, whose vegetative reproduction is seriously hindered in lower temperatures. From this perspective, the broad temperature tolerance of E. nuttallii presented in this study is alarming, as this species may already find the optimal growth conditions in central Europe (Draga et al. 2024) therefore, further climate-induced warming of water bodies may only enhance its competitiveness and expansion potential.

It is widely acknowledged that the most effective control methods rely on a deep understanding of a species’ ecology and behavior (Hussner 2017), and this is equally true for managing invasive aquatic plants. Notably, many of these species in their introduced ranges - such as C. caroliniana, E. nuttallii, and V. spiralis - do not produce viable seeds and instead depend entirely on their highly efficient vegetative reproduction. Therefore, particular attention should be paid to their process of elongation and propagule production. Our studies demonstrate a strong relationship between water temperature and the production of offshoots and ramets in species such as C. caroliniana and V. spiralis. Consequently, for these species, it is recommended to apply control methods in early spring, before water temperatures rise to levels that promote the formation of new offshoots and ramets. At the same time, the high shade tolerance observed in the tested species suggests that they may easily dominate local plant populations at the beginning of the growing season, when both the length of the day and the sun’s zenith angle limits underwater light availability. Reducing populations of these invasive species during this time of year would undoubtedly allow native, light-sensitive plants to develop undisturbed. Additionally, since species like C. caroliniana and E. nuttallii appear capable of surviving in cold water for extended periods, a second control treatment in autumn may prevent them from overwintering in dense patches. Such dense patches likely increase chances of their survival through the winter as well as accelerate the colonization process (Hoffmann et al. 2015). Furthermore, our studies indicate that although all the tested species can be considered shade-tolerant, light levels as low as 2.9 light μmol m^-2s^-1 significantly inhibited their growth, resulting in biomass levels falling below control levels early in the experiment. This biomass decline was especially evident in the case of E. nuttallii and V. spiralis suggesting that control methods based on full shading may be particularly effective for these species. Interestingly, while C. caroliniana and E. nuttallii were still able to produce a limited number of offshoots under the lowest light conditions, V. spiralis failed to produce any daughter ramets when exposed to light levels of 9.7 light μmol m^-2s^-1 or lower. If such inhibition of propagule production by shading can be replicated under field conditions, it would suggest that shading could be an extremely effective control strategy for V. spiralis.

In conclusion, both the high shade tolerance of the tested species and their resistance under low water temperatures suggest that the threat posed by these species is likely to only intensify in the future. Projections indicate that eutrophication levels in freshwater ecosystems will continue to rise (Jeppesen et al. 2009), leading to further reductions in water transparency and, consequently, limiting the availability of light for submerged macrophytes. This anticipated decline in light penetration, combined with rising water temperatures, is expected to create favorable conditions for shade-tolerant species characterized by broad thermal tolerance, such as those examined in this study. As a result, both the invasiveness and geographical range of these species are likely to expand in the coming years.