Combined effect of temporal inundation and aboveground-cutting on the growth performance of two emergent wetland plants, Phragmites australis and Bolboschoenus planiculmis

Study site

The pot experiment and plant collection were conducted in the Nakdong River estuary (NRE, N35°06′13.81″E128°56′46.81″), Republic of Korea (Figs. 1A, 1B). The study site has a humid subtropical climate based on the Köppen climate classification (Busan Metropolitan City, 2022). The annual average temperature is 14.6 °C. The annual rainfall is approximately 1,354 mm. Over half of the annual precipitation falls during the summer monsoon season (i.e., June to August). The NRE has been designated as Natural Monument (No. 179), Wetland Protection Area, and Nature Reserve due to its ecological significance as a wintering habitat for migratory waterfowl and its distinctive biodiversity.

Plant materials

Two dominant emergent plant species, P. australis and B. planiculmis, were used in the NRE pot experiments. P. australis, also known as common reed, is a perennial wetland plant of the Poaceae family that grows to 1–3 m. P. australis is found in a wide range of habitat from inland to brackish coastal environments. In the Nakdong River Estuary, the P. australis community is established along the mid-high tide line of the tidal marsh, sandbars, and elevated sand dunes (Busan Metropolitan City, 2022). Recent studies in other continents have reported aggressive behavior of invasive P. australis (e.g., haplotype M in North America; Lindsay et al., 2023; Saltonstall, 2002) by expanding its distribution or replacing native subspecies. Majority haplotypes of P. australis in the Republic of Korea were identified as the P types that is the dominant form in East Asia and Australia (Chu et al., 2011).

B. planiculmis is a tuberous wetland plant belonging to the perennial Cyperaceae family and can grow to a height of 0.4–0.9 m. The B. planiculmis community is also distributed along the mid to high tide line of the tidal marsh, partly overlapping with P. australis. Both plant species are clonal plants with a guerrilla architecture that has a strong ability to spread laterally due to its greater inter-ramet distance. The guerrilla architecture provides advantages for foraging in heterogeneous environments (Xue et al., 2018). The annual production of tuberous roots of B. planiculmis is regionally important in the NRE and the East Asian Australasian Flyway, as many waterfowl species, such as Cygnus cygnus (Whooper swan) and Anser fabalis (Bean goose), use it as a carbohydrate source during the winter season at stopover sites (Kim et al., 2013, 2016; Kim & Kim, 2021).

We collected all plant material from a nearby population established near the restored wetland in the NRE. To establish plant clones for our manipulation experiments, we first collected the rhizomes of juvenile P. austalis plants from the restored wetland in April 2010 (Fig. 1B). We cut the collected rhizomes to a length of 10 cm in with a similar wet weight. Each rhizome had a single node producing stem and roots. Prepared rhizomes were planted in near-surface sediment (2~3 cm deep) in plastic pots that were 10 cm in diameter and 15 cm in height. Each pot was filled with sediment that we collected from the tidal marsh at NRE. The collected sediment was sieved through a 2 mm filter to remove other plant material prior to use. The texture of the collected soil was classified as sandy loam. P. austalis rhizomes were grown in soil saturated with water from the restored wetland. In October 2009, B. planiculmis tubers were collected from the NRE and stored in wet condition inside a refrigerator (4 °C). As the growth rate of juvenile plants can be influenced by different tuber sizes, we screened tubers of the same size (i.e., length: 2 cm). In April of the following year, we planted selected tubers of B. planiculmis in plastic pots and kept them under saturated soil conditions for 2 weeks to facilitate tuber sprouting. All plant clones of the wetland species were grown under the same condition for 2 months before the manipulation experiments.

Pot experiments

Two mesocosm tanks measuring 6 m in length, 4 m in width, and 1 m in depth, were placed on the shore of a restored wetland in Eulsuk Island, Nakdong River Estuary (coordinates: N 35.1039°, E 128.9464°; EPSG 4326). The aforementioned restored wetland is located near the estuarine barrage and is connected to the primary river channel by two sluice gates. Therefore, the average salinity of the restored wetland remains brackish except during prolonged periods of flooding after the summer monsoon. The mesocosm tanks were filled with water from the restored wetland. The salinity of the tank water was monitored, and maintained at a brackish condition between 4–8 PSU throughout the experiment periods. Tank water was changed once at the end of inundation treatment to prevent algal blooms in the tank.

One experimental tank was used to maintain the constant water level condition during the initial growth period and the last 90-days before harvest (Fig. 1C). The other tank was used to manipulate four different water level conditions (i.e., 0, 10, 30, 50 cm depth from the soil surface) using aluminium blocks of different heights (Figs. 1D, 2). The manipulated water level conditions did not fully submerge the stem of P. australis. After 60 days of initial growth period, pots of P. australis and B. planiculmis were then moved to four different water level conditions mixed with four different inundation periods (i.e., 0, 10, 20, 30 days; Figs. 2A, 2B). To compare the growth performance of P. australis under combined stressors, we placed P. australis pots, one of which used stem-cutting as a mechanical management option, in the restoration wetlands. There were eight replicate pots for P. australis, including a stem-cutting group, and 10 replicate pots for B. planiculmis in each manipulation experiment group. In total, the P. australis experiment had 160 pots (i.e., control group: 16 pots, manipulated group: cut [2 levels] × depth [3 levels] × inundation period [3 levels] × 8 replicates) and B. planiculmis experiment had 100 pots (i.e., control group: 10 pots, manipulated group: depth [3 levels] × inundation period [3 levels] × 10 replicates). We randomly placed the replicate pots within the manipulated depth ranges in the tanks. At the end of inundation treatments, all pots were moved to the initial tank and the water level was adjusted to sediment saturation level as in the control group pots. Our experimental procedure mimics raising and maintaining the water level for 30 days to make plant inundation condition in the restored wetland, and then lowering the water level to average condition for the remainder of the growing season. To minimize the potential differences from pot position, we randomly rotated the pot position every 2 months.

We harvested all plants at the end of the 90 days growth period after the flooding manipulation experiment. We measured stem length (cm), stem diameter (cm), stem density in each pot (stems per pot), and dry weight of the entire plant for each pot. Stem volume was calculated assuming the stem was a volumetric cone by using the following equation.

(1) $$\begin{array}{rl}\mathrm{S}\mathrm{t}\mathrm{e}\mathrm{m}\mathrm{v}\mathrm{o}\mathrm{l}\mathrm{u}\mathrm{m}\mathrm{e}(\mathrm{c}{\mathrm{m}}^3\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m}\cdot \mathrm{p}\mathrm{o}{\mathrm{t}}^{-1})=& \mathrm{\pi }\times \mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m}\mathrm{r}\mathrm{a}\mathrm{d}\mathrm{i}\mathrm{u}\mathrm{s}(\mathrm{c}\mathrm{m}{)}^2\times \mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m}\mathrm{l}\mathrm{e}\mathrm{n}\mathrm{g}\mathrm{t}\mathrm{h}(\mathrm{c}\mathrm{m})÷3\\ & \times \mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m}\mathrm{d}\mathrm{e}\mathrm{n}\mathrm{s}\mathrm{i}\mathrm{t}\mathrm{y}(\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m}\cdot \mathrm{p}\mathrm{o}{\mathrm{t}}^{-1})\end{array}$$

The number of pots with surviving plants was recorded to calculate the mortality rate for different inundation depths and periods. At the end of inundation treatment after 90 days, all experimental pots were checked for the survival of plants to the end of the experiment. Plant mortality was recorded according to the following criteria; (1) no new shoot or leaf growth after the inundation treatment, (2) decomposition of stem or rhizome with losing its original shape and color. Mortality rate was calculated as the relative percentage of pots with dead plants to the number of replicate pots under the same experimental condition. All measured plant material was moved to the laboratory and cleaned with tap water to remove any attached sediment or algae. The dry weight of all plant material was measured after drying in a 70 °C convection oven for 3 days. In addition, the number of B. planiculmis tubers was counted, and their dry weight was also recorded.

Data analysis

Differences in measured response variables between P. australis and B. planiculmis were tested using an aligned-rank transform (ART) ANOVA method. The ART ANOVA is a non-parametric analysis of variance that allows testing of the combined treatment effects of two factors (Wobbrock et al., 2011). Where the main effect was significant, post-hoc comparisons for the interaction in a two-way model were carried out using ART contrasts (ART-C; Elkin et al., 2021). All statistical analysis was performed in the R environment (version 4.3.0; R Core Team, 2023).

Inundation effect on the mortality of P. australis and B. planiculmis

Mortality of two emergent plants was notably affected by different inundation depths and lengths in pot experiments (Figs. 3A–3C). All plants in the control groups survived under inundation treatments, with a mortality rate of 0%, except those in the stem cutting experiment, which had a mortality rate of 12.5%. Inundation treatments slightly increased the mortality rate of P. australis to a range of 0–37.8%, as shown in Fig. 3A. When combined with mechanical stress (i.e., stem cutting) prior to inundation treatment, these inundation effects on mortality rate were further enhanced (Fig. 3B). Mortality rate of P. australis increased to 50–87.5% with both inundation and stem cutting. Mortality rate of P. australis was higher in groups exposed to inundation for 20–30 days when compared to those exposed to short-term inundation for 10 days. The mortality rate of B. planiculmis increased with deeper inundation depths and longer inundation periods (see Fig. 3C). All pots planted with B. planiculmis survived the 10cm inundation treatment. The mortality rate of B. planiculmis increased when exposed to 30–50 cm inundation for 20–30 days.

Change of stem volume and biomass

We found that changes in inundation depth significantly decreased the stem volume and total biomass of P. australis (Table 1 and Figs. 4A, 4B). However, the growth performance of P. australis remained unchanged with increasing inundation periods (stem volume: F = 1.13, p = 0.329; total biomass: F = 0.33, p = 0.718). When combined with stem cutting treatment, stem volume of P. australis showed a significant decrease with changing inundation depths (F = 5.99, p = 0.004) and periods (F = 6.03, p = 0.004; Figs. 4C and 4D). After stem cutting, all combinations of different inundation treatments showed a decrease in total biomass compared to the control group. Therefore, no difference was observed between the different inundation treatment groups (p > 0.38).

In comparison, B. planiculmis was found to be more sensitive to different inundation conditions than P. australis (Table 1 and Figs. 4E, 4F). Stem volume and total biomass of B. planiculmis decreased with increasing inundation depth and duration. The combined effect of inundation depth and period resulted in significant changes in the growth performance of B. planiculmis across different inundation environments (stem volume: F = 22.62, p < 0.001; total biomass: F = 28.04, p < 0.001). In particular, the combined effect of inundation depth and period showed distinct growth response patterns of B. planiculmis (Figs. 4E, 4F). At inundation depths below 10 cm, stem volume and total biomass of B. planiculmis increased with longer inundation periods. Under deeper inundation conditions, between 30 and 50 cm, longer inundation periods tended to reduce stem volume and biomass of B. planiculmis. This suggests that B. planiculmis thrives in shallow habitat conditions between 10 and 30 cm.

Tuber production of B. planiculmis

We also compared how inundation treatments affected tuber production in B. planiculmis. Similar to the results for stem volume and total biomass, the density and total weight of B. planiculmis tubers exhibited a negative response to shifting inundation conditions (Table 2 and Figs. 5A, 5B). Tuber density was higher in the 10 cm inundation group and the control group than in plants that grew in deeper inundation depths (F = 52.40, p < 0.001). Tuber density increased with inundation periods in the 10 cm inundation treatment. Under deeper inundation conditions, tuber density decreased with increasing inundation periods (F = 20.50, p < 0.001). Total tuber weight exhibited a similar response pattern to the changes in tuber density under different inundation conditions.

Effect of temporal inundation in restored wetland

We tested the growth performance of P. australis and B. planiculmis for the effect of temporal inundation, aboveground cutting, and the combined effect of both. Previous studies on freshwater P. australis populations have reported that increasing inundation period, depth, or frequency can rapidly reduce the growth rate and biomass of P. australis (Tóth, 2016; Yi et al., 2020). In this study, we observed that independent short-term inundation treatments (i.e., 10–30 days, 10–50 cm depth) did not significantly reduce the growth performance of P. australis compared to the control group (i.e., saturated surface sediment). The inconsistent effects of inundation treatment observed in this study may be due to the local environmental conditions to which the experimental P. austalis has been adapted. Song et al. (2021) identified the distinct local adaptations of P. australis populations from inland and coastal wetlands. Due to periodic environmental variations associated with the hydrological regime (such as mean water level, fluctuation frequency, and tidal range), the growth characteristics of P. australis populations in coastal wetlands may experience reduced physiological stress via water level manipulation.

Likewise, populations of P. australis distributed on intertidal mudflats in the Nakdong R. Estuary have adapted to periodic tidal changes. The range of water level manipulation (i.e., 0–50 cm) used in our inundation experiments may not be a sufficient to induce physiological changes in P. australis growth. This finding indicates that a greater degree of inundation or water level manipulation is necessary to manage the growth performance of P. australis in coastal areas, where local plant populations are continuously exposed to inundation, decreased soil iron oxides (Ding et al., 2021) and water level fluctuations. This can serve as an applicable management measure to control the spread of P. australis, if the objective is only to control the expansion of P. australis. However, there may be potential differences in plant growth responses due to environmental settings between restored wetland sites and controlled mesocosms, so further field experiments are needed for in-depth understanding.

As described in our experimental site, the potential range of inundation treatments in restored wetlands may be constrained when the management objective deals with multiple species or management targets. We found that different growth responses to manipulated inundation stress were observed between P. australis and B. planiculmis. In our experiment, under 10–50 cm of inundation stress, the growth inhibition of B. planiculmis was significantly greater than that of P. australis. This suggests that the use of inundation for vegetation control may have negative impact on other endemic wetland plants that have a narrower range of adaptations to water level fluctuations or inundation periods. P. australis has a higher density of aerenchyma in its stems, rhizomes, and roots compared to B. planiculmis (Mal & Narine, 2004; Ding et al., 2021; Tshapa, Naidoo & Naidoo, 2021). The enhanced gas pathway of P. australis ensures oxygen supply to the belowground system during flood conditions. These morphological characteristics explain the different growth responses of the two emergent plants in our experiments. Our experiment result provides empirical evidence that deeper inundation cannot efficiently manage to maintain higher biomass of B. planiculmis, an important food plant resource at the study site.

In our experiment, we observed that biomass and tuber production of B. planiculmis tended to increase under shallow inundation of 0–20 cm and short-term inundation of 10–20 days. Similar experiments manipulating water levels have also shown that tuber production of B. planiculmis was higher under shallow inundation environments (less than 5–10 cm), and decreased under deeper water conditions (Liu et al., 2016; An, Gao & Tong, 2018; Ding et al., 2021). In controlled plot experiments, An et al. (2022) reported that the optimum growth range for B. planiculmis in freshwater environments is between 11–36 cm depth. They also suggested that shallow inundation (0–10 cm) during the early growing season and deeper inundation (10–25 cm) during later growing seasons can assist in B. planiculmis growth and enable it to outcompete other plant species (An et al., 2022). This could be attributed to the local adaptation of B. planiculmis communities, which are often distributed in environments with fluctuating water levels (Hroudová, Zákravský & Flegrová, 2014).

Aboveground-cutting effect on the growth performance of P. australis

Previous restoration experiments reported that cutting and leaving the above-ground part of P. australis resulted in rapid expansion of the remaining belowground plant portion within a year, while aboveground cutting had no significant effect (Greet & King, 2019). A study using various mowing regimes, such as stem cutting, on P. australis plot experiments in Central Europe found that mowing had only a short-term effect on reducing shoot size, even with various cutting trials (Güsewell, Le Nédic & Buttler, 2000). There seems to be little practical evidence supporting the long-term effect of a single mowing event. Other studies have emphasized the importance of nutrient availability in controlling excessive colonization of P. australis in wetland ecosystems (Uddin & Robinson, 2018).

In our short-term experiment, we found that combination of aboveground-cutting and inundation treatments had a significant inhibitory effect on the growth performance and survival rate of P. australis. The growth traits such as stem volume and biomass of P. australis, and the associated survival rate, were both decreased. It is noteworthy that, despite increased level of inundation after cutting, there was no increase in growth inhibition effect. A similar application in the Laurentian Great Lakes watershed showed that a cut-to-drown management strategy for P. australis greatly reduced belowground biomass production and rhizome non-structural carbohydrate content (Widin, Bickford & Kowalski, 2023). Aboveground cutting of P. australis not only affects the total biomass of P. australis patches, but also has a significant management effect in reducing the adaptive capacity of P. australis in wetland ecosystems. For example, the remaining dry shoots of P. australis (usually produced in the previous year) play an important role in aerating the belowground system through the Venturi effect (Björn et al., 2022). In addition, convective flow of gases (e.g., oxygen, carbon dioxide) is initiated in leaf sheaths of the living P. australis stem, then the gases are transmitted through culm, rhizome, and vented to old stems (Armstrong & Armstrong, 1991). This interconnected air-flow network of P. australis provide a resistant and escape mechanism against partial inundation of the entire stem-rhizome network consisting of new and old stems. In this respect, a cut-to-drown strategy or combination of cutting and inundation treatment can inhibit the supply of air through the stem-rhizome network into the drown plant part of P. australis, suppress the diffusion of oxygen, and decrease the essential metabolic processes in the stem-root network.