Effects of ammonium pulse on the growth of three submerged macrophytes

Ammonium pulse attributed to runoff of urban surface and agriculture following heavy rain is common in inland aquatic systems and can cause profoundly effects on the growth of macrophytes, especially when combined with low light. In this study, three patterns of NH4-N pulse (differing in magnitude and frequency) were applied to examine their effects on the growth of three submersed macrophytes, namely, Myriophyllum spicatum, Potamogeton maackianus, and Vallisneria natans, in terms of biomass, height, branch/ramet number, root length, leaf number, and total branch length under high and low light. Results showed that NH4-N pulse caused negative effects on the biomass of the submerged macrphytes even on the 13th day after releasing NH4-N pulse. The negative effects on M. spicatum were significantly greater than that on V. natans and P. maackianus. The effects of NH4-N pulse on specific species depended on the ammonium loading patterns. The negative effects of NH4-N pulse on P. maackianus were the strongest at high loading with low frequency, and on V. natans at moderate loading with moderate frequency. For M. spicatum, no significant differences were found among the three NH4-N pulse patterns. Low light availability did not significantly aggregate the negative effects of NH4-N pulse on the growth of the submersed macrophytes. Our study contributes to revealing the roles of NH4-N pulse on the growth of aquatic plants and its species specific effects on the dynamics of submerged macrophytes in lakes.


Introduction
In ecosystems, resource supply is rarely constant but varies in frequency and magnitude; therefore, "resource pulse" often occurs as a common ecological phenomenon [1][2][3]. Resource pulses can affect the physiological metabolism of plants, change their species pattern, and even influence the ecosystem function [4][5][6]. The effects of resource pulses have been widely explored in terrestrial ecosystems at some ecological scales, such as specific species [7,8]; wild population [9][10][11]; community [12][13][14][15][16]; and ecosystem [17,18]. However, few studies have focused on determining resource pulse effects on aquatic systems [3]. In fact, runoff of urban a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 surface and agriculture following heavy rain can lead to several-fold increases in nutrient concentrations in lake, sewage channel or river [3,19]. This phenomenon is common in the middle-lower reaches of the Yangtze River in China where many lakes are located, which will largely aggravate lake eutrophication and thus influence the survival and growth of submersed macrophytes [3,19,20].
In this study, three submersed macrophytes, namely, M. spicatum, Potamogeton maackianus, and V. natans, were used to test the effects of NH 4 -N pulse on their growth. These three species were selected because of the following: (1) they can be found in water from mesotrophic to eutrophic conditions and are widely distributed in the middle-lower reaches of the Yangtze River [44][45][46][47]; (2) they have different growth forms, that is, canopy forming for M. spicatum and rosette forming for V. natans; (3) M. spicatum tends to accumulate large amounts of nitrogen under high NH 4 -N condition. P. maackianus and V. natans show high ability of maintaining carbon constant [21,39]; and (4) V. natans is characterized by low-light compensation point [48,49].
This study aims to investigate the effects of NH 4 -N pulse on the growth of M. spicatum, P. maackianus, and V. natans under two light treatments. Three patterns of NH 4 -N pulse, differing in magnitude and frequency, were applied to the plants. Responses to NH 4 -N pulse were compared among the three species, and possible mechanisms were discussed. We tested the following hypotheses: (1) ammonium inhibits the growth of plants, and low light availability exacerbates the ammonium effects because of reduced photosynthetic carbohydrate production [34,35]; (2) ammonium pulse with high loading and low frequency will be more harmful to the growth of submersed macrophytes compared with ammonium pulse with low loading and high frequency because the former causes more severe damage to the cell structure [43]; and (3) M. spicatum will be more sensitive to NH 4 -N pulse due to its high ability to accumulate nitrogen [21].
The experimental design consisted of four replications for two light treatments (HL: high light of about 50% sunlight; LL: low light of about 25% sunlight) and three patterns of NH 4 -N loading for each species. The experiment lasted for 57 days. During plant acclimatization, the plants were incubated in buckets filled with tap water; M. spicatum and P. maackianus had developed some roots during this phase, which lasted for 28 days. In the phase of NH 4 -N pulse (Phase I), three patterns of NH 4 -N loading were applied to the plants within 16 days with total NH 4 -N loading of 6551 mg m -2 for each pattern. Pattern I [high NH 4 -N loading with low pulse frequency (HL)]: ammonium chloride solution was added to 24 buckets (eight buckets per species) in the morning with NH 4 -N loading of 1637.7 mg m -2 every four days for four times (namely 4 mg L -1 NH 4 -N each time after conversion according to water volume); Pattern II [moderate NH 4 -N loading with moderate pulse frequency (MM)]: NH 4 -N loading of 818.8 mg m -2 every two days for eight times (namely 2 mg L −1 NH 4 -N each time after conversion according to water volume) was applied to 24 buckets with eight buckets per species; and Pattern III [low NH 4 -N loading with high pulse frequency (LH)]: NH 4 -N loading of 409.4 mg m -2 every day for sixteen times (namely 1 mg L −1 NH 4 -N each time after conversion according to water volume) was applied to 24 buckets. Twenty-four buckets (eight buckets per species) without NH 4 -N loading were used as control. In the phase of releasing NH 4 -N dosing (Phase II), water in all of the buckets was refreshed with tap water and the plants were maintained for 13 days. In Phases I and II, one cup of plant in each bucket was collected every four days and growth parameters were measured. 23 June 2018 was the last sampling time of Phase I, and 5 July 2018 was the last sampling time of Phase II, and more detailed descriptions were shown in Fig 1. During the experiment, water in each bucket was renewed and filamentous algae on the plant leaves were gently removed every 7 days. The water temperature was 29.8˚C-36.9˚C. The concentrations of TN, TP, NO 3 -N, NH 4 -N, and PO 4 -P in the water column were 0.198 ± 0.079, 0.01 ± 0.018, 0.114 ± 0.036, 0.001 ± 0.004, and 0.0018 ± 0.000 mg L −1 , respectively, and maintained relatively stable during the experimental period. The concentrations of alkali-hydrolyzable N, available P, available K, and organic material in the sediments were 0.11, 0.03, 0.12, and 19.95 mg g −1 , respectively. The pH of the sediments was 6.59.
The plant samples were washed with distilled water three times and carefully separated into leaves, stems, and roots for M. spicatum and P. maackianus and separated into leaves and roots for V. natans. The samples oven dried at 80˚C to constant weight, and growth parameters were measured and recorded (shoot/plant height, leaf number, ramet number, root length, branch number, and total branch length).
SPSS software was used for statistical analyses. Values were expressed as means ± standard error (SE). All data were tested for normality and homogeneity before analyses. ANOVA was performed to evaluate the effects of NH 4 -N pulse and light treatment on the growth of the three submerged macrophytes. Means were compared by Duncan's multiple range tests. Explained variances of growth parameters were analyzed using three-way ANOVA, with biomass, height, branch/ramet number, root length, leaf number, and total branch length as dependent variables, with light treatments, NH 4 -N pulse patterns and experimental phases (NH 4 -N pulse phase and release of NH 4 -N loading phase) as fixed factors, and interactions between the three fixed factors considered.

Effect of NH 4 -N pulse on M. spicatum
The NH 4 -N pulse patterns, experimental phases, and interaction of light availability and pulse patterns significantly affected the biomass of M. spicatum (Table 1, P<0.001). In the control, biomass was higher in high light than in low light during release of NH 4 -N loading phase, while biomass was lower in high light under high NH 4 -N loading with low pulse frequency during this phase (Fig 2, P<0.01). Overall, the three patterns of NH 4 -N pulse gradually decreased the biomass (Fig 2 and Table 1). Biomass in high light was the highest among the eight groups on June 23 and July 5, whereas biomass in low light was the lowest under LH on July 5 (Fig 3).
The shoot heights of M. spicatum were markedly influenced by the NH 4 -N pulse patterns, experimental phases, interaction of light availability and experimental phases, and pulse patterns and experimental phases ( Table 1, P<0.001). Compared with LH, HL and MM evidently affected the shoot height during NH 4 -N pulse phase, showing higher shoot height in low light than in high light (Fig 2 and Table 1). The shoot height in the control was higher than that in HL on June 23 and was the highest among the eight groups on July 5. The shoot height showed an increasing trend from HL to LH (Fig 3).
The branch number of M. spicatum was significantly affected by the interaction of light availability and NH 4 -N pulse pattern only (Table 1, P = 0.049) and varied among the eight groups on June 23 and July 5, expect in the group treated with low NH 4 -N loading in high light on June 23 (Fig 2).
The root length was significantly influenced by the experimental phases, interaction of light availability and NH 4 -N pulse patterns, and interaction of the three factors (Table 1). During  the NH 4 -N pulse phase, the root length of plants in low light was longer than that of plants in high light under high NH 4 -N loading; by contrast, the root length in low light was shorter than that in high light under low NH 4 -N loading (Fig 2). On June 23 and July 5, the root length under high NH 4 -N loading in high light was significantly shorter than that in the control ( Fig  3). Moreover, the root length kept invariable among the patterns of NH 4 -N pulse, except for the control (Fig 3). The total branch length of M. spicatum was markedly influenced by NH 4 -N pulse patterns, experimental phases, interaction of light availability and experimental phases, and pulse patterns and experimental phases (Table 1, P<0.001). The total branch length in low light was longer than that in high light under high and moderate NH 4 -N loading (Fig 2). Similar to shoot height, the total branch length showed an increasing trend from high to low NH 4 -N loading in high light (Fig 3).

Effects of NH 4 -N pulse on P. maackianus
NH 4 -N pulse patterns significantly affected the biomass of P. maackianus (Table 2, P<0.05). Under moderate NH 4 -N loading, the biomass was higher in low light than that in high light during NH 4 -N pulse phase (Fig 4, P<0.05). The biomass in the control was higher than that under high NH 4 -N loading in low light on June 23, while the biomass in the other groups kept invariable (Fig 5).
The shoot height was significantly affected by light availability ( Table 2, P<0.05). The shoot height was longer in the control than under moderate NH 4 -N loading in low light on June 23 and was not significantly different among the other groups (Fig 5).
In the control, the branch number in high light was significantly higher than that in low light during the release of loading phase (Fig 3, F = 5.790 P=0.026). On July 5, the plants showed higher branch number under moderate NH 4 -N loading than under control and high NH 4 -N loading (Fig 5).
The root length was marginally higher in high light than in low light under moderate NH 4 -N loading during the release of loading phase. The leaf number was markedly affected by the NH 4 -N pulse patterns, experimental phases, and interaction of light and experimental phases ( Table 2). On July 5, the plants showed greater leaf numbers under moderate and low NH 4 -N loading than under control and high NH 4 -N loading; meanwhile, the leaf number did not change among the other groups (Fig 5).
The total branch length was affected by experimental phases (Table 2). On June 23, the total branch length was longer under control than under high and moderate NH 4 -N loading. On July 5, the total branch length under moderate and low NH 4 -N loading was significantly greater than those under high NH 4 -N loading (Fig 5).

Effects of NH 4 -N pulse on V. natans
The biomass of V. natans was markedly affected by NH 4 -N pulse patterns and experimental phase (Table 3). During pulse phase, the biomass in low light was marginally greater than that in high light under moderate NH 4 -N loading (Fig 6). The biomass was significantly lower under moderate NH 4 -N loading than under control on June 23 but did not change among all the eight groups on July 5 (Fig 7).
The shoot height was significantly influenced by light availability, NH 4 -N pulse patterns, experimental phases, interactions of light and pulse patterns, and interactions of pulse patterns and experimental phases ( Table 3, P < 0.05). The shoot height differed among treatments with varying light availability during both phases under control but was kept unchanging under the three other NH 4 -N pulse patterns (Fig 6). On June 23, the shoot height under moderate NH 4 -N loading was lower than that under control and high NH 4 -N loading. On July 5, the shoot height in low light under control was higher than that under the three other NH 4 -N pulse patterns (Fig 7).
The ramet number showed significant responses to experimental phases, interaction of light and experimental phases, and interaction of the three factors (i.e., light, NH 4 -N pulse patterns, experimental phases). Under control, the ramet number was greater in low light than in high light during pulse phase. Under moderate and low NH 4 -N loading, the ramet number was greater in high light than in low light during the release of loading phase (Fig 6). On July 5, the ramet number was greater under high NH 4 -N loading than under moderate NH 4 -N loading (Fig 7).
The root length was significantly influenced by NH 4 -N pulse patterns and interaction of experimental phases, light, and pulse patterns. Under moderate NH 4 -N loading, the root length was greater in high light than in low light during the pulse phase. On June 23, the root length was the greatest under moderate NH 4 -N loading in high and low light and under control in low light. Meanwhile, the root length was the greatest in control in high and low light (Fig 7).
The leaf number was markedly affected by experimental phases and interactions of light, NH 4 -N pulse pattern, and experimental phase ( Table 3). The leaf number was marginally higher in low light than in high light under control and moderate NH 4 -N loading during the pulse phase and under high NH 4 -N loading during the release of loading phase (Fig 6). The leaf number did not change, regardless of NH 4 -N pulse patterns in different light availability levels or experimental phases (Fig 7).

Discussion
Our experimental results showed that NH 4 -N pulse inhibited the growth of submerged plants by reducing the biomass, height and total branch length on July 5 and the root length under high NH 4 -N loading pulse in high light availability for M. spicatum and by decreasing the biomass under moderate NH 4 -N loading pulse in high light for V. natans. This finding is partly consistent with our first hypothesis. In general, excess NH 4 -N would cause stress to submersed macrophytes, and NH 4 + toxicity arising from eutrophication probably plays an important role in the degradation of the plants [20,21,29,[50][51][52]. For example, Cao et al. (2007) set 0.56 mg L −1 NH 4 -N in the water column as the upper limit for survival of V. natans in lakes in the middle-lower reaches of the Yangtze River in China [26]. In the present study, the biomass of V. natans kept invariable under NH 4 -N concentration of 0-1 mg L −1 regardless of light availability. This discrepancy could be attributed to multiple stresses from biotic and abiotic competition, herbivory, and wave exposure, except for NH 4 -N stress and low light availability in field lakes [53,54]. In our previous study, the biomass of M. spicatum was constant under stable NH 4 -N concentration of 2 mg L −1 for 4 days [55]; by contrast, in the present study, the biomass of M. spicatum declined obviously under NH 4 -N pulse of 2 mg L −1 treated for eight times within 16 days. These findings may suggest that NH 4 -N stress was partly time dependent. Short-term ammonium loading may not inhibit the growth of submersed macrophytes, but long-term exposure could significantly affect the growth of plants.
The effects of NH 4 -N pulse on plants differed among the three pulse patterns and were species specific. For P. maackianus, the branch number, leaf number, and total branch length Table 2. Percentage (%) of explained variance based on three-way ANOVA for change in biomass, height, branch number, root length, leaf number, and total branch number of P. maackianus treated with NH 4 -N pulse patterns under high and low light during NH 4 -N pulse phase and release of NH 4 -N loading phase. Boldface type indicates significant difference at the level of P < 0.05. were the lowest in high light on July 5 under high NH 4 -N loading pulse. No significant differences in these morphological traits were observed between moderate and low NH 4 -N loading pulses. These results are consistent with our second hypothesis stating the high tolerance of this species to NH 4 -N stress because of its relatively high conservative carbohydrate metabolism [39]. However, high NH 4 -N concentration may decrease the chlorophyll content and affect the electron transport system of plants [30][31][32], probably explaining the decline in some of the morphological traits for this species. For V. natans, the ramet number in low light on   Effects of ammonium pulse on the growth of three submerged macrophytes July 5 and the shoot height in high light on June 23 were the lowest under moderate NH 4 -N loading pulse. The high carbon supply for its low light saturation point may lead to the high turnover of carbon metabolism [56,57], leading to high rehabilitation capacity, which may benefit the growth of V. natans exposed to ammonium toxicity of high NH 4 -N loading pulse in the present study. For M. spicatum, no significant differences or consistent patterns were found among the three NH 4 -N pulse patterns, except for the shoot height in high light on July Effects of ammonium pulse on the growth of three submerged macrophytes 5. The carbon supply may be used for maintenance in addition to detoxification in accordance with its high sensitivity to NH 4 -N pulse and high light saturation point [21,39,49]. The effects of NH 4 -N pulse on the growth of submersed macrophytes were species specific. In the present experiment, the effect of NH 4 -N pulse on the biomass of M. spicatum was significantly greater than those on V. natans and P. maackianus, conforming to our third hypothesis. M. spicatum showed a characteristic of acquisitive strategy with relatively high NH 4 -N Effects of ammonium pulse on the growth of three submerged macrophytes absorption and growth rates [21,36,39,58]. Large amount of nitrogen could be accumulated in M. spicatum during NH 4 -N pulse phase, resulting in profound effect on their growth for a long time [21,59] and inhibited growth due to shortage of carbohydrate consumed as C-skeleton for FAA synthesis. For V. natans, the carbon supply is relatively sufficient for its low light saturation point [56,57], suggesting its high ammonium detoxification in the present study.  found that V. natans was more efficient at maintaining C-N metabolic homeostasis when faced with acute NH 4 + dosing [39]. For P. maackianus, the biomass kept unchanged regardless of NH 4 -N pulse patterns and light availability, except for that in high NH 4 -N loading pulse in low light on June 23. Imamoto et al. (2007) reported that P. maackianus showed positive RGR at only 13% transmittance [60]. This finding may be ascribed to the considerably low light requirements of this species [61]. Previous studies pointed out that P. maackianus can avoid instant ammonium toxicity by restricting photo-synthesis through uncoupling photophosphorylation, leading to low accumulation of nitrogen even in high levels of NH 4 -N [62]. Low light is a major factor determining the growth of submerged macrophytes [63,64], which could aggravate the effects of ammonium stress, especially in eutrophic lakes [26]. However, compared with high light in the present study, the biomass was higher in low light for M. spicatum during the release of loading phase under high NH 4 -N loading and for P. maackianus and V. natans during the pulse phase under moderate NH 4 -N loading. This finding partly contradicts our first hypothesis that low light availability exacerbates the ammonium effects. Cao et al. (2011) found that growth of M. spicatum would not be affected negatively by NH 4 -N under high light but decreased significantly under low light treatments [38]. Similar results were reported for P. crispus [35,40], Potamogeton amplifolius and Nuphar advena [34]. Under high light conditions, plants may show a consumptive carbon strategy, that is, high respiration rate [65,66], because of sufficient supply of photosynthetic carbon. Apart from maintaining growth, large amount of carbon may be used for ammonium detoxification [21]. By contrast, under low light conditions, plants may adopt conservative carbon strategy when facing carbon starvation by having low respiration and high carbohydrate storage [66][67][68]. These phenomena could partly promote the results in the present study.

Light (L) NH 4 -N pulse patterns (N) Phases (P) L× N L ×P N×P L×N×P
Macropyhte species can modify the morphological traits to overcome environmental and resource stresses [69,70]. In the present study, the shoot height and total branch length were higher in low light than in high light for M. spicatum during the pulse phase under high and moderate NH 4 -N loading; moreover, the shoot height and leaf number were greater for V. natans under control. These results are consistent with previous findings that terrestrial and aquatic plants would allocate more resources to the above ground part by increasing their height or number of leaves to obtain more light and improve their light competitiveness and survival fitness [71][72][73][74]. The roots act as organs for absorption and assimilation of nitrogen and phosphorus from sediments particularly in oligotrophic lakes [75,76]. In the present study, the root length of plants was inhibited under NH 4 -N pulse for M. spicatum and V. natans, consistent with previous results of decreasing allocation of underground parts under high nutrients for terrestrial plants [77,78]. M. spicatum absorbed NH 4 + mainly by the leaf rather than by the root when NH 4 -N in the water column was about 0.1 mg L −1 [75]. These results confirm that the role of the roots as a nutritive organ weakened in the hypertrophic environment [79].
To conclude, NH 4 -N pulse affected the growth of submerged macropytes in species specific and light dependent manner. The negative effects on M. spicatum were significantly greater than that on V. natans and P. maackianus. The negative effects of NH 4 -N pulse on P. maackianus were the strongest at high loading with low frequency, and on V. natans at moderate loading with moderate frequency. For M. spicatum, no significant differences were found among the three NH 4 -N pulse patterns. Low light availability did not significantly aggregate the negative effects of NH 4 -N pulse on the growth of the submersed macrophytes.