An Experimental Study of Paddy Drainage Treatment by Zeolite and Effective Microorganisms (EM)

Abstract

Eco-ditch systems have increasingly been designed and applied as a strategy to decrease the risks of water eutrophication and contamination pollution for sustainable agriculture. The main goal of this study was to evaluate the water quality of eco-ditch substrates amended with zeolite and Effective Microorganisms (EM), such as pH, dissolved oxygen concentration (DO), ammonium nitrogen concentration (NH₄⁺-N), and nitrate nitrogen concentration (NO₃⁻-N). Laboratory experiments were conducted with four single substrates (soil, none substrates, natural zeolite, and zeolite loaded with EM bacteria) and two mixed substrates (soil and varying proportions of the additives, 0, 5 and 15%, m/m). Results showed that the concentration of NH₄⁺-N was decreased with the increasing rates of additives, and zeolite loaded with EM bacteria had the highest nitrogen removal rate (97.90%) under static experimental condition. The application rate of 15% zeolite loaded with EM bacteria on the eco-ditch exerted a better effect on NH₄⁺-N and NO₃⁻-N removal without pH reduction, decreased by 87.19% for NH₄⁺-N and 30.33% for NO₃⁻-N, respectively. Path analysis showed that zeolite addition had a rapid effect (path coefficient = −0.972) on free NH₄⁺-N ions adsorption in early 1–3 days, then EM loaded at zeolite further decreased NH₄⁺-N (path coefficient = −0.693) and NO₃⁻-N (path coefficient = −0.334) via bacterial metabolism. Based on the results, the applications of natural zeolite and Effective Microorganisms (EM) at an appropriate rate (15%, m/m) can significantly improve water quality of paddy drainage via exerting effects on nitrogen removal.

1. Introduction

China is the major rice producing region with 212.8 million tons of rice in 2021 [1]. However, the high demand for fertilizer and water of rice has caused serious environmental problems such as water loss, air pollution, and non-point pollution [2,3]. There is a need to focus on mitigation strategies to minimize nitrogen loss, then decrease the risks of water eutrophication and contamination pollution for sustainable agriculture.

Nitrogen loss through surface drainage from paddy fields is the main source of non-point pollution in South China [4,5]. Intercepting the water flow from the paddy fields into drainage ditches is a direct and effective way to reduce the agricultural nitrogen loss and relieve the water environmental problems [6]. Therefore, eco-ditches have increasingly been designed and applied. Eco-ditches are ditches that are filled with soil, plants [7,8], microorganisms [9], and other substrate materials, transporting excessive water from paddy fields and intercepting nutrient losses to adjacent water bodies. Wang et al. [10] designed a concrete eco-ditch with holes on double-sided wall, which had a better total nitrogen removal efficiency than soil ditch during the earlier experimental time in static condition. In order to mitigate the nutrient losses from vegetables land, a combination of eco-ditches and wetland paddy fields experiment was conducted. Results showed that the annual nitrogen removal efficiency in runoff was 73 ± 6% [11].

The ecological benefits of eco-ditches are heavily dependent on substrates, because substrates contact with paddy drainage directly and improve water quality by filtering, plant uptake [12], and other interactions among nutrients and used substrates. Zeolite has been used to remove NH₄⁺-N for its absorption and filtration properties on an industrial scale [13,14]. Therefore, zeolite was chosen to be the ideal substrate in constructed wetlands in order to treat wastewater with a high removal of 72.99% [15]. In addition, zeolite could be used as ion-exchanger and microbial carrier simultaneously for its high porosity and large specific surface area. Zeolite could decrease the NH₄⁺-N inhibition and improve the methane production by enhancing or preserving the growth of microbial populations [16,17]. As a result, zeolite would be adapted to the paddy ditches and provide suitable conditions and support media for bacterial growth.

The bacterial assemblages attached to substrates play a crucial role in eco-ditches established for adsorption, transformation, and purification of pollutants [18]. Therefore, it is important to choose suitable bacterial assemblages for paddy drainage treatment. The biotechnology named Effective Microorganisms (EM) was first proposed by Dr. Teruo Higa in Japan in 1982, who mixed beneficial microorganisms [19]. EM bacteria has been widely applied in wastewater treatment [20], decomposing pollutants in the environment by facilitating nitrification, denitrification, ammonification, and other processes, then improving water and soil quality without damages to the environment. A field experiment showed that applying EM bacteria to constructed wetlands could enhance the removal rate and treatment time of nitrogen and phosphorus [21].

Previous studies demonstrated the potential to adopt zeolites or EM on water treatment. Given the high nitrogen concentration of paddy field drainage, zeolite and EM bacteria may be important amendments to eco-ditch substrates. However, eco-ditch systems are complex matters containing many components, such as macrophytes, substrates, and microbes [22]. The water residence and biofilm development on macrophytes [23], nutrients content of sediment, water and plants [24], and sorption properties of other components have an impact on the nutrient cycling of ditch systems, which may interfere with zeolite or EM’s ability to remove the excess nitrogen from paddy drainage in the real ditches.

For this reason, a laboratory experiment was designed to eliminate the effects of other substances on the drainage water, which would remove nitrogen in the real ditches. This research just focused on the effects of zeolite and EM bacteria on water quality from paddy drainage. For details, the objectives of this study are to: (1) observe the rules of pH, dissolved oxygen (DO), ammonium nitrogen concentration (NH₄⁺-N), and nitrate nitrogen concentration (NO₃⁻-N) from paddy drainage water under varying zeolite rates and EM bacteria treatment; (2) analyze the nitrogen removal using zeolite and the joined effects of EM bacteria and evaluate the possibility of adding zeolite and EM bacteria to eco-ditches; and (3) determine the suitable eco-ditches substrate additives and rates.

2. Materials and Methods

2.1. Study Area and Materials

In some agricultural area where there is not enough space to construct wetland, then the well-managed eco-ditch systems can function as natural wetlands to store paddy drainage for purifying nutrients, pesticides, and sediment. Those eco-ditch systems mainly consist of different designed eco-ditches. The storage period is usually short, such as 5–7 days [5], considering the ditch volume and rain-fall frequency. Thus, an experiment for 9 days (from 1 November to 9 November 2021) in static condition was conducted. The study site (118°60′ E, 31°86′ N) is located at the Water-Saving Park of Hohai University in Nanjing, China.

Natural zeolite, EM bacteria, and soil were used as the research materials. Natural zeolite was produced in Weichang County, Chengde City, Hebei Province, with a particle size of 1–3 mm and bulk density of 2.15 g·cm⁻³. Main chemical components included: SiO₂ 68.3%, Al₂O₃ 13.39%, CaO 3.42%, K₂O 2.92%, Fe₂O₃ 1.06%. EM microbial agent was purchased from Jiangxi Tianyi biology group, Jiangxi, China, consisting of photosynthetic bacteria (4.00 ± 0.80 × 10⁶ cfu·mL⁻¹), saccharomycete (1.20 ± 0.20 × 10⁷ cfu·mL⁻¹), actinomycete (9.60 ± 0.80 × 10⁵ cfu·mL⁻¹), and lactobacillus (2.80 ± 0.40 × 10⁵ cfu·mL⁻¹). The soil used was taken from the drainage ditch in the Water-Saving Park, which was classified as silty loam, with a soil bulk density of 1.40 g·cm⁻³, electric conductance of 2.18 ds·m⁻¹, pH of 6.0, total nitrogen of 1.790 g·kg⁻¹, and total phosphorus of 0.430 g·kg⁻¹. Natural zeolite and soil were repeatedly washed with deionized water and air dried.

2.2. Lab Experimental Design

In order to eliminate the influences of other factors on the drainage water, only natural zeolite (Z) and Effective Microorganisms (EM) were manipulated in this beaker experiment. There are ten treatments with three repetitions in each treatment (

Table 1. Details of different experiment treatments.

No.	Treatment	The Composition of Eco-Ditch Substrates	Mass Ratio	Simulated Objects
1	SDD	soil	100	soil drainage ditch
2	CDD	none	0	concrete drainage ditch
3	Z100	natural zeolite	100	zeolite drainage ditch
4	Z100 × EM	zeolite loaded with EM	100	zeolite drainage ditch with EM
5	Z5	natural zeolite: soil	5:95	different rates of zeolite drainage ditch
6	Z10	natural zeolite: soil	10:90
7	Z15	natural zeolite: soil	15:85
8	Z5 × EM	zeolite × EM: soil	5:95	different rates of zeolite drainage ditch with EM
9	Z10 × EM	zeolite × EM: soil	10:90
10	Z15 × EM	zeolite × EM: soil	15:85

). For details, four single ditch substrates (No. 1–4) and six mixed eco-ditches substrates (No. 5–10). Single ditch substrates including natural soil (SDD), none substrates (CDD), natural zeolite (Z100), and zeolite loaded with EM bacteria (Z100 × EM), respectively. The zeolite loaded with Effective Microorganisms (Z × EM) was obtained by soaking the washed and air-dried zeolite in Effective Microorganisms microbial agent (2.40 ± 0.36 × 10⁸ cfu·mL⁻¹) for 24 h. As for mixed eco-ditches, two mixtures with three rates were utilized in experimental eco-ditch substrates, including 5% natural zeolite (Z5, 20 g·cup⁻¹), 10% natural zeolite (Z10, 40 g·cup⁻¹), 15% natural zeolite (Z15, 60 g·cup⁻¹), and 5% zeolite × EM (Z5 × E, 20 g·cup⁻¹), 10% zeolite × EM (Z10 × EM, 40 g·cup⁻¹), 15% zeolite × EM (Z15 × EM, 60 g·cup⁻¹), respectively (Table 1).

The 2000 mL standard measuring cup was selected to simulate the static condition of stored drainage in a paddy ditch under laboratory conditions. The bottoms of the cups were filled with 0.4 kg substrates at different mass ratios, as shown in in Table 1 (soil bulk density 1.40 g·cm⁻³) to simulate the different eco-ditches. According to the ratio of drainage ditch section size to water storage, 1200 mL simulated paddy field drainage was added to the measuring cup, respectively (Figure 1). Nitrogen as 2.271 g·L⁻¹ NH₄Cl and 1.443 g·L⁻¹ KNO₃ and phosphorus as 0.878 g·L⁻¹ KH₂PO₄ were applied to deionized water to simulate the paddy field drainage (

Table 2. Main water quality indexes of simulated paddy field drainage in the experiment.

Water Type	DO (mg·L−1)	pH	NH4+-N (mg·L−1)	NO3−-N (mg·L−1)	Ec (μs·cm−1)	OPR (mV)
Deionized water	7.95	6.73	0.02	0.08	9.406	435
Simulated paddy field drainage	7.47	5.76	42.33	5.11	457.3	446

Note: Ec is electric conductance, OPR is redox potential.

).

2.3. Sampling and Analyzing

Data used in this study mainly included pH, DO, NH₄⁺-N, and NO₃⁻-N during the experimental period. On days 1, 3, 5, 7, and 9 after containing simulated paddy field drainage with three replicates in each treatment (Figure 1), DO was determined by membrane electrode process with a digital dissolved oxygen sensor (Oxymax COS51D, Endress+Hauser, Germany), pH value was determined by glass electrode method with an ion selective sensor (ISEmax CAS40D, Endress+Hauser, Germany), NH₄⁺-N and NO₃⁻-N were measured by ion selective electrode method with an ion selective sensor (ISEmax CAS40D, Endress+Hauser, Germany).

2.4. Statistical Analyses

Microsoft Excel (Office 2019, Microsoft Corp, Redmond, WA, USA) was used to organize data, perform calculations, and construct figures. The statistical package SPSS version 22.0 (SPSS Inc., Chicago, IL, USA) was used for two-way ANOVA for the independent variables: natural zeolite (Z) and Effective Microorganisms (EM). The correlations of water quality (pH, DO, NH₄⁺-N and NO₃⁻-N) were analyzed using the “Corrplot” package in R. A path analysis model was established to investigate how natural zeolite and EM affected the water quality of the paddy drainage.

3. Results

3.1. Effects of Single Ditch Substrates on Water Quality

Variations in the pH, DO, NH₄⁺-N, and NO₃⁻-N of different single ditch substrates after receiving simulated paddy field drainage from 1 to 9 days are shown in Figure 2a–d. The pH in SDD and Z100 treatments was increased throughout the experiment period, while it had no significant change in the CDD treatment (Figure 2a). The Z100 × EM had the minimum pH value, decreased sharply from 5.76 to 3.89 at the 1st day, followed by marginal changes in 3–9 days. DO was increased slowly in each treatment during the early period, peaked at the 3rd day, then decreased to varying degrees (Figure 2b). The lowest DO was resulted by the Z100 × EM bacteria at the 9th day.

Except for CDD treatment, three types of ditch substrates exhibited positive impact on the reduction of NH₄⁺-N in this study, especially for Z100 and Z100 × EM (Figure 2c). The effect of natural zeolite on NH₄⁺-N performed mostly during early stage, decreased by 86.53% compared to the simulated paddy filed drainage at the 3rd day. The duration for Z100 × EM to remove NH₄⁺-N was longer, reaching the equal level to Z100 × EM treatment at the 7th day. The NO₃⁻-N concentration during days 1–3 (5.21–7.08 mg·L⁻¹) were higher than those during the later days, and the removal efficiency was better in Z100 × EM than other treatments (Figure 2d).

3.2. Effects of Mixed Ditch Substrates on Water Quality

The results showed that the water quality (pH, DO, NH₄⁺-N, and NO₃⁻-N) was significantly different between the different EM treatments under 5%, 10%, and 15% zeolite addition rates (

Table 3. The effects of natural zeolite and EM bacteria on water quality from simulated paddy filed drainage.

Water Quality	Time (d)	Treatments						Factors
		Z5	Z10	Z15	Z5 × EM	Z10 × EM	Z15 × EM	Zeolite	EM	Z × EM
pH	1	6.29 ± 0.02 a	6.32 ± 0.02 a	6.29 ± 0.02 a	6.32 ± 0.02 a	6.30 ± 0.04 a	6.30 ± 0.04 a	ns	ns	ns
	3	6.82 ± 0.01 ab	6.82 ± 0.01 ab	6.82 ± 0.01 ab	6.88 ± 0.01 a	6.85 ± 0.01 ab	6.79 ± 0.01 b	ns	ns	ns
	5	6.97 ± 0.02 b	6.98 ± 0.02 b	6.98 ± 0.02 b	7.11 ± 0.02 a	7.12 ± 0.02 a	7.11 ± 0.02 a	ns	***	ns
	7	7.13 ± 0.01 b	7.13 ± 0.01 b	7.12 ± 0.02 b	7.29 ± 0.02 a	7.34 ± 0.02 a	7.34 ± 0.02 a	ns	***	ns
	9	7.22 ± 0.03 d	7.20 ± 0.02 d	7.22 ± 0.02 d	7.55 ± 0.01 a	7.47 ± 0.01 b	7.38 ± 0.01 c	**	***	**
DO (mg·L−1)	1	8.06 ± 0.12 b	8.21 ± 0.16 ab	8.34 ± 0.04 a	8.15 ± 0.07 ab	7.53 ± 0.07 c	7.73 ± 0.12 c	*	***	**
	3	9.03 ± 0.07 a	8.74 ± 0.15 a	8.78 ± 0.13 a	7.75 ± 0.10 b	6.83 ± 0.09 c	5.48 ± 0.18 d	***	***	***
	5	5.40 ± 0.09 cd	8.26 ± 1.47 a	7.89 ± 1.42 ab	7.35 ± 0.36 b	5.79 ± 0.32 c	5.08 ± 0.06 d	*	***	***
	7	4.68 ± 0.22 e	6.07 ± 0.57 a	5.95 ± 0.53 ab	5.72 ± 0.03 bc	5.56 ± 0.11 cd	5.43 ± 0.10 d	***	ns	***
	9	4.76 ± 0.06 cd	5.26 ± 0.19 ab	5.32 ± 0.17 a	5.09 ± 0.08 b	4.88 ± 0.08 c	4.60 ± 0.06 d	ns	***	***
NH4+-N (mg·L−1)	1	31.49 ± 0.74 a	30.74 ± 0.24 a	30.32 ± 1.03 a	31.57 ± 0.85 a	31.40 ± 0.33 a	30.56 ± 0.17 a	ns	ns	ns
	3	15.81 ± 0.36 ab	15.28 ± 0.58 bc	14.77 ± 0.49 c	16.05 ± 0.32 a	15.49 ± 0.45 ab	15.71 ± 0.36 ab	*	*	ns
	5	10.94 ± 0.45 a	10.33 ± 0.55 a	9.65 ± 0.50 b	10.82 ± 0.20 a	10.32 ± 0.45 a	9.63 ± 0.33 b	***	ns	ns
	7	9.20 ± 0.19a	7.63 ± 0.95 cd	7.23 ± 0.83 de	8.48 ± 0.23 b	7.82 ± 0.34 c	7.06 ± 0.27 e	***	ns	*
	9	7.11 ± 0.21a	6.62 ± 0.41 ab	6.15 ± 0.29 b	6.98 ± 0.27 a	6.25 ± 0.44 b	5.42 ± 0.35 c	***	*	ns
NO3−-N (mg·L−1)	1	4.26 ± 0.22ab	4.06 ± 0.21 b	2.87 ± 0.88 ab	4.95 ± 1.09 a	4.78 ± 1.42 a	5.44 ± 1.36 a	ns	**	ns
	3	6.88 ± 0.13 a	6.73 ± 0.17 a	6.76 ± 0.08 a	5.88 ± 0.12 b	5.90 ± 0.11 b	5.18 ± 0.10 c	ns	***	ns
	5	6.09 ± 0.12 ab	6.20 ± 0.11 a	5.96 ± 0.11 b	5.41 ± 0.06 c	5.15 ± 0.19 d	4.59 ± 0.19 e	***	***	**
	7	6.47 ± 0.15 a	6.75 ± 0.02 a	6.64 ± 0.14 a	5.64 ± 0.14 b	4.93 ± 0.22 c	4.42 ± 0.19 d	*	***	**
	9	6.69 ± 0.25 a	7.01 ± 0.04 a	6.91 ± 0.07 a	5.67 ± 0.19 b	4.55 ± 0.21 c	3.56 ± 0.21 d	***	***	***

Note: Values followed by different letters are significantly different (p < 0.05) among the different treatments. *, **, and *** indicate significance levels at p < 0.05, p < 0.01, and p < 0.001, respectively; ns denotes those differences were not significant.

). There was a significant effect of EM on pH value. The tested water was more likely to approach neutral with pH of 7.38 to 7.55 under low rates of Z × EM. ANOVA results showed a significant interactive effect of zeolite and EM on DO. With time, the concentration of DO in mixed ditch substrates was decreased continuously. For Z10 and Z15, the application of natural zeolite resulted in higher DO, values that were 7.71% and 15.50% higher than those achieved by Z10 × EM and Z15 × EM, respectively.

Zeolite obviously impacted the NH₄⁺-N in 5–9 days and the NH₄⁺-N was decreased with increasing zeolite rates, while Z × EM would gain lower value. Additionally, NH₄⁺-N showed a lower value in Z × EM than that for natural zeolite at three compositions of ditch substrates. NO₃⁻-N was significantly affected by EM bacteria, as well as being significantly affected by zeolite and the interactive effects of zeolite and EM bacteria for 5–9 days. There was no significant difference in NO₃⁻-N among different rates of natural zeolite. Meanwhile, NO₃⁻-N decreased significantly on every test time after the EM bacteria treatment and decreased more with raising addition.

3.3. Correlation of the Water Quality Indexes

Figure 3 shows the correlations between pH, DO, NH₄⁺-N, and NO₃⁻-N of the tested water during the experimental period. For the 1st day, DO correlated negative correlation with NO₃⁻-N (p < 0.05). As for the 3rd, 5th, and 9th day, DO correlated positively correlation with NO₃⁻-N (p < 0.05). For days 5–9, pH had a negative correlation with NO₃⁻-N (p < 0.01). DO had a positive correlation with NH₄⁺-N (p < 0.01) at the 7th day, and pH correlated positively with NH₄⁺-N (p < 0.05) at the 9th day. Only at the 9th day, NH₄⁺-N correlated negatively with NO₃⁻-N (p < 0.05). Overall, pH had correlations with NH₄⁺-N and NO₃⁻-N, while DO had a positive correlation with NH₄⁺-N in this study.

3.4. Path Analysis of Ditch Substrates Effect on Water Quality

The path analysis revealed the direct and indirect effects of natural zeolite and EM bacteria on pH, DO, NH₄⁺-N, and NO₃⁻-N (Figure 4a,b). NH₄⁺-N and NO₃⁻-N were well explained by the models at the 3rd day (R² = 0.70 for NH₄⁺-N and R² = 0.85 for NO₃⁻-N) and 9th day (R² = 0.87 and R² = 0.98, respectively).

Figure 4 shows that there was a direct and significant effect of natural zeolite on NH₄⁺-N for the red dashed lines from the 3rd day to the 9th day. EM bacteria had a positive effect on NH₄⁺-N by changing pH value directly. DO was the key factor that significantly influenced NO₃⁻-N for the early period. There were direct and significant effects between natural zeolite and DO and EM bacteria and DO, respectively. EM bacteria exhibited an important contribution to NO₃⁻-N, yet the direct path between them was nonsignificant at the 3rd day. At the 9th day, both the natural zeolite and EM bacteria showed significant effects on NO₃⁻-N.

4. Discussion

4.1. Nitrogen Removal by Zeolite and Effective Microorganisms (EM)

Results indicated that both natural zeolite and EM bacteria worked effectively on nitrogen removal of paddy drainage under laboratory conditions (Figure 2c and Table 3). Zeolite had a significant effect on NH₄⁺-N removal in the early 1–3 days. It is known that zeolite has a porous structure, which provides a situation to accommodate cations, resulting in the decrease of NH₄⁺-N [25].The adsorption of NH₄⁺-N onto the zeolite surface is rapid [15] and the removal has a positive correlation with micro porosity [26]. After the rapid-adsorption stage, NH₄⁺-N may be adsorbed slowly or desorbed to reach the ion exchange equilibrium as conditions change.

It has been concluded by previous research that filters with bacterial assemblages attached to natural zeolite carrier could improve the efficiency of wastewater treatment, especially for nitrogen compounds [27]. In this study, results showed that Z100 × EM obtained the highest removal rate in the late stage. The explanation for the increase of NH₄⁺-N removal was that zeolite provided a stable environment for biosorption [28]. Except for the adsorption of zeolite, EM bacteria might promote the nitrification and denitrification of nitrogen in water, as evidenced by the remarkable reduction of DO in 5–9 days (Figure 2b). It also explains why NO₃⁻-N only had a decreased trend under Z100 × EM, while the other treatments had no significant effects on this indicator.

The concentration of nitrogen was decreased with the increasing rates of additive materials, such as natural zeolite and Z × EM (Table 3). Correlation showed that zeolite is the main factor that effected the NH₄⁺-N removal, which was consistent with the previous research results [29]. An investigation of the pond microbiota in South China concluded that microbial metabolic activity influenced the nitrogen metabolism, the excess nitrogen of water would decrease with the increasing denitrification capacity of microbial communities [30]. Research indicated that the application of low final concentration of EM bacteria and calcium peroxide could improve water quality and significantly reduce total phosphorus, total nitrogen, and NH₄⁺-N [31]. The highest removal rate was shown in Z15 × EM, decreased by 87.19% for NH₄⁺-N and 30.33% for NO₃⁻-N.

There was a remarkable EM bacteria effect on the pH and DO of paddy drainage in this study. The H⁺ ion might be produced during microbial metabolic activity (such as nitrification process) [32], then decrease the pH value. For Z100 × EM treatment, there was a striking reduction of pH from 5.76 to 3.89 at the 1st day, then it was kept stable in the later days. Excessive EM bacteria may harm water environments for the lower pH value in this study, while 5–15 % adding rates had few effects on pH reduction. Oxygen is an important material for the nitrification process. Therefore, the changes in DO could be considered as an evidence of nitrogen removal by bacterial metabolism.

The results of this experiment suggested that it would be a feasible strategy to apply zeolite and EM bacteria on nitrogen removal of paddy drainage. Even for the low adding rates (5–15%) of Z × EM, NH₄⁺-N and NO₃⁻-N were greatly reduced in a short time (3–9 d).

4.2. Analysis of Influencing Factors to Nitrogen in Paddy Drainage

In this study, the direct and indirect effects of zeolite and EM bacteria treatments on water quality (pH, DO, NH₄⁺-N, and NO₃⁻-N) were quantitatively derived by the correlation maps (Figure 3a–f) and path analysis model (Figure 4a,b). The path analysis showed that the mechanisms by which zeolite and EM bacteria influenced the nitrogen removal were different with time. The path coefficient of nature zeolite to NH₄⁺-N (path coefficient = −0.972 for the 3rd day and −0.693 for the 9th day) indicated that the natural zeolite made a great contribution to NH₄⁺-N removal, and this effect was heavier in the early stage than later days. Therefore, it is a feasible way to obtain short-term fast removal effects by adding zeolite. The path coefficient of EM bacteria to NH₄⁺-N was small at the 3rd day (path coefficient = −0.052). This is likely because the large amount of free NH₄⁺-N inhibited the nitrifying bacteria [33]. Nitrification might be exerted with the decrease of NH₄⁺-N concentration, finally changed the path coefficient to −0.258.

The Z × EM treatment had a better removal NO₃⁻-N ability than natural zeolite (Figure 4a,b) during the experimental period, with stable path coefficients from −0.383 to −0.334. The effects of EM bacteria addition on the NO₃⁻-N removal of the paddy drainage resulted from the denitrification of biometabolism. DO is an important factor in the process of nitrification and denitrification [34]. Results also showed a significant association between DO and NO₃⁻-N (Figure 3a–f). Path analysis indicated that zeolites and EM bacteria could indirectly remove the NO₃⁻-N of paddy drainage by influencing DO (p < 0.05). In addition, pH is also an important index for water quality. Therefore, the addition rates of EM bacteria must be considered in water treatment applications, taking into account that EM has a significant influence on pH reduction in this study.

5. Conclusions

This study presents the performance of natural zeolite and Effective Microorganisms (EM) on water quality of eco-ditch systems receiving paddy field drainage, including changes in pH, DO, NH₄⁺-N, and NO₃⁻-N. It was found that Z100 × EM had the highest nitrogen removal rate under static experimental condition. These free NH₄⁺-N ions were adsorbed rapidly under zeolite application at an early stage, then the left NH₄⁺-N and NO₃⁻-N were removed by EM bacterial metabolism. The application rate of Z × EM at 15% may form an ideal downstream water environment, providing less nitrogen and benign pH value. These results suggested that natural zeolite and Effective Microorganisms (EM) have excellent potential as appropriate amendments to eco-ditches for improving the receiving water quality of paddy drainage.

It should be noted that the conclusions were based on laboratory studies, which was subject to some qualifications. During the experiment, this study ignored the effects of some indicators such as soil properties of ditches, schedules of irrigation and drainage, fertilization in the real ditches on the drainage water quality. Therefore, further studies of field experiments should be conducted.