Assessment of Physicochemical Characteristics of Surface Water in the Full-Dyke and Semi-Dyke Systems: A Case Study

The present study aimed to assess the physicochemical characteristics of surface water in the full-dyke (FD) and semi-dyke (SD) systems of the An Giang province. Water samples were collected in both seasons in 2019 (1) the first monitoring phase in the dry season (April) and (2) the second monitoring phase in the rainy season (Octo -ber). At each season, the surface water samples were collected at 15 sites in the FD system and 15 sites in the SD system to analyze the water quality parameters. The results showed that the concentrations of DO, COD, BOD, TSS, NO 3- , NH 4+ , TKN, PO 43- and TP parameters in the FD system were higher than in the SD system and there was a significant difference (p<0.05), except the value of pH and DO in the dry season. The DO, COD, BOD 5 and TSS concentrations in the FD system tended to exceed the Vietnamese standard threshold (QCVN 08:2015/ BTNMT, column B1), but in the SD system, these values were within the allowable limits of Vietnamese Standard. Other parameters, such as NO 3- , NH 4+ , TKN, PO 43- and TP exceeded the threshold of the Vietnamese standard in both systems. The results of this study provide the scientific information on water pollution inside the FD system. Therefore, it was expected that the local government of the An Giang province should have long-term strategies to manage the FD system to improve the water quality inside the FD system.


INTRODUCTION
The An Giang province is one of the largest rice producers in the Vietnamese Mekong Delta (VMD), contributing around 16.1% of the rice production in VMD in 2019 (GSO, 2019). An Giang is located upstream of the Vietnamese Mekong Delta and along the Hau rivers of the Mekong river system. Therefore, it has been affected by the natural floods that usually start from July to November (Tong, 2017) and August to October every year (Phung et al., 2017). In order to increase the rice yield and protect the crops, the government of Vietnam implemented the five-year development plan 1996 -2000 for the VND region (Government of Vietnam, 1996). As a result, the low dyke system was constructed and helped against the early flood peak arriving around mid-July to mid-August (Triet et al., 2017). After the disastrous flood in 2000, the high dyke system (i.e., the so-called FD system today) was built in the An Giang province to protect the floodplains against a flood. Initially, the FD system showed positive impacts in agriculture fields, such as increased rice crops (i.e., farmers could produce three rice crops per year) and yield and decreased the risk of flood damage (Kien et al., 2019). Moreover, it also protected the living of farmers and stabilized the livelihoods of people (Thieu and Dung, 2014). In parallel with the positive impacts, the FD system also revealed some adverse environmental effects. Previous studies suggested that the input of sediment deposits and natural fertilizers (as NPK) in sediment were prevented by the FD system (Triet et al., 2017), and the biodiversity of fish were lost in the Assessment of Physicochemical Characteristics of Surface Water in the Full-Dyke and Semi-Dyke Systems: A Case Study recent year ( Kien et al., 2019). Besides, the construction of the FD system has impacted human health (Berg, 2002), wild fi sheries (Nguyen et al., 2018), surface water quality (Nha, 2006;Chau et al., 2015), and degradation of soil quality (Xuan, 2004). Therefore, in the long term, farmers could apply too much agrochemicals in rice cultivation for rice growing and contribute to the surface water pollution (Abe et al., 2016).
Cho Moi is one of the districts of the An Giang province that has long historically constructed the FD system, especially in the My Luong townlet of the Cho Moi district ( Duyen et al., 2015). Previous studies showed that the FD system prevented the water exchange inside and outside of the system; thus, this issue could degrade the surface water resources (Duyen et al., 2015;Hoa et al., 2006). However, these studies just focused on evaluating surface water quality parameters inside the FD system, but were not co mpared with the semi-dyke (SD) system. Therefore, the present study was performed to compare the physicochemical properties of surface water between the FD and SD systems during sampling times in the dry and rainy seasons.

Site description
Cho Moi is one of the 11 districts of the An Giang province. Cho Moi consists of 18 administrative regions, including 2 townlets and 16 communes (Fig. 1). The total natural area is 36,906 ha, of which agricultural land is 30,030 ha, accounting for 81.4% of the total area (GSO, 2019). Most rice paddy fi elds were protected by diff erent Figure 1. Map of Cho Moi district in the An Giang province dyke systems. The FD systems protected the rice fi eld over the whole fl ooding period, and the SD system (i.e., the so-called August dyke systems) protected the rice fi eld only in the early peak of the annual fl ood in August.

Identifying the study site
According to the Sub-Department of Irrigation of the An Giang province (2013), Cho Moi has 85 dyke regions, including 82 regions in the FD system and 3 regions in the SD system. My Luong townlet has a long history of constructing the FD system; therefore, it was selected to represent the area inside the FD area for collecting the water samples. In addition, the An Thanh Trung district was selected as the control region for collecting samples, because it is located near the Hau river and is easily aff ected by the fl ood season.
The water samples were collected two times per year in 2019, including the fi rst phase in April, which represented the dry season, and the second phase in October, which represented the fl ood season. A total of 30 samples in each sampling time were collected, including 15 samples in the FD system and 15 samples in the SD system. The locations of water samples in both seasons were presented in Fig. 2.
The water samples were collected from 20-30 cm below the water surface and at three sampling points far from each other of 3-5 m, then mixed to collect a representative sample. The sampling plastic bottles (1 L) were rinsed three times with the water samples at the sites. All water samples were stored in the ice chest and transported to the laboratory, where they were kept at 4 o C in the refrigerator prior to analysis. The water temperature, pH, and DO were directly measured in situ with Hanna portable meters. In the laboratory, the following water parameters, such as COD (mg/L),

Statistical analysis
All environmental parameters in the present study were analyzed to compare the surface water quality between the FD system and the SD system in two seasons. All statistical analyses were performed in Statistical Package for Social Sciences (SPSS) version 26.0 (SPSS Inc., Chicago, II, USA). The data were analyzed using the two-independent samples analysis of variance to compare the water quality diff erence between the FD system and SD system in each season. The Mann-Whitney U test was employed to determine the signifi cant diff erences between means at 5% level of signifi cance (p < 0.05). All computations and the fi gures were prepared in the Origin 2021 software (Origin-Lab, Northampton, MA, USA).

Temperature , pH and DO concentration of the surface water samples in the FD and SD systems
The temperature of the water samples was presented in Fig. 3. The fi rst monitoring phase (dry season), the water temperature in the FD system ranged from 29 -30.8 o C with the average value at 30.8 ± 1.04°C and in the SD system ranged from 28.3 -29.5°C with the mean value at 29.5 ± 0.98°C (Fig. 3). The second monitoring phase (rainy season), the temperature value in the FD and SD systems fl uctuated from 28.3 -31.2°C with the average at 29.2 ± 0.78°C and 28 -32°C with the mean value at 31.0 ± 0.97°C, respectively. The results in this study corresponded to the results of previous studies, such as Giao (2020) who recorded the temperature in the Hau river fl uctuating from 27.1 -32.0°C and in the Mekong river ranging from 19.9 -32.2°C (Ongley, 2009).
The pH value of surface water in the SD system tended to be higher than in the FD system (Fig. 4a). In the dry season, there were no signifi cant diff erences between the FD and SD systems (p>0.05) in the pH value, with the mean Figure 3. Temperature of surface water in the FD system (a) and the SD system (b) value in the FD and SD system at 7.61 ± 0.28 and 7.74 ± 0.4, respectively. However, there were signifi cant diff erences (p<0.05) between the FD and SD systems in the rainy season, with the pH values reaching 6.97 ± 0.34 and 7.24 ± 0.12, respectively. According to Giao (2020), the pH value in the Tien river and Hau river fl uctuated from 7.1 -7.2 and 6.7 -7.12, respectively. Although the pH value in the dry season tended to be higher than in the previous study, it was still suitable for development of fi sh as well as aquatic animals and plants, and especially for irrigation (QCVN 08:2015/BTNMT), with the pH value in the range of 6 -8.5 (column A1) and 5.5 -9 (column B1).
The DO concentration also shows a similar trend to the pH value (Fig. 4b). The DO concentration of surface water in the dry s eason was not signifi cantly diff erent between the FD and SD systems (p>0.05; Fig. 4b).
However, the DO concentration in t he rainy season was higher in the SD system than in the FD system and signifi cantly diff erent (p<0.05).
In particular, the DO concentration reached 4.27 On the other hand, the DO concentration suitable for aquatic life should be higher 5 mg/L (Ongley, 2009). However, the DO concentration in the FD system was relatively low, showing that the surface water has been organically polluted. In contrast, the DO concentration in the SD system was suitable for the previous studies. In general, the water temperature, pH and DO at the study sites were within the National technical regulations on surface water quality (QCVN 08-MT: 2015/BT-NMT, column B1), with the limit value of pH (5.5 -9) and DO (≥ 4 mg/L). The outliers of pH were found at S11, S12, S14 sites in FD-dry and S1 sites in the FD-rainy. Note: FD: full-dyke; SD: semi-dyke Figure 5 shows the concentration of three parameters in the dry and rainy seasons. Overall, the concentration of COD, BOD 5 and TSS in the FD system were higher than in the SD system and signifi cantly diff erent (p<0.05; Fig. 5). The COD con centration of surface water was higher in the FD system than in the SD system, and these fi gures were signifi cantly diff erent (p<0.05; Fig.  5a). Specifi cally, the COD concentration in the FD system ranged from 23.0 -80.6 mg/L in the dry season and from 24 -67.2 mg/L in the rainy season. The COD value in the SD system fl uctuated from 12.5 -17.1 mg/L in the dry season and 11.2 -17.6 mg/L in the rainy season. According to Tuyen et al. (2014), the concentration of COD was higher in the dry season than in the rainy season, and the COD value in the FD system increased higher than in the SD system. Comparison with the Vietnamese standard indicates that the average values of COD in the SD system were within the National technical regulation of surface water quality for irrigation (MONRE, 2015). However, the COD concentration in the FD system exceeded the Vietnamese standard with the mean values in the dry season (45.8 ± 19.6 mg/L) and in the rainy season (34.4 ± 12.7 mg/L).

The concentrations of COD, BOD and TSS parameters of the surface water samples in the FD and SD systems
The concentratio n of BOD 5 and TSS in both seasons was signifi cantly diff erent among the FD and SD systems (p<0.05; Figs. 5b, 5c). In the dry season, the BOD 5 concentration in the FD system ranged from 7.2 -21.5 mg/L, and it ranged from 5 -9 mg/L in the SD system (Fig. 5b). In turn, the BOD 5 concentration in the rainy season ranged from 7.5 -18.5 mg/L in the FD system and 4.2 -6.9 mg/L in the SD system. The concentration in this study increased slightly compared with the study of Duyen et al. (2013), which also assessed the surface water quality in the channel of the FD system in My Luong townlet. The research recorded the BOD 5 concentration below 8.0 mg/L. Hong (2013) also found that the limited water-exchange process between inside and outside of the FD system is one of the reasons for water pollution in the FD system.
The TSS concentration in the FD system was higher than in the SD system in both seasons , and these fi gures were signifi cantly diff erent (p<0.05; Fig. 5c). In the dry season, the TSS c oncentration in the FD and SD systems ranged from 45 -112 mg/L and 20 -105 mg/L, respectively. Additionally, the TSS concentration in the rainy season fl uctuated from 40 -205 mg/L in the FD system and 40 -64 mg/L in the SD system (Fig.  5c). Boyd (1998)   The outliers of COD parameter were found at S1 sites in SD-dry and S8, S9 sites in the FD-rainy. The outliers of BOD parameter showed at S29 sites in the SD-dry. Note: FD: full-dyke; SD: semi-dyke that the number of phytoplankton species in the FD system was higher than in the SD system in the Thoai Son district. The TSS concentration in this study was within the research of Hong (2013), which was carried out in the FD and SD system in the Chau Phu district. However, the TSS concentration exceeded the Vietnamese standard (< 50 mg/L), including 14/15 sites in the FD-dry, 13/15 sites in the SD-dry, 7/15 sites in the FD-rainy, and 6/15 sites in the SD-rainy.  Fig. 6. In general, the concentration of three parameters in the FD system was higher than in the SD system, and signifi cant diff erences were found (p<0.05), except for the NH 4 + concentration in the dry season (p>0.05; Fig. 6b). After comparing these concentrations with the Vietnamese standard (QCVN 08:2015/BTNMT, column B1), the NO 3 concentration tended to be lower than the Vietnamese standard, but the NH 4 + concentration was higher than the Vietnamese standard.
The NO 3 concentration in the dry s eason fl uctuated from 0.42 -1.33 mg/L in the FD system and 0.32 -0.79 mg/L in the SD system. In the rainy season, the NO 3 concentration in the FD system ranged from 0.52 -2.26 mg/L and 0.41 -0.81 mg/L in the SD system. According to Tuyen et al. (2014), the NO 3 concentration in the FD system was also higher than in the SD system. Furthermore, previous studies have suggested that the NO 3 concentration in the dry season was higher than in the rainy season (Minh et al., 2021;Hong, 2013).
According to the Vietnamese national technical regulation on surface water (QCVN 08:2015/BTNMT, column B1), the NH 4 + concentration should be lower than 0.9 mg/L. However, the NH 4 + value at all sampling sites exceeded the Vietnamese standard in the dry season, fl uctuating from 1.54 -2.66 mg/L in the FD system and 1.26-3.78 mg/L in the SD system. In the rainy season, the NH 4 + concentration in the FD and SD systems ranged from 0.42 -2.8 mg/L and 0.14 -2.1 mg/L, respectively. In this study, the NH 4 + concentration in both seasons tended to be higher than in the previous research (Minh et al., 2021;Hong, 2013). The reason might be that the FD system in Cho Moi was built more than 20 years ago. Therefore, the concentration of water quality in the FD system accumulated inside the FD system more than in the outside FD system. This point was elucidated in the research of Gai (2012), the TKN concentration in the FD system ranged from 1 -7 years (1.83 mg/L) and 9 -10 years (2.13 mg/L). and (c) TKN concentration of surface water quality. The outliers of NO 3 parameter were found at S16, S17 sites in SD-dry. The outliers of NH 4 + parameter showed at S8, S9 sites in the FD-rainy and S21 in the SD-rainy. Note: FD: full-dyke; SD: semi-dyke Figure 6c shows that the TKN concentration in the FD system was higher than in the SD system and signifi cantly diff erent (p<0.05). In the dry season, the TKN concentration ranged from 4.76 -8.68 mg/L in the FD system and 2.1 -5.63 mg/L in the SD system. The average TKN concentration in the SD system (3.31 mg/L) was lower by half, compared with the FD system (6.55 mg/L). In the rainy season, the TKN concentration in the FD system fl uctuated from 2.66 -9.52 mg/L and 0.84 -2.49 mg/L in the SD system (Fig. 6c). The research result was consistent with the results of the research by Duyen et al. (2015), with the TKN concentration in the FD system recorded from 5.5 -11.8 mg/L in the dry season and from 3.6 -5.8 mg/L in the rainy season. Besides, Duc (2014) also indicated the amount of N fertilizers used in the FD system was higher than in the SD system, with 125 kg/ha/ crop and 113 kg/ha/crop, respectively. Therefore, the reason might be that the concentration of types of nitrogen in the FD system was higher than in the SD system. Besides, Nha (2006) also indicated that the amount of nitrogen fertilizer used inside the FD system was higher than in the outside FD system.

The concentrations of PO 4 3and TP parameters of the surface water samples in the FD and SD systems
The PO 4 3concentration in the FD system was higher than in the SD system in both seasons, with a signifi cant diff erence (p<0.05; Fig. 7a). In the dry season, the PO 4 3concentration in the FD system fl uctuated from 0.1 -3.33 mg/L and 0.06 -1.04 mg/L in the SD system. In the rainy season, the PO 4 3concentration in the FD and SD systems ranged from 0.3 -0.83 mg/L and 0.14 -0.68 mg/L, respectively. According to Tuyen et al. (2014), the mean concentration of PO 4 3in the SD system was lower than in the FD system, with 0.52 mg/L and 0.97 mg/L, respectively. Comparison with the Vietnamese standard (QCVN 08:2015/BTNMT, column B1) shows that the concentrations of PO 4 3in the FD system tended to be a higher threshold (0.3 mg/L), but these value s in the SD system were within the Vietnamese standard. Figure 7b shows that the average concentration of TP in the SD system was lower than in the FD system, and there was a signifi cant diff erence (p<0.05; Figure 7. Summary data of discriminant parameters in the FD and SD systems in both seasons: (a) PO 4 3and (b) TP concentration of surface water quality. The outliers of PO 4 3parameter were found at S16 sites in SD-dry and S27 sites in the SD-rainy. The outliners of TP parameter showed at S16 sites in the SD-dry and S16 in the SD-rainy. Note: FD: full-dyke; SD: semi-dyke commune were supported in selecting the sampling locations. Thanks to Dr. Nigel K. Downes (Researcher -CMI Integrated Expert at Can Tho University) for proofreading the manuscript. Fig. 7b). Specific in the dry season, the mean values of TP in the FD system were recorded at 1.78 ± 1.15 mg/L and 0.76 ± 0.29 mg/L in the SD system. Besides, the average values of TP in the rainy season showed a similar trend to the dry season, with 2.57 ± 0.77 mg/L in the FD system and 1.47 ± 0.46 mg/L in the SD system. The values of TP in the FD system are consistent with the results of a previous study in the Cho Moi district (Duyen et al., 2015) and Chau Phu district (Hong, 2013;Gai, 2012). In addition, the amount of fertilizer in the An Giang province was applied higher inside the FD system than outside the FD system (Duc, 2013). Specifically, farmers used 90 kg P 2 O 5 /ha/crop (equivalent to 270 kg/year) for rice cultivation in the FD system and 80 kg P 2 O 5 /ha/crop (equivalent to 160 kg/ year) in the SD system (thus, rice produced two crops per year). On the other hand, the water exchange between inside and outside of the FD system was not conducted in the long term in the FD system; therefore, it could be one of the main reasons, which made the TP concentration in the FD system become higher than in the SD system.

CONCLUSIONS
The concentration of physicochemical surface water parameters as temperature, pH, DO, COD, BOD, TSS, NO 3 -, NH 4 + , TKN, PO 4 3-, and TP showed higher values in the FD system than in the SD system and significant difference was noted (p<0.05), except the value of pH and DO in the dry season. The concentration of all parameters in the dry season tended to be higher than in the rainy season. The pH value was consistent with the Vietnamese standard (QCVN 08:2015/BTNMT, column B1). Other parameters, including the DO, COD, BOD, and TSS parameters in the FD system exceeded the threshold of the Vietnamese standard. However, the values of these parameters in the SD system were suitable for the Vietnamese standard. , and TP exceeded the threshold of the Vietnamese standard in both systems. It is expected that the government of An Giang should have essential strategies in the management of the FD system to improve the water quality inside the FD system.