REMOVAL EFFICIENCY OF NITROGEN, PHOSPHORUS AND HEAVY METALS ASSOCIATED WITH SWINE WASTEWATER USING AQUATIC MACROPHYTES

Experimental design

Four plant species consisting of Typha domengensis, Eleocharis dulcis, Acrostichum aureum, Limnophyton obtusifolium and one control without plant were assigned a completely random into 15 composite barrels with 1.2m (length) x 0.6m (width) x 0.7m (depth) with 3 replications for each treatment (Figure 1). A layer of soil was placed at the bottom of each barrel with a depth of 30cm. Plants were grown with an initial density of 6 plants/experimental barrels, exception to the control. Water level in each treatment was gradually raised to plant height after plantation, and water column-maintained depth of 30 cm from the mud surface by an effluent system with PVC Ф-34 pipes. The composite barrels were buried into the soil with a depth of 30 cm equal to the soil depth inside each barrel. Raw wastewater collected from pig farms was contained in a wastewater tank and filled into the barrels with frequency of every 3 days and 5 liters per time at an inlet zone of each treatment through the influent wastewater system as described in Figure 1. Based on the volume of water column, the surface area of barrel and the amount of wastewater entered (Adelaja et al. 2014; Dong et al. 2022), a retention time of wastewater stayed in each treatment system was estimated approximately 14 days.

Determination of soil and water temperature

In order to examine if temperature affects efficacy of pollutant removal, HOBOWare device was used to measure temperature in water column and soil of treatments during experimental time. Fifteen HOBOWare gauges (Figure 2a) were deployed into the middle of soil layer and water column within each treatment and established to automatically recorded temperature in every 30 minutes. The gauges were mounted on PVC Ф-27 pipes and fasten by plastic tubes. Every two gauges were arranged on the same PVC pipe and aligned in the middle of water column and soil layer of each treatment.

Measurement of salinity, DO, and pH

To measure salinity among experimental treatments, electrical conductivity (EC) was measured, and then be converted EC to salt concentration (Rhoades et al. 1999, NRW Service Centre 2007). Dissolved oxygen (DO), EC, and pH in effluent were measured at the outlet zone of each effluent system in every 2 weeks from 9:00 am to 10:00 am using YSI650 MDS machine (Figure 2b). YSI’s readers were deployed into water column with a depth of 10cm at the outlet zone to measure DO, EC and pH. The readers were set up stabilizing in 60 seconds before recording and displaying results. To ensure QA/QC, the machine was calibrated in one hour before conducting the measurement and checked right after measurement by following the user manual’s standard operating procedures (YSI Incorporated 2009).

Collection and analysis of effluent samples

To evaluate effectiveness of pollutant removal among experimental treatments, samples of effluent water were collected after two and four months. Collection of effluent samples was conducted at the outlet zone of effluent system in the morning by using a mini pump and collected in one day before putting the wastewater. Effluent water was directly pumped into a 2-liter plastic bottle and kept inside an ice cooler before sending to NLU’s Research Institute of Biotechnology and Environment on the same collection day for analyzing BOD5, COD, N, P, Zn, and Cu. Methods applied to analyze for these components are presented in Table 1.

Accumulation of pollutants in soil and plant

Soil and vegetation samples were collected in one day before wastewater was initially filled into barrels corresponded with one month after planting, and at the end of experiment to determine contents of N, P, Cu and Zn in soil and plant biomass. Soil samples were taken at a depth of 5cm from soil surface by a soil core. In each treatment barrel, soil was randomly collected at three different positions, mixed them together, then placed it into a ziplock bag and kept it in an ice cooler before sending to laboratory for analysis. Similarly to soil sampling, plant sample was collected from three plants in three different locations of each barrel. The sample was distinctly divided into leaves and roots. These samples were cut into 2-cm pieces, arranged neatly in paper bags, and dried in oven at 65oC for 7 days before sending for analysis. Analytical methods of N, P, Cu, and Zn applied for plant and soil are presented in Table 1.

An increasing accumulative rate of pollutants in biomass computed by the following equation:

Statistics and ANOVA analysis method

This one-factor experiment was completely randomized factor with three replications for each treatment. Software of Minitab 17 was used for ANOVA analysis to evaluate significantly differences among treatments, and for Tukey’s multiple comparision method (α = 0.05). Charts and Figures were created by using the software of SigmaPlot 12.

Concentrations of influent pollutants

Table 2 shows that concentrations of untreated pollutants in raw pig wastewater were very high. BOD5 and COD were 2687.0 and 31516.0 mg O2/L, whereas concentrations of total nitrogen (TN), total phosphorus (TP), Zn2+ and Cu2+ were 2207.0; 1335.0; 23.4; 15.4 mg/L respectively. Especially the pig wastewater contained a high electrical conductivity (EC) of 8.3 mS/cm, but a very low DO content of 1.8 mg O2/L. Concentrations of these pollutants were much higher than those of previous studies (Chang & Liu 2002, Chung et al. 2004, Nguyen & Pham 2012). The reason for that was because wastewater samples were taken from tunnels of pigsty which directly received waste from pig manures and pigsty washing, and have not been decomposed with a condensed concentration. Based on concentration of the untreated pollutants in the raw wastewater, loading rate established for each treatment, and retention time of wastewater in each treatment, initial influent concentrations of COD, BOD5, TN, TP, Zn and Cu were 806.1; 9455.0; 662.2; 400.6; 7.0; and 4.6 mg/L respectively.

Temperature variation in soil and water column

Temperature is an important ecological factor and directly affects respiration of microorganisms in a treatment system. Variation of daily average temperature in soil and water among treatments were showed in Figure 3. In general, the temperature in soil and water fluctuated from 24 – 31oC. In the first stage of development, there was a temperature difference. There was also a temperature difference among treatments in soil due to young plants, but no difference of those in water. In the final growth period, there was no different temperatures among the treatments in both soil and water. Temperatures of T. domengenis and L. obtusifolium showed lower than those of other treatments. Variations of temperatures among treatments could impact on removal efficiency of pollutants (Ab-Halim et al. 2016; Alisawi, 2020; Shatat and Al-Najar, 2011)

DO, EC and removal efficacy of salt

Dissolved oxygen (DO) in water is dependent on photosynthesis processes of aquatic plants and a microbial respiration in water column and soil. In the respiration, microorganisms consume a large amount of DO to decompose organic compounds. An amount of DO used for the microbial respiration are often supplemented by photosynthesis of aquatic vegetation. The low level of dissolved oxygen in water is not only a sign of contamination and is but also an important factor in determining water quality and pollution (Bozorg-Haddad, 2021; Kulkarni, 2016). In freshwater ecosystems, most of aquatic animals require amount of dissolved oxygen higher than or equal to 3 mg O2/L for survival. For example, fish cannot survive for long in water with dissolved oxygen less than 5 mg/L (Bozorg-Haddad, 2021). Table 2 shows that the pig wastewater contained a large amount of organic compounds with a low DO of 1.8 mgO2/L. Althought DO concentrations of each treatment fluctuated in over the experimental periods, two treatments of T. domengenis and E. dulcis generally improved the DO (Figure 4). DO concentration of these two treatments was generally ranged from 5 to 9 mg O2/L.

Most aquatic plants and animals require oxygen to survive; fish, for instance, cannot survive for long in water with dissolved oxygen less than 5 mg/L. The low level of dissolved oxygen in water is a sign of contamination and is an important factor in determining water quality, pollution control and treatment process. The DO in a saturated solution varies with the water temperature and elevation.

Electrical conductivity (EC) shows concentration of ions which exist in water is usually derived from the soluble salt ions such as Na+, K+, Cl-, SO42-. A high content of soluble ions would negatively influence and inhibit a growth of aquatic organisms. Salinity in fresh water normally ranges from 0.01 to 0.50‰. The pig wastewater contained a large amount of soluble salt ions with EC of 8.3 mS/cm (Table 2) which corresponded to salinity content of 4.6‰, ten times higher than the normal range of salinity in natural freshwater ecosystems. Due to no vegetation, EC in control treatment accumulatively increased following the treated time resulted in a gradual accumulation of the salt content during the experiment (Figure 4). Species of T. domengenis, L. obtusifolium, and E. dulcis performed a high absorption of salt ions which resulted in a very low EC and salinity. Salinity of these treatments was always lower than 0.50‰ over 14 weeks of the treated experiment (Figure 4). Exception for E. dulcis, since this species grew very quickly and had its shorter growth cycle than those of others, about 10 weeks, when this species died and returned its turnovers into water, and decomposed turnovers quickly (Marois et al. 2015, Mitsch et al. 2015). This resulted in an increasing of EC and salinity for this treatment in the last stage of the experiment.

Removal efficacy of BOD5, COD, N, P, Zn and Cu

As showed in Table 2, the pig wastewater contained a high concentration of organic matters with COD of 31,516 mg O2/L and biochemical oxygen demand (BOD5) of 2,687 mg O2/L. Based on the input load, the retention time and the total volume of each treatment system, initial influent concentrations of COD and BOD5 in all treatments were 9455.0 and 806.1 mg O2/L respectively (Table 2). Sixty days after treated, three treatments of L. obtusifolium, E. dulcis, and T. domengensis removed a large amount of organic matters, and as a result, the COD contents of these treatments reduced to 55.5, 36.0, and 176.0 mg O2/L respectively (Figure 5). But in 120 days after treated, the COD content of L. obtusifolium treatment continued dropping to 46 mg O2/L, whereas the COD of E. dulcis treatment went up to 560 mg O2/ L, which was explained by due to the 10-month growth cycle of E. dulcis species. Similar to effectiveness of COD handling, the treatments of L. obtusifolium, and T. domengensis showed the best efficacy of BOD5.

Pig wastewater was polluted with high N and P, where the concentrations of TN and TP are respectively 2207.0 and 1335.0 mg/L, while their influent concentrations at each treatment were 662.0 mg/L (TN) and 400.6 mg/L (TP) (Table 2). Decomposition of organic pollutants by microorganisms associated with treatments of L. obtusifolium, E. dulcis, and T. domengensis reduced the BOD and COD. This resulted in decomposition and transformation of organic N and P compounds into inorganic compounds for plant uptakes. These three treatments showed the best performance of N and P removal after 60 days treated. The TN contents of L. obtusifolium, E. dulcis, and T. domengensis were respectively 2.4, 10, and 3.3 mg/L. But after treated in 120 days, TN contents of all treatments increased more than doubled in comparison with those at the treated time of 60 days. Similarly to total N, the content of total P (TP) was also removed significantly after 60 days and increased after 120 days, except for two treatments of L. obtusifolium and T. domengensis (Figure 5).

Cu and Zn ussualy supplements to food of livestocks, especially in pig foods, but pollution of Cu and Zn in pig wastewater is paid less attention. Concentrations of these two heavy metals in the pig wastewater are so high, approximately 23.4 and 15.4 mg/L for Cu and Zn. Based on input load and retention time on the treatment system, the initial average concentration of Cu and Zn were 4.6 and 7.0 mg/L (Figure 5). Removal capabilities of Cu and Zn are very different among treatments. The treatment of L. obtusifolium removed completely Cu and Zn due to plant uptake. As a consequence, concentrations of Cu and Zn in this treatment was nearly equal to zero (non detection) after 60 days treated. In addition to the L. obtusifolium treatment, the A. aureum also showed a higher efficiency of Cu and Zn removal than others.

Accumulation of N, P, Zn, Cu in Soil

According to results of ANOVA analysis shown in Table 3, the initial N and P contents in soil were not different from the treatments (p = 0.186 for TN, p = 0.669 for TP). Concentration of initial soil TN ranged from 284 to 323 mg N/kg dried soil, whereas the initial TP concentration in soil ranged from 62 to 67 mg P/kg dried soil. Baseline concentrations of TN and TP used for the experiment were very low. However, both N and P concentrations in the soil were increased for all treatments. Due to without vegetation, the control treatment accumulated the lower than those of three other treatments because of its ability of low N transformation and decomposition in water, while the treatment of L. obtusifolium had a the highest accumulation of TN and TP contents in soil with 504 mg N/kg and 210 mg P/kg, respectively. This result were corresponded with the lowest TN and TP concentrations of L. obtusifolium treatment in water. Initially baseline Cu and Zn concentrations in the soil utilized for the experiment were very low and did not differ among treatments with the Cu concentration under non-detection zero and Zn concentration ranged from 3.1 to 3.9 mg/kg (p = 0.822) (Table 3). At the end of the experiment, although accumulation of Cu content in soil increased in all treatments, approximately 6 mg Cu/kg, there was no significant difference (p = 0.157). In contrast to Cu accumulation in soil, Zn concentrations were significantly different from the treatments (p < 0.000). The treatment of L. obtusifolium showed the highest Zn accumulation with 11.4 mg Zn/kg. This is resulted from the high Zn removal efficiency of L. obtusifolium treatment in water which assisted by organic matter breakdown associated with microorganisms in this treatment.

Contents of N, P, Cu and Zn in plant biomass

Since the initial N baseline content in foliages and roots are significantly different among treatments (Table 4). The treatment of L. obtusifolium contained the highest N concentrations in root (0.97% ± 0.15) and foliage (1.90% ± 0.10). This difference could be explained that because in first month after planting (not input the wastewater to treatments), the L. obtusifolium species grew faster than others treatments resulting in higher accumulation of N content in its root and foliage. This corresponded with the lowest initial N content of L. obtusifolium treatment in the soil (Table 3). After treated in 120 days, root and foliage N concentrations of all treatments significantly increased in comparison with their initial baseline concentrations (Table 4). The increasing percentage of N absorption rate by vegetation treatments within root and foliage were presented in Figure 6. All treatments showed a high N absorption rate in roots with greater than 50%, but there was no a significant difference of the rate among the treatments. The highest N absorption rate in roots reached 67% for the A. aureum treatment while the lowest N absorption rate was 50% for the L. obtusifolium treatment. On the other hand, the L. obtusifolium treatment showed the highest foliage N absorption with the 28.9%, followed by the treatment of E. dulcis (19.6%), from which the rates of two these treatments were significantly different from two other remaining treatments. If combined with the root and foliage absorbed N contents, the treatment of L. obtusifolium performed the highest capacity to accumulate nitrogen in its biomass. Generally, under a normal condition, all treatments accumulated lower N in roots than in foliages, but with the addition of N increasing amount from pig wastewater, N tended to be accumulated more in roots than in foliage in order to create a balance of N content in their biomass.

Similar to N status accumulated in biomass, the initial root and foliage P concentrations were very different among treatments. The treatment of L. obtusifolium always performed the best ability to uptake P into its biomass, from which the initial P baseline concentration were 0.329% ± 0.004 in root and 0.339% ± 0.094 in foliage and that the P concentration accumulated in its roots and foliage after 120 days were respectively 1.078% ± 0.017 and 0.486% ± 0.017 (Table 4). In contrast to N absorption results, phosphorus mainly accumulated in roots of treatment’s plants after the 4-month handling period. If it takes into account of absorbed P rate in biomass, the treatment of T. domengensis showed the best performance of P absorption rate in its biomass, about 68% in root and 62% in foliage (Figure 6).

The N/P ratio in plant biomass is an ecological indicator, which is used to assess a level of excess or deficiency of N or P content provided for growth of trees in different ecosystems. N/P ratio of 14-15 reflects a balance between N and P contents in an ecosystem. The N/P ratios in biomass of all treatments in this study were rather low, approximately 3 in root and 8 in foliage for treatments (Figure 6). These results indicated that the pig wastewater was more N polluted than P in a term of N/P ratio for plant uptake.

According to the analyzed results presented in Table 4, both Cu and Zn were only accumulated in roots of treatments. Cu and Zn concentrations in foliage were very low and even under detectable in some treatments. At the same rate of initial Zn and Cu influent loading, two treatments of T. domengensis and L. obtusifolium accumulated the highest Zn content in their roots, 154.0 ± 15.0 and 154.3 mg/kg ± 43.8 respectively (Table 4), where only the treatment of L. obtusifolium showed the highest Cu absorption in its roots with Cu concentration of 80.0 mg/kg ± 6.