Nitrogen and phosphorus fluxes in three soils fertigated with decentralised wastewater treatment effluent to field capacity

The Decentralised Wastewater Treatment System (DEWATS) provides low cost onsite sanitation to residents living in informal settlements. Wastewater management through agriculture prevents environmental pollution and promotes sustainable agriculture. This study investigated the effects of fertigation with DEWATS effluent to field capacity in three South African soils under a banana crop. The experiment was conducted as a complete randomised design in a greenhouse with two irrigation water treatments (DEWATS effluent vs municipal tap water irrigationþ fertiliser) × three soil types (Ia, Cf and Se) and four replicates over 728 days. Data were collected on crop growth, nitrogen (N) and phosphorus (P) uptake and dynamics in the soil. The DEWATS effluent significantly (p< 0.05) increased N and P uptake and soil NH4 -N and extractable P concentrations. Furthermore, DEWATS effluent fertigation significantly (p< 0.05) increased N leaching from the Ia soil and P leaching from the Cf soil. Nitrogen and phosphorus leaching from DEWATS was lower than the tap water irrigationþ fertiliser treatment. There was, however, excess N and P accumulation from the DEWATS than the irrigationþ fertiliser treatment, which would cause environmental concerns from runoff and leaching losses in the medium to long term. This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/). doi: 10.2166/wrd.2019.025 om https://iwaponline.com/jwrd/article-pdf/9/2/142/553803/jwrd0090142.pdf 2020 W. Musazura (corresponding author) A. O. Odindo Crop Science Discipline, University of KwaZulu-Natal, P. Bag X01, Scottsville 3209, South Africa E-mail: wmusazura@gmail.com E. H. Tesfamariam University of Pretoria, P. Bag X20 Hatfield, Pretoria 0028, South Africa J. C. Hughes Soil Science Discipline, University of KwaZulu-Natal, P. Bag X01, Scottsville 3209, South Africa C. A. Buckley Pollution Research Group, Chemical Engineering, University of KwaZulu-Natal, Durban 4041, South Africa


INTRODUCTION
Municipalities in South Africa are considering provision of proper sanitation to all residents, including those in informal settlements, in a move towards the fulfilment of the millennium development goals (MDGs) (Roma et al. ).
The eThekwini (Durban) municipality in KwaZulu-Natal (KZN) commissioned community ablution blocks that can be connected to a decentralised wastewater treatment system (DEWATS) as an interim solution to sanitation problems (Crous et al. ). The DEWATS is a modular water-borne sanitation system which consists of the settler, anaerobic baffled reactor (ABR) þ anaerobic filter (AF) and planted gravel filters (Gutterer et al. ). The treatment process involves anaerobic degradation of organic matter within the ABR followed by the AF. The AF effluent is further passed to planted gravel filters which consist of vertical flow constructed wetland (VFCW) and horizontal flow constructed wetland (HFCW) for further polishing.
The final effluent must comply with the stringent South African DWA () discharge standards hence any failure to the wetland may lead to discharge of poorly treated wastewater.
The use of treated wastewater in agriculture has been recommended as a major way to fulfil MDG number seven of fighting against hunger (WWAP ). For an effective wastewater use programme in agriculture, practical guidelines that will be used to inform policy makers on how to maximise benefits and mitigate risks must be developed (Pescod ). Practical guidelines consider technical aspects such as land area requirements, effluent management in different seasons and environmental sustainability in different soils (Pescod ; USEPA ).
Effluent production occurs throughout the whole year and crop water requirements are variable with seasons. Therefore, crops that can fully utilise all the water and nutrient from effluent and irrigation methods that allow soils to absorb all the effluent produced are required.

Experimental site
The study was conducted under controlled conditions in a growing tunnel (greenhouse) at Newlands-Mashu Research Centre, Durban, KwaZulu-Natal, South Africa (29 58 0 S; 30 57 0 E). Durban is in the east coast of South Africa and experiences cool dry winters and hot wet summers. Soil physical properties were analysed before planting while chemical properties were analysed before planting and after harvest (728 days after planting). Bulk density was determined from undisturbed soil cores collected from a depth of 0-300 mm. The field capacity and permanent wilting points for the respective soils were calculated based on particle size using a calculator from the SWB Sci model (Annandale et al. ) ( Table 1). Soils were air dried, ground and sieved to pass through a 2 mm mesh.

Soils and analyses
A representative sub-sample of each soil was analysed for soil properties chemical and physical properties at the Soil Inorganic N (NH þ 4 -N and NO À 3 -N) was determined from freshly collected soil samples by extraction in 1:5 soil: 2M KCl and filtering using Whatman® No. 2 paper following methods by Mynard & Kalra () and analysed using Merck Nova 60 Spectroquant® (Merck Millipore, Germany) following standard methods (APHA ). Phosphorus was extracted from freshly collected soil using the Ambic 2 solution followed by filtering using Whatman® No. 1 paper.
Phosphorus was then determined from the filtrate using the molybdenum blue procedure following standard methods (Non-Affiliated Soil Analysis Work Committee ).

Experimental design management practices
A 2 × 3 × 4 factorial experiment was carried out in a complete randomised design. The experiment comprised of two irrigation treatments (DEWATS effluent vs municipal tap water irrigation þ fertiliser) × three soil types (Cf; Typic Haplaquept, Ia; Rhodic Hapludox, Se; Aquic Haplustalf) × four replicates. Inorganic fertiliser was applied to the tap water þ fertiliser treatment soils; they were mixed with urea (46% N), single superphosphate (10.5% P) and potassium chloride (52% K) based on soil analysis recommendations (Table 2). Dolomitic lime was added at a rate of 1.03 g kg -1 to the Ia and Cf soils to adjust soil pH to a permissible acid saturation of 1%. The soils had different bulk densities (Table 1)

Effluent characterisation
For the first 210 days after planting (3 April-29 October 2015), the pots were fertigated with DEWATS effluent from the horizontal flow constructed wetland (HFCW).
Thereafter the effluent used was that obtained after the AF of the DEWATS. Effluent chemical oxygen demand (COD), suspended solids, pH, and nutrients (NH þ 4 -N, NO À 3 -N and P) were monitored throughout the growing Extractable Cu (mg kg -1 ) 3.6 0.2 9.5  (2)): where TDM ¼ total dry biomass of the whole plant (g); DM ¼ plant tissue dry mass (%) for each plant part; FM ¼ fresh biomass (g) for each plant part.
Samples for plant tissue nutrient analysis were collected from the third uppermost leaf after harvest. Plant tissue samples were oven dried at 70 C for 72 hours. Dried plant tissues were then crushed and sieved through a 1 mm sieve.
The leaf tissues were analysed for total N using the LECO ® TruSpec Micro CNS analyser and P using the acid digestion method followed by inductively coupled plasma optical emission spectroscopy (ICP-OES) Vista MPX following standard methods (Riekert & Bainbridge ).

Nutrient leaching and drainage
Sampling of leachates commenced 181 days after planting.
Leachates were collected from the WFDs at random periods four hours after irrigation and analysed for NH þ 4 -N, NO À 3 -N and PO 3À 4 -P using a Nova 60 Merck Spectroquant ® (Merck Millipore, Germany) according to standard methods (APHA ). Soil drainage rates were quantified by measuring the volume of water leached 4 hours after random irrigation events.

Data analysis
All data were analysed using GenStat ® 18th edition (VSN International, UK). The data were subjected to analysis of variance (ANOVA) and standard error of mean differences were used to separate differences between means at the 5% significance level.

Effluent characterisation
The N and P concentrations of effluents used during the study are reported in Table 3.

Crop growth and yield
The interaction between soil type and irrigation treatment on banana plant height, total leaf area, fresh and dry biomass are presented in Table 4. The plant height and total leaf area were significantly high in Se compared to other soils for both irrigation treatments (Appendix 1, available with the online version of this paper). These plant growth variables were also comparable between the two irrigation treatments under Ia soil as well as to Cf soil fertigated with DEWATS effluent.
Least plant height and total leaf area were reported in Cf soil in tap water þ fertiliser treatment.
The fresh and dry biomass of banana measured at harvest (728 days after planting) are also reported in Table 4. Both fresh and dry biomass were significantly low in Cf soil under tap water þ fertiliser treatment compared to other soil and irrigation treatment combinations (Appendix 2, available online). Highest fresh and dry biomass was recorded in Se soil under tap water þ fertiliser treatment.
Generally speaking, fresh and dry biomass under tap water þ fertiliser treatment was significantly higher than DEWATS treatments planted to similar soil types. The only exception was for Ia soil, which was not statistically significant but was still higher under tap water þ fertiliser.

Soil chemical properties
There was a significant (p < 0.01) interaction between irrigation treatments and soil type on soil NH þ 4 -N content (Appendix 3, available online). Irrigation treatments significantly differed (p < 0.01) with respect to extractable P (Appendix 3). The NH þ 4 -N and extractable P concentrations in three different soils and irrigation treatments are described in Figure 3. Fertigation with DEWATS effluent significantly increased NH þ 4 -N content in all soils compared to tap water þ fertiliser treatment. The least NH þ 4 -N concentrations values were found in the Cf and Se soils under the tap water þ fertiliser treatment.

Nitrogen and phosphorus leaching and drainage
There were significant differences in P leached between the three soils (p < 0.05) see Appendix 4 (available online).
A significant interaction (p < 0.001) between soil type and irrigation treatment on N leaching over time was also reported (Appendix 4). The amount of P leached from each pot amongst the three soils is shown in Figure 4. High P was leached from Cf soil compared to both Ia and Se.
The interaction between soil type and irrigation treatment over time on inorganic N leached is shown in Figure 5. Very high N leaching occurred in Se soil under  the tap water þ fertiliser treatment compared to DEWATS effluent. Comparisons amongst different soils within the DEWATS effluent treatment showed that N leaching was higher in Ia than the Se and Cf soils.

Irrigation and nutrients
The quantities of N and P supplied through fertigation using DEWATS effluent in relation to the crop fertiliser requirements are shown in Table 5. Fertigation using DEWATS effluent to maintain soil field capacity added excessive N and P, more than was required by the crop.
There was a significant difference (p < 0.001) in P uptake between soils and N and P uptake between the irrigation

Crop growth and yield
The DEWATS effluent increased banana vegetative growth (plant height, dry mass and leaf area) in the Cf soil, although highest growth occurred in the Se soil (Table 4). This was due to nutrients supplied through fertigation (Table 5) and their subsequent uptake by crops (Table 6), which agreed with several studies using the same type of wastewater (Bame et al. ). Banana yield could not be determined due to delayed and erratic flowering exceeding the experimental time frame, probably due to excess N from the effluent (Table 5).

Soil chemical properties
High soil P content in Se soil compared to Ia and Cf reported in Figure 3 was probably due to low drainage of the soil and retention by soil Al/Mn/Fe minerals. According to findings by Bame et al. (), Ia soils retain more P due to their high organic matter content while Cf loses more due to its course texture, but Figure 3 showed that P content was comparable between Ia and Cf soils. Comparisons between the irrigation treatments showed that soil P content significantly increased in the DEWATS treatment compared to tap water þ fertiliser treatment regardless of soil type ( Figure 3). Therefore, fertigation with DEWATS effluent to field capacity increased soil P content regardless of soil type. Such excess accumulation of P above crop nutrient requirements warrants for DEWATS effluent application according to crop nutrient requirement instead of crop water requirement.
The NH þ 4 -N concentrations increased significantly in all soils under DEWATS effluent fertigation (Figure 3). This is expected in soils with high cation exchange capacity (CEC) due to adsorption by the soil colloids as reported by some authors (Bame et al. ; Hernández-Martínez et al. ). On the other hand, NH þ 4 -N content also increased in the low CEC Cf, probably due to increased fertigation which applied more N into soils (Table 5). This could lead to enhanced volatilisation, especially at soil pH exceeding 7 (Dendooven et al. ). The pH values of all soils used in this study ranged between 4.11 and 5.20 (Table 1), hence pH driven volatilisation losses are expected to be very low.

Nitrogen and P leaching
The leaching of P was high in Cf compared to the other two soils (Figure 4) due to low P sorption capacity of sandy soils.
High organic matter in Ia soils and clay loam soils (Se)  High amounts of N were leached from the tap water þ fertiliser treatment on the Se soil at 181 DAP ( Figure 5) due to fast hydrolysis of the urea fertiliser. In DEWATS effluent fertigated soil, the low N leaching losses from the Se and Cf soils compared to the Ia were probably due to the lower N content in these soils (Table 1). According to Egiarte et al.
(), high concentrations of NO À 3 in leachates results from nitrification, especially in acidic soils. Therefore, high N leaching from the Ia soil (DEWATS) was likely caused by fast nitrification resulting from acidity of that particular soil, as also reported by Bame et al. ().

Irrigation and nutrients
High banana leaf tissue N and P concentrations in DEWATS effluent treatment (Table 6) are directly linked to nutrients applied through fertigation (Table 5)

CONCLUSIONS
Crop growth significantly increased in Cf soil fertigated with DEWATS effluent. Fertigation with AF effluent up to soil field capacity loaded more N and P to the soil, which even exceeded crop fertiliser requirements. Soil extractable P and NH þ 4 -N increased significantly in all DEWATS effluent fertigated soils. Soil P leaching differed between soils, Cf soil losing more compared to Ia and Se. There was a significantly high N leaching in tap water þ fertiliser treatment than in DEWATS effluent treatment. Therefore, the use of DEWATS effluent to fertigate banana according to crop water requirement may potentially lead to excess accumulation of N and P in the soil profile which could eventually enhance leaching below the root zone. This warrants for crop nutrient requirement based DEWAT effluent application under the given climatic conditions and soil types for sustainable recycling of resource. Nitrogen leaching differed amongst three soils under DEWATS effluent fertigation, highest leaching was reported in Ia soil compared to other soils. The banana leaf tissue N and P concentrations were significantly higher in DEWATS effluent compared to tap water þ fertiliser implying that banana plants may benefit with nutrients supplied by the effluent.
Care must be taken, especially in high drainage soils such as Cf and Ia, where irrigation scheduling with room for rainfall can be opted to prevent N and P leaching. Considering that the study was conducted under controlled conditions, further investigations are recommended at field scale to accommodate various climatic zones, soil forms and crop types.