Domestic Sewage Effluent Increases Plant Growth, Yield, and Fiber Quality of Cotton Cv. BRS 335

ABSTRACT Semi-arid regions have low water availability, and water reuse in irrigation is an alternative to enable satisfactory agricultural yields in the Brazilian semi-arid region. To evaluate the effects of irrigation with sewage effluent and phosphate fertilization on upland cotton (Gossypium hirsutum L. cv. BRS 335), we irrigated the cotton plant only with well water (control) or irrigated it with a solution containing 50 and 100% effluent sewage, in soil with and without pre-planting phosphate fertilization. For plant growth analysis, we considered time as the third factor. The experimental design consisted of randomized blocks, with six replicates, during two cotton cycles in the open field. Plant growth, yield, and fiber quality increased with the application of sewage effluent compared to the control. The control obtained cotton fiber yields of 740 and 741 kg ha−1 in the first and second cycles, using 50% sewage effluent outperformed 240 and 290% of the control and using 100% sewage effluent outperformed 235 and 300% of the control, respectively. Pre-planting phosphate fertilization failed in upland cotton cv. BRS 335. Irrigation using 100% sewage effluent promotes more significant plant growth, production, and quality of cotton fiber.


Introduction
Cotton is a crop of great economic importance globally, and the textile industry highly values it due to its high fiber quality (Yao, Yang, and Liu 2006). However, environmental conditions and the genetic potential of cultivars influence this crop's productivity and fiber quality. Zonta et al. (2015) found the highest cotton yield to FiberMax 993, BRS 286, BRS 336, and BRS 335 cultivars, with 130% ETc (1179 mm), but it is possible to obtain good water efficiency with 70% ETc (635 mm). Thus, highly efficient

Characterization of the experimental area
The experiment was conducted in the field of the Milagres settlement project, Apodi, Rio Grande do Norte, Brazil (5°35"18.82""S latitude and 37° 54" 08.48'' W longitude and an altitude of 109 m). The prevailing climate is Tropical of the Equatorial Zone, with three subtypes of semi-arid climate (mild, medium, and severe). Regarding climatic parameters, the region has an average annual temperature of 28.5°C, average insolation of 3,041 hours year −1 , average evaporation of 2,190 mm year −1 , average relative humidity of 66.8%, and average precipitation of 767 mm year −1 (INMET: National Meteorological Institute).
The soil of the experimental area has a flat relief with a dominant slope of less than 2% and was classified as Inceptisol. Soil is Loamy sand on the surface and Sand clay loam in the subsurface. It has a 1.68 kg dm −3 soil density, 36.6% porosity, 13.42 mm total water storage capacity, and low initial phosphorus content (7.0 mg dm −3 ). Table 1 shows nitrogen, potassium, phosphorus, calcium, magnesium, and organic matter contents in the soil at the beginning of the experiment, after the first and second years of cultivation.
The soil was irrigated with tubular well water and domestic sewage effluent from the water treatment plant installed in the Milagres Rural Settlement, Apodi, RN, Brazil. Two water sources were stored in two tanks, pressurized by two 1.5 hp electric pump sets. Lemos et al. (2021) show details on the sewage treatment plant. Samples from the water sources were collected once a month for physical-chemical and biological characterization (Table 2), following the recommendations described in the "Standard Methods for the Examination of Water and Wastewater" of the American Public Health Association -APHA (Rice, Baird, and Clesceri 2012). Table 1. Nitrogen (N), phosphorus (P), potassium (K+), calcium (Ca 2+ ), magnesium (Mg 2+ ), and organic matter (O.M.) contents in the soil at the beginning of the experiment, after the first and second years of cultivation.  *Mean values and standard deviation (SD), WW -well water, TDE -Treated domestic effluent, pH(H 2 O)-Hydrogen potential in water, EC -Electrical conductivity, SAR -Sodium adsorption ratio, NH 4 +-N -ammoniacal nitrogen, NO 3 -N -nitric nitrogen, OGC -oil and grease content, BOD -biochemical oxygen demand, and COD -chemical oxygen demand.

Experimental design, treatments, and plant material
During two cotton cycles in the open area, Cotton cv. BRS 335 was cultivated in the open field using an experimental design of randomized blocks with six replicates. We irrigated the cotton plant with three water sources (T 1 = well water -control, T 2 = treated domestic sewage effluent diluted by 50% in well water, and T 3 = raw effluent of treated domestic sewage) soil with and without phosphate fertilization at pre-planting. For plant growth analysis, we considered time as the third factor. Each experimental unit consisted of four rows in double spacing (1.10 m × 0.5 m × 0.1 m) with 8.5 meters in length, totaling an area of 27.2 m 2 (3.2 m × 8.5 m), considering 7.0 m of the two central rows as the usable area, disregarding 0.75 m at the ends. Phosphorus dose was added to the soil and incorporated in the first 0.05 m deep. We obtained a phosphorus dose of 100 kg ha −1 of triple superphosphate (46% P 2 O 5 -) for maximum crop yield, according to Ribeiro, Guimarães, and Alvarez (1999), for soils with a low initial phosphorus content.
The irrigation depth was calculated according to crop evapotranspiration (ETc) using Equation (1). The reference evapotranspiration (ETo) was calculated by the Penman-Monteith equation parameterized by FAO (Allen et al. 2006). The climatic data were obtained at the weather station of the National Institute of Meteorology (INMET) (Apodi -RN) and based on the crop coefficient presented in the FAO 56 Manual (Allen et al. 2006), with a 2-day irrigation interval. At the end of each day, the climatic data necessary to feed the irrigation worksheet were collected and used to determine the irrigation depth of the following day (Equation 2). When cotton plants had 60% of open bolls, irrigation was suspended.
Where: L -irrigation depth (mm); ETc -Crop evapotranspiration (mm); f -Application efficiency (considered equal to 0.85, the value of the distribution uniformity coefficient -DUC).
Irrigation was performed using a modified bubbler; a low-pressure system developed to avoid clogging of biological origin when using domestic sewage effluent (Hao et al. 2017). In the system, microtube emitters with a 3 mm internal diameter were used to avoid clogging, not requiring sophisticated filtration systems.
The irrigation system consisted of both primary and secondary lines, PVC pipes, with diameters of 32 and 50 mm, respectively. The polyethylene lateral lines of 16 mm in diameter were installed with a spacing of 3.2 meters between them. 2.0-m-long microtube emitters were connected to the lateral lines to feed the two furrow lines.
All treatments applied irrigation with well water before planting until soil moisture was close to field capacity. Then, upland cotton (cultivar BRS 335) was sown by planting 10 seeds per linear meter at a 3 cm depth and alternated on the sides of the irrigation furrow in double rows (1.10 m × 0.5 m × 0.1 m), totaling a population of 125,000 plants ha −1 . When germination stabilized, treatments with irrigation solution started only 10 days after planting (DAP). Weeds were controlled using a hoe twice, at 15 and 25 days after germination (DAG). Pests were sampled weekly, and chemical control was carried out when necessary. Table 3 shows agronomic data and irrigation parameters during the cotton cultivation cycles of 2016 and 2017.

Growth, production, and fiber quality
We evaluate plant growth, production, and quality of cotton fiber. Plant height and diameter were measured over two cultivation cycles, and the measurements were always performed on the same three plants of each plot at 27, 44, 64, and 78 days after germination (DAG) in the first cycle and at 27, 43, 57, 71, and 85 DAG in the second cycle.
Yield (Y) was determined based on the fiber percentage (FP) of all the bolls collected in the usable area of the experimental units. We determine the percentage and quality of the fiber in each cultivation cycle. A sample of 20 bolls was collected from each experimental unit. The bolls were analyzed using HVI (High Volume Instruments) model 900 from USTER, according to NBR ISO 139:2008 (standard for conditioning and testing) of the Brazilian Association of Technical Standards (ABNT), which are: temperature 20 ºC (±2) and humidity 65% (±2). We have obtained the fiber quality characteristics: UHM -fiber length (mm), UNF -fiber uniformity (%), SFI -short fiber index (%), STR -strength at rupture (g tex −1 ), ELG -elongation at rupture (%), MIC -micronaire index (µg in −1 ), MAT -maturity (LORD), RD -reflectance (%), +B -degree of yellowing, and CSP -count-strength product.

Statistical analysis
The data obtained were subjected to analysis of variance by the F test. When significant, the means were compared by the Tukey test at the 5% probability level. All analyses were performed with the free statistical software SISVAR® version 5.6.

Cotton growth
There was a significant effect (p <.01) of the interaction between irrigation with sewage effluent and time on the variables plant height (PH) and stem diameter (D) in the two cultivation cycles (Table 4).
In the first cycle, the highest values of plant height were verified in plants from T 3 at 44, 64, and 78 DAG, and at 27 DAG plants in the treatments, T 2 and T 3 obtained higher plant height than those in T 1 (Table 5). In the second cycle, the highest values of plant height were verified in the treatment T 3 at 43, 57, 71, and 85 DAG. At 27 DAG, the treatments T 2 and T 3 had higher plant height than T 1 (Table 5).
For stem diameter in the first cycle, the highest values were obtained in T 3 at 27 DAG, and at 44, 64, and 78 DAG, the treatments T 2 and T 3 obtained higher stem diameters than T 1 (Table 6). In the second cycle, the highest values of stem diameter were verified in plants from T 3 at 27, 43, 57, 71, and 85 DAG (Table 6).
In cotton growth, phosphate fertilization was significant (p < .05) for stem diameter in the first cycle, and phosphate fertilization failed in the other cotton growth variables (Table 4). Phosphate    fertilization increased the stem diameter of cotton by 6% in the first cycle, but in the second cycle, phosphate fertilization failed to stem diameter (Table 4).

Cotton fiber production
There was a simple effect (p <.01) of irrigation with a solution containing sewage effluent on the variables yield and fiber percentage in the two cultivation cycles (Table 7). Sewage effluent irrigation increased cotton fiber yield (Y) in both cultivation cycles compared to the control (T1) ( Table 7). The control obtained a cotton fiber yield of 740 and 741 kg ha −1 in the first and second cycles, respectively (Table 7). For 50% of the effluent in the irrigation solution (T2), cotton fiber yield values were 240 and 290% higher than the control in the first and second cycles, respectively (Table 7). In total irrigation sewage effluent (T3), cotton fiber yield values were 235 and 300% higher than the control in the first and second cycles, respectively (Table 7). Cotton FP did not differ significantly in the two cultivation cycles for both solutions and control. On the other hand, adding 50 and 100% of effluent in the irrigation solution (T 2 and T 3 ) reduced cotton FP by about 1.8 and 3.2%, respectively, compared to the control (T 1 ). Between T 2 and T 3 , FP decreased by 1.4%, and this difference was significant (Table 7).

Cotton fiber quality
The treatments influenced the fiber quality of upland cotton. In the first cycle, irrigation had simple effects on STR, ELG, and CSP and phosphate fertilization on SFI and ELG. At the same time, for MIC, the interaction between the factors was significant (Table 8).
In the second cycle, the interaction between the factors was nonsignificant for any variable studied. However, there were simple effects of the irrigation factor on the variables SFI, STR, and RD and phosphate fertilization on +B (Table 9).
Irrigation with treated domestic sewage effluent increased the variables STR, ELG, and CSP of cotton fiber compared to the control treatment (irrigation with well water) in the first cultivation cycle and the variables STR and RD in the second cycle (Tables 8 and 9). However, there was variation in the response of the variables to the effects of sewage effluent because for STR and CSP in the first cycle and RD in the second cycle, and higher mean values were obtained in T3 (100% sewage effluent). In contrast, for ELG in the first cycle and STR in the second cycle, higher values were recorded in plants of T2 (50% sewage effluent). On the other hand, the SFI variable in the second cultivation cycle decreased significantly with the increase of effluent in the irrigation solution, and the lowest mean value was observed in T3 (Tables 8 and 9). Regarding the analysis of the interaction for MIC in the first cycle, it was found that cotton plants fertilized with phosphorous at pre-planting did not differ significantly between the irrigation solutions tested. However, for plants without phosphate fertilization, the mean value of MIC obtained in T 1 was higher than that obtained in T 3 (irrigation with 100% sewage effluent). Within the irrigation solutions, Table 8. F test and Tukey test for fiber length (UHM), fiber uniformity (UNF), short fiber index (SFI), strength at rupture (STR), elongation at rupture (ELG), Micronaire index (MIC), maturity (MAT), reflectance (RD), degree of yellowing (+B), and count-strength product (CSP) of the fiber of upland cotton subjected to irrigation solutions with sewage effluent and with and without phosphate fertilization, in the first cultivation cycle.  29.7 a 85.8 a 6.8 a 31.9 a 5.5 a 4.4 a 0.9 a 84.1 a 9.6 b 3008 a Without P 29.8 a 86.0 a 6.6 a 32.7 a 5.7 a 4.4 a 0.9 a 83.8 a 9.8a 3087a ** = 1% significance level (p < .01); * = 5% significance level (p < .05); ns = not significant (p > = 0.05). Different letters indicate differences by the Tukey test at a 5% probability level (p < .05). Uppercase letters in columns compare the means as a function of the effects of irrigation. Lowercase letters in the columns compare the means of the phosphate fertilization management.

Sources of variation
there was no significant difference in the treatments with sewage effluent (T 2 and T 3 ), while in T 1 , a higher mean value of MIC was obtained in the absence of phosphate fertilization at pre-planting (Tables 8 and 9). Regarding the effects of phosphate fertilization on the quality of cotton fiber, it was found that in the absence of phosphorus at pre-planting, the values of SFI and ELG in the 2016 cultivation cycle and +B in the 2017 cultivation cycle were higher than those obtained in the experimental units that received phosphate fertilization (Tables 8 and 9).

Discussion
In semi-arid regions, water and fertilizer management study are necessary for agricultural sustainability Lira et al. 2021). In these regions, water shortages for irrigation jeopardize the development of crops tolerant to water deficits, such as cotton (Soares et al. 2018(Soares et al. , 2020. We studied the domestic sewage effluent for cotton irrigation, with and without phosphate fertilization. We verified that the domestic sewage effluent is viable for cotton irrigation, and we found that cotton cv. BRS 335, irrigated with domestic sewage effluent, has no responsibility for phosphate fertilizer. Irrigation with sewage effluent (T 3 , 100% ETc) increased the growth of cotton from 43 DAG. Sewage effluent met cotton's water demand, increasing cotton growth compared to well water. Sewage effluent increases cotton growth due to the nutrient supply, mainly nitrogen. However, phosphate fertilization only affected the stem diameter in the first cultivation cycle. Thus, phosphate fertilization failed in the growth of cotton for study conditions. Santos et al. (2016) found that irrigated cotton with effluent grew similar to cotton that received top-dressing mineral fertilization.
In both cultivation cycles, the mean cotton yield irrigated with sewage effluent (T 2 and T 3 ) was higher than the average obtained in Brazil in the same evaluation period (1,629 kg ha −1 in the 2016/ 2017 season), according to CONAB -Companhia Nacional de Abastecimento (2018). In the second cycle of cotton cultivation, we obtain better results with 2,224 kg ha −1 in T 3 . The yields obtained in the second cultivation cycle were similar to those obtained by Alikhasi, Kouchakzadeh, and Baniani (2012). They found that irrigation only with treated domestic sewage effluent promoted a cotton yield of 2,200 kg ha −1 .
Our second-cycle cotton yield results indicate a residual phosphorus effect of sewage effluent and organic matter degradation on soil from the first to the second cropping cycle. When using T 2 (50% well water and 50% sewage effluent), the cotton yield increased by about 420 kg ha −1 in the second cycle compared to the first cropping cycle, and when using T 3 , the cotton yield increased by 482 kg ha −1 . However, T 1 (well water) obtained a similar cotton yield in both cropping cycles. The positive effect of using sewage effluent on cotton occurs due to the supply of nutrients from the effluent, mainly N, P, and K. According to Carvalho, Ferreira, and Staut (2011), ideal amounts of P and K favor photosynthesis and the accumulation and translocation of carbohydrates to fruits in cotton, resulting in a good lint yield. In addition, P favors the maturation of bolls, accelerating their opening, which contributes to a good quality of lint.
Irrigation with sewage effluent promoted a decrease in cotton fiber percentage, but the increases of 235 to 300% in cotton yield outweighed the 8% decrease in cotton fiber percentage. Silva et al. (2013) also found that cotton fiber productivity decreases when irrigated with domestic sewage effluents. These authors indicate that the excess nitrogen present in treated domestic effluents can cause a drop in cotton fiber percentage because it extends the vegetative period, resulting in excessive plant growth.
Thus, the photoassimilates are directed to the vegetative structures to the detriment of reproductive structures.
Regarding cotton fiber quality, the phosphate fertilization at pre-planting in the first cycle and irrigation with sewage effluent in the second cycle reduced cotton SFI. However, all values of SFI observed are between 6 and 9%, within the low range (Fonseca and Santana 2002), which is typical for the cv. BRS 335 (Morello et al. 2012).
The good acceptance of cotton lint in the market occurs when the SFI is rated as low or very low. Short fibers reduce the efficiency of textile processing and the quality of the yarns, influencing their uniformity, imperfections, hairiness, strength, and appearance, representing 10 to 20% of the waste in the spinning units. Thus, the lower the values of SFI in cotton fiber, the better this fiber is for the textile industry (Fonseca and Santana 2002;Morello et al. 2012).
We obtain fibers more resistant to rupture in treatments T 2 and T 3 in the two cultivation cycles. However, regardless of the water source used in irrigation or phosphate fertilization, all STR values obtained in the present study were higher than 30 g tex −1 , so the fibers were classified as very resistant to rupture (Fonseca and Santana 2002).
ELG indicated very fragile fibers (ELG <5%) for treatments T 1 , T 3 , and with the presence of phosphorus at pre-planting in the first cycle of cultivation. In the second cycle, the values of ELG in all treatments were higher than 5%, classified as fragile (Fonseca and Santana 2002), indicating that the residual effect of nutrient supply from the first to the second cultivation cycle was positive for this variable.
We obtain the best MIC at T 1 without phosphate fertilization in the first cycle at pre-planting. However, all treatments achieved similar MIC in the second cycle (Tables 7 and 8). The MIC values ranged from 4.2 to 4.6 μg in −1 , classified as of regular quality (Fonseca and Santana 2002). This index (MIC) represents a criterion in the choice of lint for yarn production, considered in establishing the value of cotton lint in the market.
The indices RD and +B also represent an essential factor in the definition of the price of cotton lint, as they reflect its whiteness. The higher these indices, the greater the acceptance of the product by the industry (Fonseca and Santana 2002). For the cv. BRS 335, the treatments affected these variables only in the second cultivation cycle. However, regardless of the treatments, the mean values of RD and +B observed in the two cultivation cycles were higher than 83% and 9.5, respectively, indicating that the lint produced is of high quality.
We obtain the highest CSP (3255) values in T3 in the first cultivation cycle. However, in all treatments, the observed values are higher than 2500, classified as very high CSP (Sestren and Lima 2007) The fiber produced was of excellent quality, within the standards currently required by the national and international textile industry. Irrigation with sewage effluent (T 3 ) increased cotton yield by up to 300%, besides improving the fiber quality of the cv. BRS 335, especially the STR, ELG, CSP, and RD values, reduces the SFI. Phosphate fertilization failed in the growth, productivity, and cotton fiber quality, mainly in the second cropping cycle. These results indicate that sewage effluent supplied the phosphorus demand of cotton.

Conclusions
Sewage effluent can partially or fully supply the water and nutritional demands of the cotton cv. BRS 335 significantly increases plant growth, yield, and fiber quality. There is a residual effect of domestic sewage effluent from the first to the second cultivation cycle, increasing upland cotton production. We obtain the best cotton growth, yield, and fiber quality results using 100% sewage effluent to meet cotton water demand. Pre-planting phosphate fertilization failed in upland cotton cv. BRS 335.