Assessment of seasonal and spatial variation of surface water quality, identification of factors associated with water quality variability, and the modeling of critical nonpoint source pollution areas in an agricultural watershed

Surface water quality impairment in agricultural watersheds is a major environmental concern in the United States. To assess seasonal and spatial variability of surface water quality and identify factors associated with surface water quality variability, we monitored surface water quality at seven locations in Bayou Plaquemine Brule Watershed in Louisiana twice monthly from March of 2002 to February of 2008 and performed multivariate analyses of the dataset. Using the Soil and Water Assessment Tool (SWAT) model, we identified critical areas of nonpoint source pollution in the watershed. While temperature, turbidity, dissolved oxygen (DO), conductivity and pH were determined in the field using YSI Sonde (YSI Incorporated, Yellow Springs, Ohio), surface water samples were analyzed for total nitrogen (TN), total phosphorus (TP), nitrate/nitrite-N (NO3/NO2-N), soluble reactive phosphate (SRP), total suspended solids (TSS), and five-day biological oxygen demand (BOD5) in laboratory. The monthly water quality sampling included a regular sampling and an after-rain-event sampling. Average DO for the summer months, March through November, was 4.91 ± 0.08 mg L–1 (4.91 ± 0.08 ppm), while average DO for the winter months, December through February, was 8.32 ± 0.12 mg L–1 (8.32 ± 0.12 ppm). Dissolved oxygen was negatively correlated with TN (r = –0.22, p ≤ 0.001), SRP (r = –0.17, p ≤ 0.001), TP (r = –0.17, p ≤ 0.001), BOD5 (r = –0.25, p ≤ 0.001), and surface water temperature (r = –0.70, p ≤ 0.001). Turbidity was strongly correlated with TSS (r = 0.59, p ≤ 0.001), suggesting that most turbidity in the water body comes from suspended solids. Similarly BOD5 was significantly positively correlated with TN (r = 0.43, p ≤ 0.001), NO3/NO2-N (r = 0.26, p ≤ 0.001), TP (r = 0.25, p ≤ 0.001), and SRP (r = 0.18, p ≤ 0.001). Results of factor analyses showed sediment, phosphorus (P), nitrogen (N), surface water temperature, dissolved solids, and acidity/alkalinity as the most important factors associated with surface water quality variability in this watershed. Although relatively higher concentrations of sediments, TP, and TN were observed in the upper reaches of the watershed based on water quality monitoring, the SWAT simulation results showed the critical nonpoint source pollution areas of sediment, P, and N in the lower reaches of the watershed. Lower reaches of the watershed have mainly rice and crawfish production, while the upper reaches include primarily sugarcane, pasturelands, and soybean production. Information on seasonal variability of surface water quality, factors associated with surface water quality variability, and the critical areas for nonpoint source pollution will be valuable inputs for developing a watershed management plan for effective nonpoint source pollution control in an agricultural watershed. Kew words: agricultural watershed—factor analyses—nonpoint source pollution—SWAT model— total suspended solids—water quality monitoring Agriculture is the leading cause of nonpoint source pollution and water quality impairment in United States (USGAO 1990; Hamilton and Miller 2002; James et al. 2007). Erosion from agricultural lands, excess amount of fertilizer and manure applications, higher animal stocking densities, channel erosion, and constructional activities are responsible for nonpoint source pollution in the waterways (Carpenter 1998; Poudel et al. 2010). In addition, residential areas are also an important source of nonpoint source pollution (Poudel and Jeong 2009). Decomposition of organic materials in the waterways also results in the addition of nitrogen (N), phosphorus (P), and other minerals to water bodies. Low dissolved oxygen (DO) in a water body is often the result of elevated oxygen-demanding substances, elevated total suspended solids (TSS), and algal bloom. Nonpoint source pollution from dissolved substances such as chloride (Cl), sulfate (SO4), and carbonate (CO3) also contaminate water bodies due to application of fertilizers, limestone, rock weathering, pesticides, or from urban runoff. Seasonal and spatial variability in surface water quality has been reported by several researchers (Clarke et al. 2002; FranceLanord et al. 2003; Mankin et al. 2003; Nagano et al. 2003; El-Kaddah and Carey 2004). The seasonal pattern of surface water quality characteristics, especially nutrients and suspended solids, corresponds to the seasonality of agricultural production. Jaynes et al. (2004) reported a higher level of nitrate (NO3) concentration from late fall to midsummer and lower level of NO3 concentration in late summer in Walnut Creek Watershed, Iowa, United States. Randall et al. (2003) found that 62% of the annual drainage and 69% of the annual NO3 loss doi:10.2489/jswc.68.3.155 C opright © 2013 Soil nd W ater C onsvation Soiety. A ll rhts rerved. w w w sw c.org 68(3155-171 Jornal of Soil nd W ater C onsvation