Journal of Ecosystems

Journal of Ecosystems / 2014 / Article

Research Article | Open Access

Volume 2014 |Article ID 898054 | https://doi.org/10.1155/2014/898054

Salim Aijaz Bhat, Gowhar Meraj, Sayar Yaseen, Ashok K. Pandit, "Statistical Assessment of Water Quality Parameters for Pollution Source Identification in Sukhnag Stream: An Inflow Stream of Lake Wular (Ramsar Site), Kashmir Himalaya", Journal of Ecosystems, vol. 2014, Article ID 898054, 18 pages, 2014. https://doi.org/10.1155/2014/898054

Statistical Assessment of Water Quality Parameters for Pollution Source Identification in Sukhnag Stream: An Inflow Stream of Lake Wular (Ramsar Site), Kashmir Himalaya

Academic Editor: Guangliang Liu
Received01 Jun 2013
Accepted04 Dec 2013
Published20 Jan 2014

Abstract

The precursors of deterioration of immaculate Kashmir Himalaya water bodies are apparent. This study statistically analyzes the deteriorating water quality of the Sukhnag stream, one of the major inflow stream of Lake Wular. Statistical techniques, such as principal component analysis (PCA), regression analysis, and cluster analysis, were applied to 26 water quality parameters. PCA identified a reduced number of mean 2 varifactors, indicating that 96% of temporal and spatial changes affect the water quality in this stream. First factor from factor analysis explained 66% of the total variance between velocity, total-P, NO3–N, Ca2+, Na+, TS, TSS, and TDS. Bray-Curtis cluster analysis showed a similarity of 96% between sites IV and V and 94% between sites II and III. The dendrogram of seasonal similarity showed a maximum similarity of 97% between spring and autumn and 82% between winter and summer clusters. For nitrate, nitrite, and chloride, the trend in accumulation factor (AF) showed that the downstream concentrations were about 2.0, 2.0, and 2.9, times respectively, greater than upstream concentrations.

1. Introduction

River water quality is of great environmental concern since it is one of the major available fresh water resources for human consumption [1, 2]. Throughout the history of human civilization, rivers have always been heavily exposed to pollution, due to their easy accessibility to disposal of wastes. However, after the industrial revolution the carrying capacity of the rivers to process wastes reduced tremendously [3, 4]. Anthropogenic activities such as urban, industrial, and agricultural as well as natural processes, such as precipitation inputs, erosion, and weathering of crustal materials affect river water quality and determine its use for various purposes [15]. The usage also depends upon the linkages (channels) in the river system, as inland waterways play a major role in the assimilation and transportation of contaminants from a number of sources [68]. Besides linkages, the seasonal variation in precipitation, surface runoff, interflow, groundwater flow, and pumped in and out flows also have a strong effect on the concentration of pollutants in rivers [912]. In view of the limited stock of freshwater worldwide and the role that anthropogenic activities play in the deterioration of water quality, the protection of these water resources has been given topmost priority in the 21st century [1315]. Research-wise, one of the important stages in the protection and conservation of these resources is the spatiotemporal analysis of water and sediment quality of the aquatic systems [16]. The nonlinear nature of environmental data makes spatio-temporal variations of water quality often difficult to interpret and for this reason statistical approaches are used for providing representative and reliable analysis of the water quality [17]. Multivariate statistical techniques such as cluster analysis (CA) and factor analysis (FA) have been widely used as unbiased methods in analysis of water quality data for drawing out meaningful conclusions [18, 19]. Also it has been widely used to characterize and evaluate water quality for analyzing spatio-temporal variations caused by natural and anthropogenic processes [2022]. In this paper we present a methodology for examining the impact of all the sources of pollution in Sukhnag stream (Kashmir Himalayas) and to identify the parameters responsible for spatiotemporal variability in water quality using CA and FA.

2. Materials and Methods

2.1. Materials
2.1.1. Study Area

The present study was carried out on Sukhnag stream in Kashmir Himalaya. It is among the five major inflows of the Lake Wular. This lake is the largest fresh water lake of Indian subcontinent and has been designated as a Ramsar site in 1990 under the Ramsar convention of 1975. The Sukhnag, a torrential stream, flows through Budgam district, in the state of Jammu and Kashmir (Figure 1). It flows from the mountain reaches of the Pir Panjal mountain range located in the southwest of Beerwah town. The Sukhnag stream drains the famous Toshmaidan region in the higher locales of Pir Panjal range. It has a glacial origin and covers a distance of about 51 kms from head to mouth. Descending from the mountains, the Sukhnag passes through a sand choked bed across the Karewas, finally merging with the outlet of Hokersar wetland (Ramsar site). The Sukhnag drainage system spreads over an area of about 395.03 km2 and about 1551 streams cascade the waters for the whole watershed into this stream. During flash rains the water in this stream flows with the tremendous velocity in the upper reaches causing soil wastage of the left and right embankments of the stream and greatly damaging standing crops, plantation, houses, and road communication. The stream passes through a large area of high socioeconomic importance to North Kashmir. These areas include Rangzabal, Zagu, Bras, Arizal, Chill, Zanigam, Sail, Kangund, Goaripora, Siedpora, Beerwah, Aarwah, Aripanthan, Rathson, Makhama, Nawpora, Check-kawosa, Botacheck, and Narbal. The stream serves as a life line of this vast area as it serves as a source of water for both domestic and agricultural purposes. The current study is therefore a step forward in addressing the deteriorating conditions of the stream so as to recommend concrete measures for its sustainable management.

2.2. Methods
2.2.1. Sampling and Analysis

Samples were taken flow proportionately (i.e., more frequently during peak flow periods than during low flow periods) to capture nutrient pulses during runoff events from February 2011 to January 2012. The surface water samples were collected in midchannel points between 10.00 and 12.00 hours from each of the sampling sites and placed in prerinsed polyethylene and acid-washed bottles for the laboratory investigations. The parameters such as depth, transparency, temperature, pH, and conductivity were determined on the spot while the rest of the parameters were determined in the laboratory. These include orthophosphorus, total phosphorus, ammoniacal nitrogen, nitrite-nitrogen, nitrate-nitrogen, organic nitrogen (Kjeldahl nitrogen minus ammonical nitrogen), alkalinity, free CO2, conductivity, chloride, total hardness, calcium hardness, magnesium hardness, sodium, and potassium. They were determined in the laboratory within 24 hours of sampling by adopting standard methods of Golterman and Clymo (1969) and APHA (1998) [2325].

2.2.2. Statistical Analysis

Data for physicochemical parameters of water samples were presented as mean values and analyzed using descriptive analysis. We used coefficient of correlation (CV) and -test, for describing the temporal variations of the observed water quality parameters. Prior to investigating the seasonal effect on water quality parameters, we divided the whole observation period into four fixed seasons: spring (March, April, and May), summer (June, July, and August), autumn (September, October, and November), and winter (December, January, and February). Regression analysis (RA) was carried out in order to know the nature and magnitude of the relationship among various physicochemical parameters. First, we determined the best-fit model (the largest ) for exploring whether there was any significant relationship among water quality parameters or not.

Accumulation factor (AF), the ratio of the average level of a given parameter downstream (following source discharge) to the corresponding average level upstream (prior to the source discharge) [26], was used to estimate the degree of contamination due to anthropogenic inputs.

The degree of river recovery capacity (RRC) for this stream was calculated using the mathematical equation by Ernestova and Semenova [27] and modified by Fakayode [26]; that is, where is the level of a parameter downstream (i.e., immediately after the discharge point) and is the corresponding average level upstream where the water is relatively unpolluted.

2.2.3. Multivariate Statistical Methods

With the objective of evaluating significant differences among the sites for all water quality variables, data was analyzed using one-way analysis of variance (ANOVA) at 0.05% level of significance [28]. Stream water quality was subjected to two multivariate techniques: cluster analysis (CA) and principal component analysis (PCA) [29]. CA and PCA explore groups and sets of variables with similar properties, thus potentially allowing us to simplify our description of observations by allowing us to find the structure or patterns in the presence of chaotic or confusing data [30]. All statistical analyses were performed using the SPSS (v. 16) and PAST (v. 1.93) software applications.

Cluster Analysis (CA). Cluster analysis is a multivariate statistical technique, which allows the assembling of objects based on their similarity. CA classifies objects, so that each object is similar to the others in the cluster with respect to a predetermined selection criterion. Bray-Curtis cluster analysis is the most common approach of CA, which provides intuitive similarity relationships between any one sample and the entire dataset and is typically illustrated by a dendrogram (tree diagram). The dendrogram provides a visual summary of the clustering processes, presenting a picture of the groups and their proximity with a dramatic reduction in dimensionality of the original data [31].

Factor Analysis/Principal Component Analysis (PCA). Factor analysis is applied to reduce the dimensionality of a data set consisting of large number of interrelated variables, and this reduction is achieved by transforming the data set into a new set of variables—the principal components (PCs), which are orthogonal (noncorrelated) and are arranged in decreasing order of importance. In this study we used principal component analysis (PCA) of factor analysis. The PCA is a data reduction technique and suggests how many varieties are important to explain the observed variance in the data. Mathematically, PCs are computed from covariance or other cross-product matrixes, which describe the dispersion of the multiple measured parameters to obtain eigenvalues and eigenvectors. Moreover, these are the linear combinations of the original variables and the eigenvectors [32]. PCA can be used to reduce the number of variables and explain the same amount of variance with fewer variables (principal components) [33]. Also PCA attempts to explain the correlation between the observations in terms of the underlying factors, which are not directly observable [34].

Prior to modeling, all the nutrient concentrations were log-transformed to make the distribution closer to the normal. Statistical conclusions and tests were made on the basis of a multiparametric model. We have used CV, -test, ANOVA, RA, CA, and PCA to evaluate the impact of anthropogenic activities and spatio-temporal variations on physicochemical characteristics of Sukhnag stream.

3. Results and Discussion

The mean values of physicochemical parameters at different sampling sites in Sukhnag stream during the period of 12 months (February 2011–January 2012) are presented in Table 1. Water temperature, pH, and DO demonstrated a seasonal cycle during the period of study. High temperature values were recorded (24.33 ± 2.52°C) in summer season at site V and low values (2.33 ± 1.23°C) were recorded in winter at site I. pH ranged from 7.26 (±0.07) to 8.07 (±0.21) with the highest values in winter and the lowest in summer at most of the study sites. The pH values at the tail site (site V) of the stream showed a decreased trend from wet to dry season while upstream values were higher during the dry season. DO values were generally higher at upstream sites and the lowest at the downstream sites. There was however a progressive increase in DO (6.46 ± 0.31 to 13.4 ± 0.45) at all the sampling sites during the transition to rainy, winter season. Variation in EC was significant (CV = 5.2–18.5%, ) among seasons and at all sampling sites (). Higher values of EC were recorded in spring (396.6 ± 20.8 μS/cm) at the tail site (site V) and lower in winter (208.0 ± 8.54 μS/cm) at the headstream site (site I). The higher EC is attributed to the high degree of anthropogenic activities such as waste disposal and agricultural runoff. The seasonal variations in depth during one year of study showed that it was highest in the spring season. By autumn the depth starts to decrease and is lowest in the winter. The water depth at all the sites varied both spatially () and temporally (CV = 4.6–60.3%, ). Maximum surface water velocities for the five sites were recorded in spring season (peak flow season) and minimum were recorded in winter season (snowy season). Surface water velocities showed greater variability in the winter season (CV = 10.5–36.1%) than in autumn season (CV = 4.2–16.2%). Ortho-phosphorus and total phosphorus concentrations were highest (0.13 ± 0.01 mg/L and 0.41 ± 0.02 mg/L, resp.) in spring at site V and lowest (0.03 ± 0.01 and 0.17 ± 0.01 mg/L, resp.) in autumn at site I. Concentrations of NO3–N and NO2–N were highest (0.78 ± 0.05 and 0.021 ± 0.0 mg/L, resp.) in summer at site V and lowest (0.26 ± 0.02 and 0.007 ± 0.00 mg/L, resp.) in winter at site I. Ammonical-N showed the highest values in winter at site V (0.16 ± 0.05) and lowest values in summer at site I (0.03 ± 0.00). The highest values of organic-N and total nitrogen (0.62 ± 0.06 and 1.34 ± 0.14 mg/L, resp.) were found in summer at site V and lowest (0.13 ± 0.01 and 0.49 ± 0.01 mg/L, resp.) in winter at site I. Total hardness, calcium hardness, and magnesium hardness were observed highest in spring and lowest in autumn and winter at all the sites. Lower levels of total hardness, calcium hardness, and magnesium hardness were observed at upstream sites compared to the downstream sites and among the seasons; the lowest values were recorded in the winter at all the sites. The highest values of Ca2+, Mg2+, Na+, K+, and TDS (46.18 ± 4.41, 9.82 ± 0.93, 15.39 ± 1.47, 3.56 ± 0.26, and 353.3 ± 50.5 mg/L, resp.) were recorded at the tail site (site V) in spring season and lowest (16.83 ± 5.28, 3.91 ± 1.22, 6.23 ± 1.95, 1.55 ± 0.48, and 116.6 ± 27.3 mg/L, resp.) at the headstream site (site I) in the winter season. TDS and TSS values were recorded highest (487.0 ± 93.3 and 145.66 ± 43.13 mg/L, resp.) in spring at site IV and lowest (139.3 ± 40.7 and 22.66 ± 13.42 mg/L, resp.) in winter at site I. The overall values of CV showed significant difference of concentrations from head to tail. On the basis of molar concentrations, among the cations, Ca2+ and Na+ were dominant and Mg+2 and K+ were found in minor concentrations. Chloride was the dominant anion observed. Overall we observed significant degree of spatial and temporal variations in the concentration of water quality parameters using ANOVA () and -test () analysis.


S. no.ParametersSeasonsSite ISite IISite IIISite IVSite V

1Water temperature (°C)Winter
Cvar.%52.9029.0647.1965.4774.60
Spring
Cvar.%32.9130.7028.4831.9936.33
Summer
Cvar.%13.668.245.876.1610.34
Autumn
Cvar.%48.3044.7040.2037.5039.05

2Depth (m)Winter
Cvar.%49.6438.6038.3460.3255.53
Spring
Cvar.%29.5815.9823.3914.1413.80
Summer
Cvar.%4.6510.8339.8237.473.03
Autumn
Cvar.%9.097.6914.1731.0833.25

3Velocity (m/s)Winter
Cvar.%16.9616.6220.8610.5336.14
Spring
Cvar.%13.4713.3412.8415.337.24
Summer
Cvar.%12.7316.5012.6411.2529.55
Autumn
Cvar.%7.424.283.0516.229.93

4DO (mg/L)Winter
Cvar.%3.351.861.020.570.59
Spring
Cvar.%5.897.496.266.466.61
Summer
Cvar.%3.295.353.534.324.72
Autumn
Cvar.%5.935.7511.4312.1811.36

5Free CO2 (mg/L)Winter
Cvar.%4.079.868.325.014.01
Spring
Cvar.%12.0010.665.689.9514.43
Summer
Cvar.%16.935.598.723.825.09
Autumn
Cvar.%5.573.486.5613.2110.26

6pHWinter
Cvar.%2.310.574.480.355.05
Spring
Cvar.%1.471.462.050.681.03
Summer
Cvar.%2.644.361.970.411.08
Autumn
Cvar.%0.640.130.231.081.04

7Alkalinity (mg/L)Winter
Cvar.%8.167.9111.486.7010.91
Spring
Cvar.%14.1211.9610.336.105.66
Summer
Cvar.%6.032.902.813.952.81
Autumn
Cvar.%12.406.578.7411.876.46

8Chloride (mg/L)Winter
Cvar.%22.926.7421.3914.959.07
Spring
Cvar.%10.6031.688.6312.7311.08
Summer
Cvar.%10.648.4014.0311.363.47
Autumn
Cvar.%22.8810.008.779.9412.30

9Conductivity (µS/cm)Winter
Cvar.%4.1118.5022.3013.359.96
Spring
Cvar.%7.288.3417.048.775.25
Summer
Cvar.%4.6312.945.515.694.70
Autumn
Cvar.%7.7513.4210.676.206.10

10Total hardness (mg/L)Winter
Cvar.%44.7448.9242.9841.7734.71
Spring
Cvar.%9.6717.5714.9812.2713.15
Summer
Cvar.%9.9715.2513.6815.5614.38
Autumn
Cvar.%3.757.614.958.986.58

11Calcium hardness (mg/L)Winter
Cvar.%31.4134.8732.3524.6223.97
Spring
Cvar.%6.3112.714.399.259.55
Summer
Cvar.%15.5316.9911.1813.8413.92
Autumn
Cvar.%4.2615.833.518.997.95

12Magnesium hardness (mg/L)Winter
Cvar.%62.2568.2656.3360.9445.77
Spring
Cvar.%28.6843.2733.7027.5517.61
Summer
Cvar.%15.1717.5123.6026.7516.39
Autumn
Cvar.%3.956.9810.3911.974.70

13Total solids (mg/L)Winter
Cvar.%29.2327.3726.5027.1826.90
Spring
Cvar.%22.4217.2518.3519.1718.53
Summer
Cvar.%17.5015.7411.2212.5812.62
Autumn
Cvar.%6.066.326.337.637.06

14TDS (mg/L)Winter
Cvar.%23.4021.4919.6021.9420.95
Spring
Cvar.%20.0313.2713.6514.7114.30
Summer
Cvar.%12.038.966.718.8010.63
Autumn
Cvar.%2.886.937.298.847.68

15TSS (mg/L)Winter
Cvar.%59.2462.7863.5355.4071.47
Spring