Multivariate techniques, discriminant analysis, and WQI were applied to analyze a water quality data set including 27 parameters at 5 sites of the Lake Wular in Kashmir Himalaya from 2011 to 2013 to investigate spatiotemporal variations and identify potential pollution sources. Spatial and temporal variations in water quality parameters were evaluated through stepwise discriminant analysis (DA). The first spatial discriminant function (DF) accounted for 76.5% of the total spatial variance, and the second DF accounted for 19.1%. The mean values of water temperature, EC, total-N, K, and silicate showed a strong contribution to discriminate the five sampling sites. The mean concentration of NO2-N, total-N, and sulphate showed a strong contribution to discriminate the four sampling seasons and accounted for most of the expected seasonal variations. The order of major cations and anions was
Today surface water is most vulnerable to pollution due to its easy accessibility for disposal of pollutants and wastewaters. Worldwide surface water quality is governed by complex anthropogenic activities and natural processes [
In view of the spatial and temporal variations in the hydrochemistry of surface waters, regular monitoring programs are required for reliable estimates of the water quality [
In this study, physicochemical parameters of surface water quality directly affected by different pollution sources were monitored over two-year period. The data sets obtained were subjected to multivariate statistical technique, namely, discriminant analysis (DA), to obtain information about the similarities or dissimilarities among the monitoring periods and sites and to identify water quality variables responsible for spatial and temporal water quality variations in surface water. Besides multivariate statistical analysis, water quality index (a multifactor mathematical tool) was used to interpret water quality of studied lake numerically. It is regarded as one of the most effective ways to communicate water quality [
During the last decades, widespread deterioration in water quality of Wular lake has been reported due to anthropogenic influences (agricultural practices, increased exploitation of water resource, sewage runoff, agriculture, and urban sprawl) and natural processes (changes in precipitation, erosion, and weathering of crustal materials) [
The valley of Kashmir lies on the northern fringe of the Indian subcontinent and is lacustrine basin of the intermontane depression formed between the lesser and the greater Himalaya. It abounds a vast array of freshwater bodies, streams, lakes, ponds, and rivers famous for its beauty and natural scenery throughout the world. These numerous but varied freshwater ecosystems are of great aesthetic, cultural, socioeconomic, and geological value besides playing an important role in the conservation of genetic resources of both plants and animals. However, anthropogenic activities have resulted in heavy inflow of nutrients into these lakes from the catchment areas [
Geographically the Wular Lake, one of the largest wetlands of Asia, is situated at an altitude of 1,580 m (a.m.s.l), between 34°16′–34°20′N latitudes and 74°33′-74°44′E longitudes (Figure
Sampling station locations and their coordinates.
Study sites | Latitude | Longitude | Elevation | Location | Water depth (m) | |
---|---|---|---|---|---|---|
Makhdomyari | Site I | 34°-17′-44.2′′ | 74°-37′-24.2′′ | 1597 | Southeastern | 1–3 |
Vintage | Site II | 34°-24′-08.1′′ | 74°-32′-39.1′′ | 1583 | Eastern side | 1–4 |
Ashtang | Site III | 34°-24′-3.8′′ | 74°-32′-41.7′′ | 1583 | Northwestern side | 0.5–4.5 |
Watlab | Site IV | 34°-21′-29.4′′ | 74°-01′-59.2′′ | 1577 | Western side | 1–5.5 |
Ningle | Site V | 34°-17′-16.6′′ | 74°-30′-26.6′′ | 1574 | Northern | 0.5–4.4 |
Morphometric features of Lake Wular.
Max area | 61.6 Km2 |
Min area | 12.24 Km2 |
Average area | 31.415 Km2 |
Max volume | 371.825 × 106 m3 |
Min volume | 187.735 × 106 m3 |
Average volume | 267.675 × 106 m3 |
Elevation | 1,580 m (amsl) |
Maximum length | 16 km |
Minimum breadth | 7.6 km |
Shape | Elliptical |
Max depth | 5.8 m |
Minimum depth | 0.9 m |
Showing layout of study area and surface water quality monitoring stations in Lake Wular.
Surface water samples (0.5–1.0 m) were collected from five sites on monthly basis from February 2011 to January 2013. On each sampling date, three replicates were collected at each sampling site. The water samples were preserved in prerinsed 1-L polypropylene, acid-washed sampling bottles at 4°C in darkness and analyzed within 24 h. A saturated mercuric chloride solution was used at a final concentration of 0.2 mlL−1 to stop all microbiological activities in the water samples. The parameters including depth, transparency, temperature, pH, and conductivity were determined on spot while the rest of the parameters were determined in the laboratory. The parameters including orthophosphorus, total phosphorus, ammoniacal nitrogen, nitrite nitrogen, nitrate nitrogen, organic nitrogen (Kjeldahl nitrogen minus ammoniacal nitrogen), alkalinity, free CO2, conductivity, chloride, total hardness, calcium hardness, magnesium hardness, Na, K, silicate, sulphate, iron, and TDS were determined in the laboratory within 24 hours of sampling. The analysis was done by adopting standard methods of Mackereth, Golterman and Clymo, and APHA [
Data for physicochemical parameters of water samples were presented as mean values and analyzed using descriptive analysis. We used standard deviation for describing the spatiotemporal degree of variations of the observed water quality parameters in Lake Wular, in different months and seasons. 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).
Stepwise discriminant analysis (DA) which is also a multivariate statistical technique was used for spatiotemporal analysis of water quality data. Discriminant analysis (DA) is used to classify cases into categorical-dependent values, usually a dichotomy. If discriminant analysis is effective for a set of data, the classification table of correct and incorrect estimates will yield a high correct percentage. In DA, multiple quantitative attributes are used to discriminate between two or more naturally occurring groups. In contrast to CA, DA provides statistical classification of samples and is performed with prior knowledge of membership of objects to a particular group or cluster. Furthermore, DA helps in grouping samples sharing common properties. The DA technique builds up a discriminant function for each group, which operates on raw data [
The weight coefficient maximizes the distance between the means of the criterion (dependent) variable. DA was performed, on each raw data matrix using stepwise modes in constructing discriminant functions, to evaluate both the spatial and temporal variations in water quality of the lake. The sites (spatial) and the seasons (temporal) were the grouping (dependent) variables, whereas all the measured parameters constituted the independent variables. Linear discriminant functions were used to describe or elucidate the differences between the sampling sites and the influence of season on water quality of each sampling site. The relative contribution of all variables to the separation of groups was highlighted [
Accurate and timely information on the quality of water is necessary to shape a sound public policy and to implement the water quality improvement programmes efficiently. One of the most effective ways to communicate information on water quality trends is with indices. The WQI is a mathematical instrument used to transform large quantities of water quality data into a single number which summarize different quality parameters. The WQI is an index of water quality for a particular use. Mathematically, the index is an arithmetic weighting of normalized water quality measurements. The weightings are different for different water usages [
Water quality index is
When
Pollutants are: (i) completely absent when
The more harmful a given pollutant is, the smaller is its permissible value for drinking water. So the “weights” for various water quality parameters are assumed to be inversely proportional to the recommended standards for the corresponding parameters; that is,
Based on the range of WQI values, water is grouped into the following categories [ WQI less than 25: water is not polluted and fit for human consumption (excellent), WQI between 26 and 50: slightly polluted (good), WQI between 51 and 75: moderately polluted (poor), WQI between 76 and 100: polluted (very poor), WQI above 100: excessively polluted and unfit for human use (unsuitable).
In this study, multivariate statistical and mathematical analysis methods (DA and WQI approaches) were applied to evaluate the impact of anthropogenic activities and spatiotemporal variations in physicochemical characteristics on water quality of Wular Lake. Statistical conclusions and tests were made on the basis of a multiparametric model, specifying how water quality parameters are changed with different seasons and nature of polluting source in the studied aquatic system.
During the two years of this study, the lake behavior was explored by measuring 27 parameters to assess the quality of this aquatic system. All these parameters were measured from samples collected from study stations in the Wular Lake, as indicated in Figure
(a) Box and whisker plots showing spatiotemporal dynamics of physical parameters. (b) Box and whisker plots showing spatiotemporal dynamics of chemical parameters. (c) Box and whisker plots showing spatiotemporal dynamics of nutrient (N and P) parameters. (d) Box and whisker plots showing spatiotemporal dynamics of ionic parameters.
Discriminant analysis is one of the more advanced multivariate classification techniques used to define the variables, discriminating between the identified clusters, by specifying the weight (i.e., discriminating power) to these variables [
Spatial variations in water parameters were evaluated through stepwise discriminant analysis (DA) method. Four discriminant functions (DFs) were found to discriminate the quality of the five sampling sites used in this study (Table
Discriminant function coefficients and Wilks Lambda for spatial variations in water parameters of Lake Wular.
Standardized canonical discriminant function coefficients | Eigenvalues | Wilks Lambda | ||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Function | WT | Depth | Trans | EC | T_Alkalinity | Ca_HArd | NO2-N | Org_N | Total-N | Na+ | K+ | Silicate | Sulphate | TDS | Eigenvalue | % Variance | Cumulative % | Canonical Correlation | Wilks’ Lambda | Chi-square | df | Sig. |
1 | 1.06 | −2.01 | 0.31 | 0.94 | 0.20 | 0.12 | 0.78 | 0.29 | −1.22 | 0.87 | 1.07 | 1.22 | 0.89 | 0.64 | 20.82 | 76.50 | 76.5 | 0.98 | 0.00 | 635.6 | 56.0 | 0 |
2 | −0.87 | 1.86 | 0.10 | 0.50 | 0.63 | 0.53 | 0.52 | 0.98 | −0.27 | 0.84 | 1.00 | 0.51 | 0.18 | −0.38 | 5.19 | 19.10 | 95.6 | 0.92 | 0.07 | 298.0 | 39.0 | 0 |
3 | −0.74 | 0.59 | 0.47 | 0.45 | −0.03 | 0.25 | 0.40 | 0.48 | 0.29 | −0.57 | 0.28 | −2.22 | 2.04 | 0.67 | .950 | 3.50 | 99.1 | 0.70 | 0.41 | 98.2 | 24.0 | 0 |
4 | −0.86 | −0.46 | −0.68 | 0.31 | 0.45 | 0.65 | 0.84 | 0.15 | −0.17 | −0.25 | −0.11 | 0.28 | 0.89 | −0.49 | .259 | 0.90 | 100.0 | 0.45 | 0.80 | 25.1 | 11.0 | 0.009 |
From the canonical discriminant plot (Figure
Discriminant plot showing spatial variation of water parameters in Lake Wular (2012–2003).
Temporal DA was performed on synthesized data after dividing the whole data set into four seasonal groups (spring, summer, autumn, and winter). Three discriminant functions (DFs) were found to discriminate the quality of the three sampling seasons used in this study (Table
Discriminant function coefficients and Wilks Lambda for temporal variations in water parameters of Lake Wular.
Standardized canonical discriminant function coefficients | Eigenvalues | Wilks Lambda | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Function | NO2-N | NH4-N | Total-N | Total-P | TDS | Sulphate | Eigenvalue | % of Variance | Cumulative % | Canonical Correlation | Wilks’ Lambda | Chi-square | df | Sig. |
1 | 2.17 | −1.63 | 2.51 | −1.88 | 0.09 | 2.28 | 94.22 | 85.90 | 85.90 | 1.00 | 0 | 159.038 | 18 | 0 |
2 | 1.40 | 1.54 | −0.90 | −0.76 | −1.16 | 0.75 | 9.69 | 8.80 | 94.80 | 0.95 | 0.014 | 77.028 | 10 | 0 |
3 | 0.64 | 1.88 | −1.16 | −0.15 | −0.09 | 1.02 | 5.76 | 5.20 | 100.00 | 0.92 | 0.148 | 34.384 | 4 | 0 |
Drinking water standards, unit weights, average water quality, and water quality index of five study stations in Lake Wular.
Chemical parameters | Standards | Recommending agency | Unit weight (wi) | Makhdoomyari | Vintage | Ashtang | Watlab | Ningle | |||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Av Wq |
|
Av Wq |
|
Av Wq |
|
Av Wq |
|
Av Wq |
|
||||
pH | 7.0–8.5 | ICMR | 0.037 | 7.84 | 4.19 | 7.76 | 3.80 | 7.80 | 3.97 | 7.75 | 3.73 | 7.77 | 3.87 |
Electrical conductivity | 750 |
WHO | 0.000 | 252.9 | 0.01 | 254.2 | 0.01 | 255.4 | 0.01 | 282.3 | 0.01 | 248.5 | 0.01 |
Chloride | 250 mg/L | ISI | 0.001 | 12.35 | 0.01 | 13.34 | 0.01 | 14.33 | 0.01 | 14.11 | 0.01 | 12.85 | 0.01 |
Dissolved oxygen | 5.0 mg/L | WHO | 0.052 | 9.03 | 3.04 | 8.93 | 3.10 | 8.93 | 3.10 | 8.94 | 3.09 | 9.12 | 2.99 |
Free CO2 | 22 mg/L | WHO | 0.012 | 9.89 | 0.54 | 10.02 | 0.54 | 10.00 | 0.54 | 10.04 | 0.54 | 10.43 | 0.57 |
Total alkalinity | 120 mg/L | USPHS | 0.002 | 112.7 | 0.21 | 107.0 | 0.19 | 99.54 | 0.18 | 114.5 | 0.21 | 102.0 | 0.19 |
Total hardness | 500 mg/L | WHO | 0.001 | 145.5 | 0.02 | 143.9 | 0.02 | 142.2 | 0.01 | 161.7 | 0.02 | 128.2 | 0.01 |
Nitrate-N | 45 mg/L | WHO | 0.006 | 0.47 | 0.01 | 0.48 | 0.01 | 0.48 | 0.01 | 0.46 | 0.01 | 0.42 | 0.01 |
Calcium | 75 mg/L | ICMR | 0.003 | 21.16 | 0.10 | 20.88 | 0.10 | 20.61 | 0.10 | 26.05 | 0.12 | 19.85 | 0.09 |
Magnesium | 50 mg/L | ICMR | 0.009 | 13.50 | 0.39 | 13.66 | 0.24 | 13.23 | 0.23 | 14.09 | 0.25 | 11.47 | 0.20 |
Sulfate | 200 mg/L | ICMR | 0.001 | 6.10 | 0.00 | 5.50 | 0.00 | 5.03 | 0.00 | 4.77 | 0.00 | 3.90 | 0.00 |
Iron | 0.3 mg/L | ISI | 0.874 | 0.14 | 40.7 | 0.13 | 38.5 | 0.13 | 39.1 | 0.11 | 32.6 | 0.10 | 29.1 |
Total dissolved solids | 500 mg/L | WHO | 0.001 | 154.1 | 0.02 | 146.9 | 0.02 | 133.3 | 0.01 | 106.8 | 0.01 | 77.08 | 0.01 |
|
|||||||||||||
|
1.00 |
|
49.2 |
|
46.5 |
|
47.3 |
|
40.6 |
|
37.1 |
Discriminant plot for temporal variations (Figure
Discriminant plot showing temporal variation of water parameters in Lake Wular (2012-2013).
Evaluation of overall water quality is not an easy task, particularly when different criteria for different uses are applied. Moreover, the classification of water quality follows various definitions with respect to the contents of different water quality parameters. Dozens of variables have been developed and are available to be used in management governmental or environmental programs, but the high price because of water analysis to attend these programs generally makes it difficult to use them. In this study, the application of the water quality index approach to the Lake Wular has the objective of providing a simple and valid method for expressing the results of several parameters in order to more rapidly and conveniently assess the water quality. Combining different parameters into one single number leads to an easy interpretation of the index, thus providing an important tool for management purposes. As described, WQI employing thirteen parameters can give an indication of the health of the water body at various points and can be used to keep track of and analyze changes over time, but other options can be used in an economic way.
The values of water quality indices are taken as the standards for drinking water according to Table
In this study, statistical and mathematical exploratory techniques were utilized to evaluate variations in surface water quality of Lake Wular. This study has shown that the highest sources of variation in water quality are both seasonal factors as well as anthropogenic factors. The results exhibit that the DA technique is useful in present accredited classification of surface waters in the whole lake basin; hence, the number of sampling sites and respective cost in the future monitoring plans can be lessen. The water quality index provided a numeric expression, used to transform large number of variables data into a single number, which represented the water quality level of whole Wular Lake basin. Thus, the study illustrates the useful application of chemometric techniques for the analysis and interpretation of lake water quality data and identification based on pollution status and identification of pollution sources as part of the efforts towards management of sustainability of this lake. The main sources of pollution came from domestic wastewater and agricultural activities and runoff; however, they contributed differently to each station in regard to pollution levels. These results provide fundamental information for developing better water pollution control strategies for the Wular Lake.
The authors declare that there is no conflict of interests regarding the publication of this paper.
The authors are indebted to Director, Centre of Research for Development (CORD), and Head Department of Environmental Science, University of Kashmir, for providing full support and necessary laboratory facilities for carrying out the chemical analysis. Also the authors gratefully acknowledge the help from the State Irrigation and Flood Control and Indian Meteorological Department for providing necessary data.