Assessment of the impact of pesticide usage in groundwater aquifer of an agriculturally dominated area in North West India

Groundwater plays an important role as a decentralized source of drinking water for millions of rural and urban families across the globe. In India, it accounts for nearly 80 percent of the rural domestic water needs and 50 percent of the urban water needs (Kumar & Shah 2004). Although groundwater is less susceptible to contamination compared to surface water resources, several anthropogenic activities such as the improper disposal of municipal and industrial effluents and agricultural leachate, are causing groundwater pollution. Among various sources of groundwater pollution, the extensive and improper utilization of pesticides has also become a source of groundwater pollution.

Development of agriculture is assisted by the usage of pesticides. Pesticide use has helped in preventing the losses caused by pest attack and has improved the net production potential of crops, but the surplus amount unutilized by the crops leaches down to ground water causing its pollution (Zhao & Pei 2012). Pesticide contamination in ground water is related to the persistence of pesticides in soil. The ability of a pesticide to be absorbed by the soil media decides whether it will leach down to ground water or not. The pesticides with poor adsorption or absorption on soil surface will have a higher tendency to leach down to ground water thus leading to its pollution. Groundwater once polluted will take several decades to get cleaned on its own and clean up processes are uneconomic, tedious and cumbersome.

Several studies have been reported showing groundwater pollution due to pesticides usage. Pesticides cause serious health hazards to living systems due to their rapid solubility in fat and thus accumulating in target organisms (Agrawal et al. 2010). The vulnerability of groundwater to pesticide pollution is generally governed by the soil properties such as its texture, total organic matter present, and pesticides usage and their degradation products (Kumar & Shah 2004). Aktar et al. (2009) have discussed in details the potential effects of pesticides usage viz. advantages and disadvantages, over the surface water and groundwater resources of the corresponding area.

Thakur et al. (2015) assessed the groundwater contamination through pesticide usage in the vegetable growing areas in Delhi. Eight groundwater samples from different farms were collected and analyzed for the pesticide residues (Organochlorine) in it. Authors detected the presence of pesticides in groundwater in the area. Kole & Bagchi (1995) conducted a survey by drawing drinking water samples from various hand pumps and wells around Bhopal and found that more than half of the samples were contaminated with Organo Chlorine pesticides above the EPA standards. Chaudhary et al. (2002) showed the elevated levels of pesticides in the groundwater aquifer in Howrah district rendering the water unfit for drinking purposes. Sankararamakrishnan et al. (2005) conducted the groundwater quality analysis of Kanpur, India and reported the presence of high concentration of organochlorine and organophosphorus pesticides. Kumari et al. (2008) collected groundwater samples from the tube wells in farm fields around Hissar and found the pesticide residues (organosulphates, only chloropyriphos) more than the prescribed drinking water limits rendering the groundwater in the area unfit for drinking purposes.

Mohapatra et al. (1995) analyzed the ground water samples in rural areas near Farrukhabad (U.P) and concluded that the possible sources of groundwater contamination are the groundwater recharge by contaminated Ganga River and downward movement of pesticide residues along with rain water. Tariq et al. (2004) analyzed the groundwater samples in the cotton growing districts of Bahwalanagar, Muzafargarh and Rajanpur and found the presence of pesticide residues in groundwater.

Li et al. (2013) reported the presence of organo-chlorine pesticides in the shallow groundwater, samples collected from Taibu basin of China. Goncalves et al. (2007) studied the impact of intensive horticultural practices on ground water contamination in Portugal and found the presence of these pesticides in the groundwater – Lindane, Pendimethaline, endosulfan sulphate and endosulfan. Hernández-Romero et al. (2004) evaluated the water quality of the Pozuelos–Murillo lagoon system in southern Mexico with particular emphasis on the detection of organochlorine and organophosphate pesticide residues in water and sediments. Residues of organochlorine compounds were detected in the study area. As per a study conducted by Belluck et al. (1991), Atrazine and its metabolites have become the main toxic contaminants in the ground water of USA/Canada.

The above mentioned studies are based on the experimental detection of the pesticide residues in groundwater by analyzing the groundwater samples in an area. There have been only a few studies based on prediction of the residues employing computational tools like the Pesticide Root Zone Model (PZRM) (Carsel et al. 1985; Trevisan et al. 1993; Cogger et al. 1998), Pesticide Emission Assessment at Regional and Local scales (PEARL) Model for assessing leaching of pesticide into groundwater or its retention in the soil (Leistra et al. 2001; Tiktak et al. 2002a, 2002b) and Pesticide Impact Rating Index (PIRI) (Kookana et al. 2005; Aravinna et al. 2017).

PRZM is a finite difference based model used in simulating the movement of a contaminant (pesticide in the present case) in an unsaturated soil system. The model utilizes the data related to the soil characteristics, climatological data and data related to pesticide application, its degradation rate etc. to predict the movement of the pesticide through soil and ultimately to the groundwater. PEARL is also a finite difference based model which describes the fate of a pesticide and relevant transformation products in the soil-plant system. The model utilizes data related to pesticide application, pesticide transport processes (convective, dispersive and diffusive), pesticide sorption (equilibrium and non-equilibrium) and transformation, uptake of pesticides by plant roots, lateral discharge of pesticide with drainage water, and volatilization of pesticide from soil and plant surfaces.

These tools give quantitative assessment of the pesticide impact on groundwater, which is based on the pesticide usage in the area. This assessment helps in formulating regulatory policies and practices related to pesticide usage having least detrimental impact on the environment and specifically groundwater, so that proper pesticide usage policies can be developed.

The present study deals with assessment of the impact of the pesticide usage in terms of its residues/load in groundwater in an area (a few villages) lying in Jalandhar district in the state of Punjab, India. At present, few studies have been carried, out in only limited regions in India, to assess the impact of pesticides and agrochemicals. Although the entire country seems to be affected by the over usage of pesticides, the problem appears to be more acute in the state of Punjab because the state is dominant in agricultural activities. Increased productivity due to modernization of agriculture has been driven by the excessive use of pesticides, which has consequently resulted in accumulation of pesticide residues in the groundwater of the region. Numerous studies on estimating the pesticide residues in groundwater have been performed experimentally, but little information exists on a probable relationship between the quantity of pesticide used and its impact on quality of groundwater. A need to carry out comprehensive monitoring across the country to regulate the usage of pesticides with suitable mathematical models, under corresponding Indian conditions, is required.

Accordingly, certain villages of Jalandhar district were chosen as the study area to assess the relationship between pesticide usage and its impact on quality of groundwater. A software package, Pesticide Impact Rating Index (PIRI), has been used for this purpose (Kookana et al. 2005). PIRI is based on pesticide use, the pathways through which the pesticides are expected to migrate to the water resources (the assets), and the value of the asset.

Based on the frequency of usage of various pesticides in the sampling stations, four pesticides, namely atrazine, chlorophyrifos, phorate and monocrotophos, were selected for detailed analysis employing PIRI. PIRI software was evaluated for its suitability under Indian conditions by arriving at suitable correction factors related to local conditions. Correction factor is the ratio of observed pesticide residue values in groundwater and corresponding model-estimated residues.

Three villages, viz. Boparai, Samailpur, and Malwal in Nakodar Tehsil in Jalandhar district, India, were chosen as the study area for the present study. The study area is densely populated and is a part of the prosperous Satluj Beas Doab Region. The region has fertile alluvial deposits and therefore is primarily an agricultural-dominated area. Apart from agriculture, the subsidiary occupations include dairy, followed by poultry, fishery and beekeeping. The major problems being faced by the region are depletion of the water table, poor soil fertility and small land holding, traditional methods of agriculture, improper use of pesticides, over fertilization of crops and improper spray techniques.

Figure 1 depicts seven different farm locations lying in three villages in Nakodar Tehsil, which were identified for sampling. The soil data, groundwater samples and pesticide usage data is collected from the farmers from the seven sampling stations (SS). The details of the sampling stations (SS) are as follows –

• (a)

  Village Boparai – SS-1 (Sugarcane), SS-2 (Rice – Potato-Winter Maize)

• (b)

  Village Samailpur – SS-3 (Sugarcane), SS-4 (Vegetables), SS-5 (Rice-Wheat)

• (c)

  Village Malwal – SS-6 (Rice – Potato-Winter Maize), SS-7 (Rice – Potato-Winter Maize)

PIRI is a quantitative approach for assessing the potential impact of pesticide on groundwater and surface water and consequently its detriment to living organisms in reach, by employing the data related to pesticide usage, the pathways through which the pesticides are expected to migrate to the water resources (considered as assets), and the value of the corresponding asset (Kookana et al. 2005). Each component is quantified using site hydrogeological conditions, viz. type of soil, its organic matter content, soil porosity, land slope, soil loss, recharge rate and water table depth, and corresponding hydro-meteorological conditions, viz. rainfall and temperature of the area.

This factor is applicable to both surface water and groundwater, but here only transport to groundwater is considered, which is in line with the objectives of the present study.

This factor is important when risks associated with pesticides to water bodies located in different regions are compared. This value is insignificant when pesticides threatening the same asset are compared with each other.

A survey was carried out to collect information from farmers regarding usage of pesticides in specific fields along with the cropping pattern over the last two years (2015–2017). It was observed that the farmers used a variety of pesticides for the same crop due to proliferation of the various brands and types of pesticides, different recommendations by the shopkeepers/company representatives and resistance developed by the target pests to certain pesticides. A total of 21 different pesticides were used in four cropping patterns that were prominent in the study area (Table 1).

Based on the frequency of usage of various pesticides in the sampling stations, out of the above 21 pesticides, a total of four pesticides, namely Atrazine, Chlorophyrifos, Phorate and Monocrotophos, were selected for detailed analysis to PIRI. Table 2 depicts the pesticide properties and application data relevant to the study (rate at which the pesticides are applied and the nozzle size for the spray application), at all seven sampling stations (SS). Environmental persistence (half-life, t_1/2) and the soil sorption coefficient (K_oc) of the four pesticides were obtained from the available literature.

Groundwater samples were collected from various tube wells located in the seven farms for pesticide residue analysis. Tube wells tapped the aquifer between 20 m and 50 m below ground level. Samples were collected after flushing for 5 minutes in the case of bore wells, in order to obtain fresh aquifer water. Sampling bottles made up of high quality dark glass with Teflon stoppers were used. Plastic or polyethylene containers were used initially but later water samples were transferred to glass bottles to avoid the pesticides present in water samples being adsorbed on the inner walls of the bottles.

Liquid–liquid extraction followed by gas chromatography was used for the determination of pesticide residues. A 500 ml groundwater sample was taken in a well-rinsed 1 litre separator funnel and 10 g of NaCl was added to it. The funnel was shaken to dissolve the NaCl completely. The residues were extracted thrice with dichloromethane (50:25:25 ml), shaking vigorously for 2–3 minutes with intermittent pressure release. The separator funnel was kept undisturbed to separate the two layers. The lower aqueous layer was drawn from a 1 litre separator funnel. Three extracts were combined and dried by passing through an adsorbent (2.5 cm ID and 15 cm long) containing anhydrous Na₂SO₄ over a small pad of glass wool at the bottom and collected in a well-rinsed 250 ml flat-bottom flask. The extracts were concentrated up to 1.0 ml with a vacuum rotary evaporator and 10 ml of n-hexane was added to the combined extract and concentrated to 1.0 ml again. The final volume was made up to 2.0 ml with n-hexane solvent and with acetonitrile solvent. A concentrated 2.0 ml sample was analyzed with the help of a gas chromatograph (GC). The instrument detection limits were established by using 3:1 signal to noise ratio to determine a peak as a valid quantifiable peak. Each sample was analyzed in duplicate and the average was used in analytical calculations. Concentrations below the limit of detection were assigned zero values for the statistical analysis. The results obtained are given in Table 3.

The study area was divided into different homogenous units based on the visual observation and farmer's experience, in order to collect the soil samples representing the soil condition of the study area. Surface litter at the sampling spot was removed. At least five samples from each sampling unit were collected. A ‘V’ shaped cut was made up to a depth of 15 cm at the sampling spot and thick soil slices from top to bottom of the exposed face of the ‘V’ shaped cut were removed and placed in a clean container for soil analysis (Figure 2).

After this, the samples were mixed thoroughly and foreign materials (like roots, stones, pebbles and gravels) were removed. A representative soil sample was obtained by applying a well-known quartering and coning method. The results of the parameters for soil have been tabulated (Table 4).

Average annual rainfall in the study area for a period of one year (01 April 2016–31 May 2017) was taken as 551.3 mm (MoEIT 2017). Details of the water supplied to farms by tube wells (total irrigation), depth of borewells (given in Table 5) were ascertained from the farmers, which may not be very accurate, but can be presumed to be reasonably good. These have been tabulated below. The average maximum temperature during this year was 29.93 °C and average minimum temperature during this period was 17.27 °C.

The software package as received from Commonwealth Scientific and Industrial Research Organisation (CISRO), Australia, relates to PIRI version 6. The input parameters necessary for the software run for the assessment of pesticides leaching into the groundwater are organic carbon partition coefficient (Koc), environmental persistence (t_1/2) of pesticide, water table depth (D), total rainfall, total irrigation, soil type, percentage organic content in soil (f_oc). The soil bulk density and moisture content can be given manually and if not, like in the present case, the software uses the suitable values according to the given soil type. The recharge rate (q) is also calculated by the software (if not provided manually) on the basis of the soil type, total rainfall and total irrigation for flat terrain.

For benchmarking of the guideline value for drinking water, a thorough study of the existing guidelines for drinking water for various pesticides was carried out including Draft Indian Standard Drinking Water – Specification (Second Revision of IS 10500), WHO Guidelines for Drinking-water Quality Third Edition, 2008 and USEPA guidelines. The adopted acceptable values for the present study are 35 ppb, 20 ppb and 1 ppb for atrazine, chlorophyrifos and monocrotophos respectively (USEPA 2000). For phorate the limiting values was assumed to be 25 ppb.

Applying the data, model was run to assess the groundwater pollution potential for a period of one year (01 April 2016–31 May 2017) and the results obtained are given in the following section.

The model (PIRI) estimated the pesticide residue/load and its mobility in the groundwater. The model assigns the risk categories to each pesticide and category names were given according to the scoring system based on the application of Equation (9). The groundwater pesticide load risk rating and mobility for each sampling station as assigned by model are shown in Figures 3–9. The attenuation factor and the groundwater pollution potential in terms of pesticide residue/loads obtained for each sampling station are given in Table 6.

The pesticide residues/loads estimated by PIRI are further compared with the observed pesticide residues of the four pesticides under consideration (atrazine, chloropyrifos, monocrotophos and phorate) in all the seven water samples in order to compare the actual risk with PIRI prediction.

The risk rating in terms of total groundwater load was ‘Very high’ for atrazine, ‘High’ for phorate and ‘Medium’ for chloropyriphos. As monocrotophos was not applied in field 1, its risk was classified as ‘Very low’ by the software. With the type of soil being ‘Silt’ and borewell depth 40 m, the mobility was calculated as ‘Medium’ for atrazine, and ‘Low’ for phorate, whereas chloropyrifos and monocrotophos were classified as ‘Very low’ (Figure 3).

The risk rating in terms of total groundwater load was ‘Very high’ for atrazine, ‘High’ for monocrotophos and phorate and ‘Low’ for chloropyriphos. Chloropyrifos, although applied in minor dosage, was not observed. With the type of soil being ‘Silty clay’ and borewell depth 40 m, the mobility was calculated as ‘Medium’ for atrazine, ‘Low’ for phorate and monocrotophos, whereas chloropyrifos was classified as ‘Very low’ (Figure 4).

The risk rating in terms of total groundwater load was ‘High’ for all the four pesticides. With the type of soil being ‘Silty clay’ and borewell depth as 25 m, the mobility was estimated to be ‘Low’ for atrazine, phorate and monocrotophos. Chloropyrifos was classified as ‘Very low’ (Figure 5).

Monocrotophos and chloropyrifos were rated as ‘High’, phorate was rated as ‘Medium’ and atrazine was rated as ‘Low’ groundwater pollution risk. With the type of soil being ‘Silty clay’ and borewell depth 20 m, the mobility was calculated as ‘Low’ for monocrotophos and chloropyrifos, whereas atrazine and phorate were classified as ‘Very low’ (Figure 6).

Phorate was rated as ‘Medium’ and for the rest pesticides it was rated ‘Very low’ as they were not applied in the field. With the type of soil being ‘Silt’ and borewell depth as 30 m, the mobility was calculated as ‘Low’ for phorate (Figure 7).

Monocrotophos was rated as ‘High’, phorate was rated as ‘Medium’, whereas chloropyrifos was rated as ‘Very low’ groundwater pollution risk as they were not applied in the field. With the type of soil being ‘Sandy silt’ and borewell depth 45 m, the mobility was calculated as ‘Low’ for monocrotophos, whereas chloropyrifos and phorate were classified as ‘Very low’ (Figure 8).

In Sample 7, monocrotophos was rated as ‘Very high’, phorate, atrazine, chloropyrifos were rated as ‘Medium’ groundwater pollution risk. With the type of soil being ‘Silt’ and bore well depth 45 m, the mobility was calculated as ‘Low’ for monocrotophos, whereas atrazine, chloropyrifos and phorate were classified as ‘Very low’ (Figure 9).

From the results of observed residues of the four pesticides (atrazine, chloropyrifos, monocrotophos and phorate) in the groundwater samples and the pesticide residues/load estimated by the model (PIRI software), it can be inferred that the observed values were comparatively higher than the model estimated values for most of the cases. The higher observed pesticide load in groundwater could be due to accumulation of pesticide residues despite degradation by various processes, because of the prolonged use of these pesticides over the years. PIRI software has been evaluated only for one year time period and therefore the estimated values are reflecting the residues during a limited time period.

Although the absolute values of observed and model estimated residues did not match, they seemed to bear consistent ratios fluctuating between reasonable ranges. Therefore the correction factor can be applied to the results of the model estimates, which may comply with the Indian scenarios for the chosen study area. This correction factor was estimated by dividing the observed values by the model estimated values, as shown in Table 7. The spatially averaged correction factors along with the range for each pesticide are shown in Figure 10. Multiplicative correction factors have to be imposed to the model estimates so as to predict the realistic pesticide residues in an area.

Pesticide usage has become an inseparable part of modern agriculture. If only the potential impacts of its usage over groundwater resources can be determined, thorough monitoring policies can be made for regulating pesticide application and also the local farmers can be made aware of the problems of improper pesticide usage. This will lead to conservation of ground water quality for various uses as well as future generations.

Based on the results of the study, it is recommended that PIRI software can be introduced for assessment of residues and mobility potential of various pesticides being used in India. For this, the suitable correction factors have to be incorporated based on corresponding local conditions and considered time frame of analysis.