Impact of irrigating with arsenic contaminated water on farmers’ incomes in west bengal

IMPACT OF IRRIGATING WITH ARSENIC CONTAMINATED WATER ON FARMERS’ INCOMES IN WEST BENGAL Madhavi Marwah Malhotra* Abstract With high arsenic contamination of groundwater in West Bengal much beyond permissible limits for irrigation water, and institutional measures aimed at enhancing groundwater pumping to meet the growing food requirements in the country, the long-run sustainability of agricultural production and farmers’ livelihoods in arsenic affected areas are under threat This study undertakes a comparison of the net incomes of farmers earned from crop production between arsenic affected and non- arsenic affected areas’ agricultural situation To analyse the differences in the agricultural situation in detail, the non-parametric Mann-Whitney U test for comparing two samples is used In conclusion, the study finds evidence that farmers using arsenic contaminated water for irrigation for over two decades in West Bengal are now facing triple impoverishment on account of having to adopt a less profitable cropping pattern, lower yield of crops and higher input costs per unit of cultivated land area Key words: arsenic, agriculture, groundwater Introduction Much of Eastern India, underlain with alluvial aquifers receiving plentiful rainfall for recharge, is not posed with a challenge as far as volume of groundwater availability is concerned Particularly, for the state of West Bengal, the stage of groundwater development1(SGD) was about 42 percent in 2013, as compared to Punjab and Haryana with SGD close to 140 percent and 120 percent, respectively (CGWB, 2013) However, there are serious issues on account of groundwater quality deterioration due to contamination with several dangerous chemicals such as arsenic and fluoride Arsenic is well-known to be a carcinogen (cancer-causing substance) While cancer takes a longer time (more than to 10 years) to develop, the most common clinical manifestations of arsenic poisoning are skin lesions Skin abnormalities caused by arsenic ingestion include hyper-pigmentation and keratosis Hyper-pigmentation leads to discoloured spots, dark brown spots and darkening of limbs, whereas keratosis shows up as thickening of palm and feet soles (Mazumder et al, 2010) The extent of arsenic impact on human health is related to the nutritional status of a person, such that the effects are more prominent for those who are malnourished and have low immunity (Das, Roy and Chakrabarti, 2016) The arsenic-led groundwater crisis has burdened the lives of the affected populations, particularly the rural poor who use wells and tube-wells for sourcing drinking, domestic as well as irrigation water, in such a way that it has physical, economic and social impacts While physical impacts include weakness and other illnesses, the economic impacts due to added medical treatment, costs and                                                              Stage of groundwater development = (Annual groundwater draft for all uses / Net groundwater availability) * 100 * Madhavi Marwah Malhotra is a PhD scholar at the Centre for Economic Studies and Policy, Institute for Social and Economic Change, Bengaluru This paper is part of her ongoing PhD research She is grateful to the guidance received from her PhD supervisor, Prof Krishna Raj and doctoral committee members: Prof R S Deshpande, Prof S Madheswaran and Dr A V Manjunatha Madhavi is also thankful to the two anonymous reviewers for their constructive suggestions on the paper The usual disclaimers apply social impacts such as social exclusion, polygamy, threat of divorce and dowry demands, have further worsened their situation (Das, Roy and Chakrabarti, 2016) Wage loss due to arsenic poisoning in rural populations worsens the households’ economic condition (Indu, Krishnan and Shah, 2007) The causal relation between learning outcomes of secondary school children and drinking arsenic-contaminated water has also been established (Asadullah and Chaudhury, 2011) Additional economic impacts of arsenic contaminated groundwater arise on account ofits use for irrigation purposes Several crops and crop varieties have been scientifically tested and proven to be susceptible to arsenic loading in irrigation water and soil in different quantities Crop vulnerabilities are known to be in the form of reduced yield, lower grain weight and uptake of arsenic in the crop, resulting in its entry into the food chain In India, the state of West Bengal is severely affected with the groundwater aquifer in 11 out of 23 districts (104 blocks) having reported arsenic levels much higher than the permissible limit for drinking water set by World Health Organisation at 0.01 mg per litre and even the less restrictive limit set by the Bureau of Indian Standards at 0.05 mg per litre These districts are Malda, Murshidabad, Nadia, North 24-Parganas, South 24-Parganas, Howrah, Hooghly and Bardhaman It can be observed that arsenic presence (orange and red area) is concentrated in the region to the east of the Ganges (see Figure 1) While there are many competing explanations on the occurrence of arsenic in groundwater aquifers of the Indo-Gangetic Plain region, it is well established that arsenic is present in the underground rock formations and gets mixed with the groundwater There are two types of rocks in this region – deep Pleistocene and Holocene depositions – where arsenic presence is known It gets mixed with groundwater due to processes like oxidations of different chemical compounds, pumping of groundwater from the aquifer, etc   Figure 1: Arsenic map of West Bengal Source: Public Health Engineering Department, Government of West Bengal (2014) This paper examines the impacts of arsenic-contaminated water irrigated on famers’ incomes through a field survey in four blocks in West Bengal characteristically differentiated in terms of the groundwater situation The study is relevant in the light of institutional developments aimed at increasing the use of groundwater for irrigation Based on the descriptive analysis of primary data supplemented with econometric analysis including Mann-Whitney U test for comparison of means of selected variables between arsenic and non-arsenic areas, we observe a negative influence of using arsenic contaminated irrigation water on farmers’ incomes Findings from the study are used in deducing policy suggestions for groundwater management in arsenic affected areas   Statement of the Problem Arsenic contamination in groundwater is widespread and found to be much above the permissible/threshold limits for drinking (0.01 mg per litre by World Health Organization) and irrigation purpose (0.05 mg per litre by Food and Agriculture Organization) in the state of West Bengal Given the adverse impacts of using arsenic concentrated water on human health and agriculture, policies and programmes towards increasing groundwater extraction for agriculture in this region can have serious implications on the sustainability of agriculture as West Bengal is the largest rice producing state in India Since the states which were once known as the ‘Food Bowl of India’ are unable to fulfil the food requirements of the growing population, the government has moved towards the eastern states to bring about a second Green Revolution in India (Government of India, 2015) In this regard, the ‘Bringing Green Revolution to Eastern India’ scheme was launched in 2011 under the Rashtriya Krishi Vikas Yojana, covering the eastern States of Assam, Bihar, Chhattisgarh, Jharkhand, eastern Uttar Pradesh, Odisha and West Bengal Under this scheme, “100 percent assistance is provided for construction activities (INR 30,000/ dug well/ bore well and INR 12000/shallow tube well)” (BGREI scheme guidelines) In addition, a 50 percent subsidy on cost of pumpsets, up to INR 10,000 has also been provided (ibid.) Moreover, the state government of West Bengal is itself motivated to encourage groundwater extraction for irrigation purposes, to ensure agriculture sector growth To achieve this goal, two important initiatives have been taken in 2011 with the relevant government department committing to provide access to pumping facilities at reduced setup costs Specifically, in February 2011, the West Bengal State Electricity Distribution Company Limited (WBSEDCL) passed a policy resolution stating that it would provide new electricity connections to farmers against a payment of a fixed fee amounting between Rs 1,000 and Rs 30,000 per connection, depending on the connected load This meant a reduction in expenses for farmers who would no longer bear the full cost of wires, poles and transformers, as required earlier (Mukherji, Shah and Banerjee, 2012).In addition, the Water Resources Investigation and Development Directorate (WRIDD) in Bengal, vide a memo dated November 9, 2011, changed a provision of the aforementioned Act As per this amendment, farmers located in “safe” groundwater blocks and owning pumps of less than horsepower (HP) and tube wells with discharge of less than 30 cubic metres per hour would no longer need a permit from the State Water Investigation Department (SWID) to apply for electricity connection from WBSEDCL With this new amendment, farmers other than those in the semi-critical and critical blocks (53 semi-critical and critical as of 2011; 76 semi-critical and critical as of 2013) would be outside the purview of the Act, making it easier to put in an application for electricity connection in safe groundwater blocks (Mukherji, Shah and Banerjee, 2012; CGWB, 2013) As per the CGWB assessment, groundwater extraction for irrigation in West Bengal has increased from 9.72 billion cubic metres (bcm) to 10.84 bcm between 2011 and 2013 The point to be noted here is that the blocks in ‘safe’ category are assessed only with respect to the depth of water level by the Central Ground Water Board (CGWB) It has no bearing on the quality aspect of groundwater, which means that the policies not restrict groundwater extraction in ‘safe’   but arsenic contaminated blocks Table below shows groundwater extraction for irrigation in arsenic affected blocks of West Bengal over the years from 2009 to 2013 Table 1: Groundwater Extraction for Irrigation (bcm) in Arsenic Contaminated Districts of WB Districts 2009 2011 2013 Murshidabad 163042 165522 186187 North 24-Parganas 77292 78061 84524 Nadia 167617 168092 179222 Source: CGWB 2011, 2014, 2017 With respect to linkages between the level of arsenic content in groundwater and groundwater extraction, Ghosh and Singh (2009) report that arsenic gets mobilized into the aquifer with continuous groundwater pumping The concentration of arsenic, however, is known to be higher in the shallow aquifers of the Bengal Delta Plain The deep, confined aquifers are relatively free from arsenic, whereas the shallow unconfined waters are the contaminated ones (Government of West Bengal, 2005) Study Objective Pitt, Rosenzweig and Hassan (2012) studied the impact of arsenic ingestion on incomes of rural households in Bangladesh using household expenditure data and found evidence of lower labour productivity due to the health impacts of arsenic intake They further estimated its impact on women’s productivity in doing household work Similarly, men, who are the primary wage earners, had a lower labour productivity and therefore had lower than usual working and earning capacity due to the intake of arsenic The consumption of purchased goods and production of household goods were found to be significantly lower for households which use arsenic contaminated tubewell water for drinking and cooking purposes Impacts in terms of decline in crop productivity and deteriorating soil fertility are likely to increase the cultivation costs and squeeze the profit margins for the already impoverished agriculturebased population However, to the best of the authors’ knowledge, there is no study to have estimated the loss of farmers incomes on account of lower crop yields and higher cultivation costs other than labour To fill this gap in literature, the study aims to estimate the monetary impact on farmers’ incomes of irrigating with arsenic-contaminated water The study has its limitations in the sense that it considers the impacts only in terms of realised costs of cultivating crops The impact on family labour productivity and costs due to additional hired labour thereof is complex and therefore a statistical significance test is carried out to compare all variables of interest across the arsenic and non-arsenic sample areas Gross returns to a farmer from production include only the returns from selling the main crop output and not those received on account of by-products Further, it fails to attribute the income losses to different factors causing a yield decline or low selling price, i.e how much is on account of arsenic and how much is due to other factors   Review of Literature ™ Occurrence and mobilization of arsenic in groundwater Developments in natural sciences have been engaged in understanding the occurrence and mobilisation of arsenic in groundwater Primarily, arsenic occurs in two forms: trivalent and pentavalent states, i.e Arsenite As(III) and arsenate As(V) respectively The toxicity of various forms of arsenic strongly depends on their oxidative states and chemical structures The inorganic forms of arsenic present in soil, when taken up and transported through the food chain, turn out to be toxic, affecting various life forms Among the two oxidation states As(III) and As(V), As(V) is less toxic and mostly present in immobile mineral forms, whereas the As(III) form is more toxic and gets mobilized into water and enters living cells (Shrivastava et al 2015) Ravenscroft et al (2009) have described four geochemical mechanisms of natural arsenic pollution: reductive dissolution, alkali desorption, sulphide oxidation and geothermal activity, of which the first is the most relevant in the South Asian context Reductive dissolution occurs when arsenic adsorbed2 to iron oxyhydroxides in sediments gets mixed with groundwater because of microbial degradation of organic matter which reduces ferric iron to the soluble ferrous form The presence of arsenic is in the relatively un-withered alluvial sediments derived from igneous and metamorphic rocks in the Himalayas and related young mountain chains (Brammer and Ravenscroft, 2009) Similar results were reported by Nickson et al (2000) and UNICEF (2008) who find reductive dissolution of iron oxyhydroxides present in sediments as the source of arsenic release into groundwater in the Ganges Plain area of West Bengal and Bangladesh In another study, Smedley and Kinniburgh (2002) found that high-arsenic groundwater was not related to areas of high arsenic concentration in the source rock Two key factors were identified: first, there should be very specific biogeochemical triggers to mobilize arsenic from the solid phase to groundwater, and second, the mobilized arsenic should have sufficient time to accumulate and not be flushed away, that is, it should be retained in the aquifer In other words, arsenic released from the source should be quick, relative to the rate of groundwater flushing There are a number of processes for the mobilization of arsenic in groundwater namely, (i) mineral dissolution, (ii) desorption of arsenic under alkaline and oxidizing conditions, (iii) desorption and dissolution of arsenic under reducing conditions, (iv) reduction of oxide mineral surface area, and (v) reduction in bond strength between arsenic and holt mineral surface (Smedley and Kinniburgh, 2002) Oxidation of sulphide minerals (pyrite-FeS ) is another hypothesis which has been conceived as the cause of groundwater arsenic contaminationin West Bengal According to this hypothesis, arsenic is released from the sulfide minerals (arseno-pyrite) in the shallow aquifer due to oxidation (Mandal et al, 1998) The lowering of the water table owing to over- exploitation of groundwater for irrigation is the cause of the release of arsenic Some investigators explained that excessive use of water for irrigation and use of fertilizers have caused the mobilization of phosphate from fertilizers down below the shallow aquifers, which have resulted in the mobilization of arsenicdue to anion exchange onto the reactive mineral surfaces Sikdar and Chakraborty (2008) asserted that the combined processes of recharge of                                                              Adsorption: the adhesion in an extremely thin layer of molecules (as of gases, solutes or liquids) to the surfaces of solid bodies or liquids with which they are in contact (Merriam Webster dictionary)   groundwater from rainfall, sediment water interaction, groundwater flow, infiltration of irrigation return water (which is arsenic rich due to the use of arsenic-bearing pesticides, wood preservatives, etc and the pumping of arsenic-rich groundwater for agriculture purpose), oxidation of natural or anthropogenic organic matter and the reductive dissolution of ferric iron and manganese oxides, are the key factors in the evolution of groundwater arsenic contamination in the area It seems that there are a number of hypotheses, which have their own discrepancies and limitations to explain the physical processes Shrivastava et al (2015) report that arsenic mobilization in the Bengal Basin can happen due to the discharge of arsenic into alluvial sediments by the oxidation of arsenic-containing pyrite, and displacement of anions of arsenic present in aquifer sedimentary minerals by phosphate anions used in fertilizers which are applied on the soil surface, and discharge of arsenic in anoxic conditions by the reduction of iron oxyhydroxide during sediment burial Specifically, for India, Ghosh and Singh (2009) report that the release of arsenic, by the natural processes in groundwater, has been recognized, from the Holocene sediments comprising sand, silt and clay in parts of the Bengal Delta Plains (BDP), West Bengal and in the Gangetic plains of Bihar Several isolated geological sources of arsenichave been recognized, viz Gondwana coal seams in Rajmahal basin, Bihar mica-belt, pyrite-bearing shale from the Proterozoic Vindhyan range, Son valley gold belt and Darjeeling Himalayas belt The contaminated aquifers are of Quaternary age and comprise micaceous sand, silt and clay derived from the Himalayas and the basement complexes of eastern India These are sharply bound by the River Bhagirathi-Hooghly in the west, the rivers Ganges and Padma in the north, the flood plain of the River Meghna and the River Yamuna in the northeast (Ghosh and Singh, 2009) Another important finding in the literature is the high degree of spatial variation in arsenic levels over a few metres of the same aquifer as well as lateral variation in the same well at different depths (The World Bank, n.a.) This means that generalization with a few sample tests lacks credibility and rather a more wide scale individual testing of all wells is needed to ascertain the extent of contamination Furthermore, the linkage between groundwater pumping and arsenic mobilization, although not well established, has been brought under discussion In this context, it is suggested that groundwater development should be undertaken cautiously after thorough laboratory tests and assessment of potential threats of contamination in unexploited areas or of worsening arsenic levels in already exploited regions (UNICEF, 2008) ™ Agronomic impacts of irrigating with arsenic contaminated water While arsenic is well-known to be extremely harmful to human health through both external exposure to the skin as well as consumption of arsenic-contaminated water, another branch of literature deals with understanding the impacts of using the same for irrigation purposes Since the last decade or so, research on irrigation with arsenic-rich water has received much attention, primarily due to the increasing dependence on groundwater as a major source of irrigation and its role as a critical input in producing food for the burgeoning population In South Asia,   Bangladesh is the worst affected by arsenic contamination in terms of both population at risk as well as areal extent, followed by India Here, we explore the existing literature to understand the following relevant issues: What are the impacts on crops of irrigating with water contaminated with arsenic? Within each impact, what are the heterogeneities across different crops and crop varieties? What are the impact mechanisms? What are the possible impacts on other agricultural inputs like soil and labour? There are three main impacts on crops identified from the literature These are: Crop uptake of arsenic and impact on crop yield We also find a negative impact on tangible agricultural inputs like soil The following three sub-sections discuss each of these impacts and the related heterogeneities Arsenic uptake by crops Several studies have confirmed the uptake of arsenic by different crops when irrigated with arsenic-rich water and grown in arsenic accumulated soil However, the amount and type of arsenic found in crops tends to differ, depending on several other factors including crop type, crop variety, soil type and so on Therefore, it is safe to say that the relationship between the arsenic level in irrigation water and soil, and the amount of arsenic uptake in crops, is extremely complex A majority of the arsenic affected areas in India and Bangladesh have a rice-based cropping pattern with three rice seasons in a year and studies show differences in arsenic uptake of rice grown during different parts of the year Analysis of rice samples of both winter and summerseasons of year 2000 from four selected villages in Bangladesh, representative of different agricultural and soil conditions as well as different rice varieties, was undertaken by Duxbury et al (2003) The study revealed arsenic content in summer rice (108 to 331 mg/kg) was higher than that grown in winter season (72 to 170 mg/kg) and mean arsenic level in summer rice was 1.5 times more as compared to winter rice (at percent level of significance) Similarly, a statistically significant difference was found in grain arsenic content of winter and summer rice samples collected from the same districts, such that arsenic level in summer rice was 1.3 times higher than winter rice (Williams et al 2006) The primary reason for such a result is that summer rice is entirely dependent on irrigation, which is largely groundwater based and so there is much more arsenic addition during this time of cultivation The arsenic uptake by winter rice is a residual impact of arsenic-rich irrigation during summer cultivation Winter rice is majorly rainfed but the arsenic accumulation in soil during summer cultivation causes the arsenic uptake during this season Another important factor in paddy cultivation is the soil condition – aerobic or anaerobic In the aerated soils, arsenic presence is in arsenate (AsV) form, which is mostly unavailable for crop uptake as arsenate gets adsorbed by iron hydroxides In anaerobic soils like flooded paddy cultivated lands, arsenic is readily available to crops since it is present in its reduced form, i.e arsenite (AsIII) (Brammer and Ravenscroft 2009; Xu et al 2008) Arsenic accumulation in crops is also determined by the longevity of groundwater irrigation in the affected regions This can be explained by the long-term arsenic buildup in the soil due to continuous irrigation with the contaminated water In a Bangladesh study, Williams et al (2006) analyzed 330 winter and summer rice samples, 94 vegetables, 50 pulses and spices samples for arsenic   and reported the highest arsenic rice grain levels in the samples from south-west districts These districts are characterized by high arsenic water concentrations, extensive use of shallow tube wells for rice cultivation and a much longer history of groundwater irrigation Not all crops take up equivalent amounts of arsenic In a China-based study, comparing arsenic accumulation in wheat and rapeseed, it was found that wheat accumulates more arsenic than rapeseed (Liu et al 2012) While both wheat and rapeseed were found suitable for cultivation given soil arsenic content of to 60 mg/kg, rapeseed is preferable beyond 80 mg/kg A more general finding is that plants belonging Cruciferae type such as mustard, rapeseed etc have high tolerance to arsenic (Liu et al 2012; Chaturvedi 2006; Zhong et al 2011) Williams et al (2006) report substantial differences in the arsenic levels between and within different types of vegetables, the mean maximum arsenic concentration being higher in the root and tuberous vegetables than the fruit vegetables, and lowest in leafy vegetables Similar findings were reported from West Bengal in the report of the Inter-Ministerial Group (IMG) on Arsenic Mitigation (2015) They state higher arsenic accumulation in potato, brinjal, arum, amaranth, radish, ladies finger, and cauliflower, and relatively low levels of arsenic accumulation in beans, green chilli, tomato, bitter gourd, lemon and turmeric The major oil seeds and pulses have been found to contain a high arsenic content Moreover, the high yielding rice varieties have more arsenic accumulation than the local varieties (IMG on Arsenic Mitigation, 2015) Furthermore, even within crops, different parts of the plant take up different arsenic quantities In rice crop, the arsenic accumulation in ascending order of plant parts follows the order: economic produce, leaf, stem, root (Das et al, 2013) For rice in particular, the accumulation followed the order: root, straw, husk, grain with highest arsenic accumulation in the root and lowest in grain (Abedin et al, 2002) Although the present study does not look into human health or food chain implications of arsenic contamination, understanding these dynamics within arsenic uptake by crops would prove to be useful in developing holistic adaptation strategies which address both human health and economic impacts including productivity in terms of cropping pattern in affected areas Arsenic impact on crop yields Theoretically, Heikens (2006) explains the process and trajectory of arsenic accumulation in soil and the resultant impact on crop yields due to continuous irrigation with arsenic contaminated water (see Figure 2) yield  As in crops  As in soil  Figure 2: Arsenic Accumulation in Soil and Impact on Crop Yields Time  As in soil   As in soil As arsenic from irrigation water accumulates in the soil, crop concentrations of arsenic also tend to increase beyond a point, depending on bioavailability, uptake and transport However, after a certain level of soil arsenic content, the plant growth becomes severely inhibited and arsenic concentration in the plants are then no longer relevant, as shown by the dotted line Crop yields may not be severely affected until a threshold level, beyond which yields will fall sharply (Heikens, 2006) A substantial number of scientific research studies point towards crop yield reduction due to continuous irrigation with arsenic concentrated water Khan et al (2010) for instance found that arsenic addition in either irrigation water or as soil-applied arsenic resulted in yield reductions from 21 to 74 percent in summer rice and to 80 percent in winter rice, the latter indicating the strong residual effect of arsenic on subsequent crops Hossain (2005) also found yield reductions of more than 40 and 60 percent for two popular rice varieties (BRRI Dhan-28 and Iratom-24), when 20 mg/kg of arsenic was added to soils, compared to the control In a controlled pot experiment study, Abedin et al (2002) found that arsenatecontaminated irrigation water accounted for 26, 38, 56 and 65 percent rice yield reduction by the addition of 1, 2, and mg arsenic respectively for BR-11 variety Moreover, the number of rice grains (filled spikelets) also decreased significantly (at percent level of significance) to 120, 106, 77 and 61 with 1, 2, and mg arsenic treatment, respectively as compared to 160 in the control group Abedin et al (2002) also observed a decline in the weight of rice grain with arsenic addition such that the highest thousand grain weight (i.e., mass of 1,000 grains) of 19.8 grams was found in the control case which decreased to 18.3 grams in the highest arsenate treatment case Azad et al (2012), Panaullah et al, (2009), Huq et al (2006) and Pigna et al (2009) all present similar findings In another study by Das et al (2013), the number of active seedlings per pot was unaffected until 15 mg/kg of arsenic addition in soil, but was reduced significantly beyond this level in the case of summer paddy The study found that filled and matured grains per panicle were reduced substantially after 10 mg/kg of soil arsenic as compared to the control pot, the decline being65.2 percent with 60 mg/kg of soil arsenic As far as yield was considered, the decline was as high as 80.8 percent in the case of 60 mg/kg arsenic content in soil (Das et al, 2013) Yields of sixteen potato varieties were tested for arsenic accumulation in Bangladesh, which revealed that there was a negative influence of arsenic on potato yield in different degrees (Haque et al 2015) The impact was as high as 66 percent reduction in ‘Jam Alu’ variety with 50 mg/kg arsenic addition in soil (ibid.) Wheat and mustard seem to be less sensitive than potato to arsenic contaminated soils, which possibly explains the cropping pattern across the sample blocks In a study in China, Liu et al (2012) found that there was no impact of arsenic until 60 mg/kg of arsenic concentration in soil and yield reduction was observed only beyond 80 mg/kg of arsenic For mustard, there is no conclusive evidence indicating adverse impact on its yield The point at which arsenic accumulation in soil reaches a ‘toxic level’ or ‘upper limit’ is not clear and varies with crop and crop variety Das et al (2008) for instance, show that only arsenic levels exceeding10 mg/kg arsenic in soils affect plant productivity and growth Meharg and Rahman (2003) 10   The usual tests for comparing two samples are the paired samples t-test, independent samples t-test and one-way ANOVA These tests come in the category of parametric tests, whose results are sensitive to assumptions about the two sample distributions These assumptions are as follows: Dependent variable should be a continuous variable Eg Height, weight, temperature Independent variable should consist of two categorical, unrelated groups Eg Male/female; treated/untreated Independence of observations such that there is no relationship between observations in group There are no significant outliers Dependent variable is normally distributed for each of the categories of the independent variable Homogeneity of variances and group or within each group In case any of these assumptions not hold for the samples, we must use non-parametric tests, i.e tests which not make any strict assumptions about the data The non-parametric alternative to two independent sample t-tests is the Mann-Whitney U test Since our sample distributions not meet the necessary assumptions for using parametric tests, we use the Mann-Whitney test for comparison between arsenic and non-arsenic samples • Wilcoxon ranked sum test or Mann-Whitney U test: How does it work? The same test was independently constructed by Mann and Whitney using U-statistic and then by Wilcoxon based on the sum of ranks Consider two independent samples – one is the experimental group (E) having received a certain treatment and the other is the control group (C) If the median of E is larger than the median of C, then in general, the sample scores of E would be larger than those of C To check if this holds true, we write the scores of the two samples in one array in ascending order and assign ranks to each score The sum and mean of ranks for each group is calculated separately and compared If null hypothesis (Ho: ME = Mc) holds true, then the medians are almost equal and two average rank sums are almost equal The alternative hypothesis is that rank sums are not equal (Gibbons, 1993) Rejecting the null hypothesis implies that the two samples come from different populations and also shows which of the two has a greater median Although the test statistic compares the medians of the two groups, we can conclude that the populations from which the samples were collected are different, which implies that the difference in other statistics such as mean are also statistically significantly different Results and Discussion Comparing the average net returns per hectare of cultivated land in each of the three cropping seasons: Kharif, Rabi and Summer reveals huge differences between that for farmers in arsenic affected blocks and the blocks not affected by arsenic The study observes that the difference between mean (or average) net returns per hectare of cultivated land for a farming household in the non-arsenic and 17   arsenic areas are Rs 18,943 in Kharif season, Rs 36,686 in Rabi season and Rs 36,686 in summer season Mann-Whitney U test confirms that these findings are statistically significant at percent level of significance (see Tables 1A, 2A and 3A) Since this is a crude comparison, we explore the possible explanations for these differences by considering one season at a time Net returns per hectare directly depend on the gross returns and the costs of production/cultivation Gross returns depend on type of crop, yield, per unit selling price of the crop output and by-product Costs of cultivation concept used is in line with ‘cost A1’, ie actual expenses in cash or kind for production, of the Commission for Agricultural Costs and Prices (CACP) The price of output used in calculating the net returns is the market selling price of output by the producer, ie the price at which he sold his crop output; whereas production takes into account the total output of a crop produced per hectare of cultivated land without excluding the quantity that is kept for household consumption For the purpose of our analysis, we only consider the returns from selling the main product, and not by-products The cropping pattern is analysed on the basis of the crop having the highest share in the gross cropped area of the household Table 10 shows season-wise primary crop in the block along with percentage of sample households growing that crop Table 10: Block-wise Cropping Pattern Kharif Rabi Summer Arsenic, depleted Arsenic, non-depleted Non-arsenic, depleted Non-arsenic, nondepleted 89 (paddy) 95.5 (paddy) 100 (paddy) 98 (paddy) 70.1 (wheat) 91.2 (mustard) 93.4 (potato) 53 (mustard) 47 (potato) 81.9 (jute) 89.5 (jute) 62.7 (paddy) 84.3 (paddy) The other important component determining net returns is the cost of cultivation Comparing the costs of cultivation ie fertiliser cost, cost of irrigation, seed cost and labour cost, between arsenic and non-arsenic areas is of relevance While fertiliser cost can be compared on average between the two groups, the irrigation cost needs to be further segregated by taking into account the different categories depending on their participation in the water market • Kharif season (July to September) In the Kharif season, the crop grown is uniform across all sample blocks Since West Bengal experiences high rainfall, paddy cultivation in this season is easy and therefore, widely prevalent across all areas Analysing paddy output per hectare of cultivated area, the average output for arsenic blocks is 3,623 kg per hectare, whereas the same is 4,850 kg per hectare in non-arsenic blocks The lower selling price of paddy output is also one of the factors leading to lower net returns in arsenic blocks (Rs 10 per kg) visà-vis non-arsenic blocks (Rs 11 per kg) Mann-Whitney U test confirms the statistical significance of these results at percent level of significance (refer to Table 4A and 5A) Therefore, gross returns for paddy cultivation are reduced in arsenic areas on account of both lower yields and lower selling price of the output 18   As for fertiliser costs of paddy cultivation, the average cost per hectare in arsenic areas is Rs 8,230 and Rs 5,874 in non-arsenic areas Statistical significance at percent level of significance of the difference in fertiliser costs of the two groups is confirmed by Mann-Whitney U test (refer to Table 6A) Although the cost of groundwater irrigation per hectare of paddy cultivated area is higher in our sample from non-arsenic blocks than that from arsenic blocks, this is not a reliable indicator since the proportion of water buyers and pump owners is not equal across all the blocks What is of relevance is that the comparison between irrigation costs of water buyers and pump owners between the arsenic and non-arsenic areas reveals higher cost for water buyers as well as pump owners This is primarily due to the fact that there is a higher rate of electrification in the non-arsenic blocks which allows for cheaper groundwater extraction as compared to the diesel pumps which are more popular in our sample of arsenic blocks Table 11: Costs and returns of Kharif paddy cultivation in arsenic and non-arsenic blocks (in Rs) Yield of paddy (kg/ha) Arsenic Non-arsenic 3,623 4,850 Selling price of paddy (kg) Gross returns 10 11 37,731 53,927 Seed cost 4,465 3,582 Fertiliser cost 8,210 5,875 Irrigation cost 3,160 5,375 Labour cost 8,048 6,394 Net returns 13,847 32,790 Source: Author’s calculations While irrigation cost forms a relatively small component in the kharif season, it is the lower productivity and output prices together with higher fertiliser costs which drain the kharif agricultural incomes of farmers in arsenic areas • Rabi season (October to February) Rabi is winter season with scanty rainfall and temperatures are too low to support the cultivation of a wide variety of crops The main crops grown during this season in arsenic sample areas include wheat and mustard, whereas in the non-arsenic sample areas, farmers mostly grow potato and a few grow mustard Yields of sixteen potato varieties were tested for arsenic accumulation in Bangladesh which revealed that there was a negative influence of arsenic on potato yield in different degrees (Haque et al 2015) The impact was as high as 66 percent reduction in ‘Jam Alu’ variety with 50 mg/kg arsenic addition in soil (ibid.) Wheat and mustard seem to be less sensitive than potato to arsenic contaminated soils, which possibly explains the cropping pattern across the sample blocks In a study in China, Liu et al (2012) found that there was no impact of arsenic until 60 mg/kg of arsenic concentration in soil and yield reduction was observed only beyond 80 mg/kg of arsenic For mustard, there is no conclusive evidence 19   indicating adverse impact on its yield In fact, our study finds that yield of mustard was higher in arsenic affected block (1100 kg per hectare) as compared to the control (728 kg per hectare) indicating that mustard yield may be resistant to arsenic in soil and irrigation water The findings are also in line with the fact that irrigation water requirement among these three crops is highest for potato, followed by wheat and the lowest for mustard Since comparison of yield and selling prices is not possible for all crops given that the primary crops grown are not the same across the two groups (arsenic and non-arsenic), we compare the gross returns Mean gross returns per hectare of cultivated area during Rabi in the arsenic areas is Rs 35,802 and Rs 64,107 for the non-arsenic areas Average fertiliser costs per hectare are Rs 5,595 for non-arsenic blocks and Rs 6,720 for arsenic blocks This result is statistically significant at percent based on Mann-Whitney U test (refer to Table 7A) Table 12: Costs and returns of Rabi cultivation in arsenic and non-arsenic blocks Arsenic Non-arsenic Yield of mustard (kg/ha) 1,103 728 Selling price of mustard 31 31 35,802 64,107 1,962 1,451 6,720 5,595 11,980 6,580 Average gross returns # Seed cost# Fertiliser cost # Irrigation cost# # 6,445 5,103 Net returns# 8,692 45,378 Labour cost #across all crops;inrupees Source: Author’s calculations Therefore, the higher costs of cultivation – seed, fertiliser, labour, irrigation and lower returns from produced output due to lower productivity and the chosen cropping pattern explain the lower net returns in the arsenic sample areas • Summer (March to June) Farmers in the arsenic affected sample blocks primarily grow jute, whereas in the non-arsenic blocks, paddy cultivation is predominant during the summer season As highlighted earlier, Azad et al (2012), Panaullah et al, (2009), Huq et al (2006), Das et al (2013) and Pigna et al (2009) have observed a decline in the yield of paddy when irrigated with arsenic contaminated water Similar observation is made from our study as close to 40 percent of sample farmers in one of the surveyed arsenic blocks reported that they were forced to diversify to jute and vegetables from paddy, since paddy yield was shrinking over the years and eventually summer paddy cultivation became unprofitable This is consistent with the trend observed from the latest available district- level secondary data on area under principal crops in Murshidabad (where Raninagar-II block is located) and North 24Parganas (where Basirhat-I block is located), as shown in table below 20   Table 13: Area (in thousand hectares) Under Some Principal Crops in North 24-Parganas and Murshidabad (2008-09 to 2012-13) Crops 2008-09 2009-10 2010-11 2011-12 2012-13 131.3 121.6 113.5 114.1 172.0 157.7 157.6 159.5 Murshidabad Summer paddy 135.7 Jute 147.2 North 24-Parganas Summer paddy 95.2 88.7 77.6 72.0 70.5 Vegetables 68.69 68.79 72.87 71.16 71.56 Fruits 19.66 20.05 20.36 20.93 21.22 Jute 54.3 53.7 50.2 56.7 47.4 Source: District Handbooks 2014, Government of West Bengal For jute, grown only in the arsenic blocks, we find that average yield is similar in the two blocks i.e 2,281 kg per hectare in groundwater depleted block and 2,106 kg per hectare in the nondepleted block Mean gross returns per hectare for farmers in arsenic blocks are estimated at Rs 53,009 and that for farmers in non-arsenic blocks at Rs 60,065 The Mann-Whitney test shows the significance of the difference in samples at percent level Average fertiliser costs per hectare are Rs 6,211 for arsenic blocks and Rs 4,850 for nonarsenic blocks, the result being statistically significant at percent level based on the Mann-Whitney U test (refer to Table 8A) While all costs of cultivation, i.e seed, fertiliser, labour, irrigation are higher for those in arsenic areas, irrigation costs between the two areas have the maximum differential During the summer season, since crop irrigation requirements are high for both paddy and jute and since arsenic areas are diesel dependent, irrigation costs are much higher Table 14: Costs and Returns of Summer Cultivation in Arsenic and Non-arsenic Blocks (in Rs) Arsenic Non-arsenic Average gross returns 53,009 60,065 Seed cost 1,454 1,520 Fertiliser cost 6,211 4,850 Irrigation cost 44,065 11,812 Labour cost 10,477 8,557 Net returns 15,806 41,197 Source: Author’s calculations Overall, we find that as compared to other seasons, there is a higher percentage of farming households in both categories who not cultivate at all in summer This is particularly so in the diesel dependent areas due to high cost of irrigation for both diesel pump owners and their water buyers 21   Perception-based analysis of long-term changes in agricultural parameters We begin by analysing long term changes in cropping pattern Table 15: Change in Crop Yields of Major Kharif Crop (paddy) Over 10 Years Paddy (Kharif) Arsenic, depleted Arsenic, non-depleted Non-arsenic, depleted Non-arsenic, non-depleted Increase Decrease Households % change 19 (38 %) 49 19 (32 %) 20 (35 %) Households -35 11 (22 %) -30 25 (42 %) -37 27 (48 %) -26 31 (54 %) 15 (26 %) (2 %) 48 (11 %) 55 No change % change 20 (40 %) 45 28 (50 %) Households Source: Author’s calculations It is clearly evident that the proportion of paddy farmers reporting an increase in crop yield over the last 10 years is much higher in the non-arsenic blocks Moreover, the average paddy yield after the increase is also substantially higher Next, the proportion of paddy farmers reporting a decline in the yield over the last decade is significantly higher in Raninagar-II and then Basirhat-I, i.e the arsenic blocks vis-à-vis the non-arsenic ones As earlier, the post-decline average yield is also relatively much lower in the arsenic blocks Table 16: Change in Crop Yields of Major Rabi Crops Over 10 Years Increase Potato Non-arsenic, depleted Non-arsenic, non-depleted Decrease Households % change 17 (30 %) 40 (8 %) Households No change % change Households (2 %) 38 (68 %) (4 %) 21 (88 %) Mustard Arsenic, non-depleted 18 (36 %) 42 (18 %) -25 23 (46 %) Arsenic, depleted (11 %) 35 (35 %) -18 14 (54 %) Source: Author’s calculations In the case of mustard yield, the results show the opposite trend as in the case of kharif paddy In the arsenic block, the percentage of farmers experiencing an increase in yield is more than in the non-arsenic block Moreover, both the before and after the increase average yields are higher for the former Furthermore, the percentage of farmers reporting decrease in the yield is lower in the arsenic block whereas the average yield both before and after the decline is higher, in comparison to the non-arsenic block These results corroborate the fact that mustard is a less water-intensive crop and hence its production remains unaffected by the quality of irrigation water applied 22   Table 17: Change in Crop Yields of Major Summer Crops Over 10 Years Increase Paddy Households Decrease % change Households No change % change Households Non-arsenic, depleted 13 (39 %) 50 0 20 (61 %) Non-arsenic, non-depleted (20 %) 38 (7 %) -20 32 (73 %) Arsenic, depleted 11 (27 %) 53 17 (41 %) -28 13 (32 %) Arsenic, non-depleted (16 %) 51 11 (25 %) -30 26 (59 %) Jute Source: Author’s calculations Summer paddy, which is grown only in the two non-arsenic blocks, shows higher before and after increase average yield as well as higher percentage of cultivating farmers reporting an increase, in the groundwater ‘critical’ block as compared to the groundwater ‘safe’ block The reasons reported for the increase are change of seed variety and increased fertiliser application In jute production, the arsenic and depleted block shows higher percentage of farmers indicating an increase as well as decrease as compared to the arsenic and non-depleted block The primary causes of decline in jute yields reported in the two blocks are poor quality of irrigation water and declining soil fertility Table 18: Change in Quantity of Fertiliser Applied (per unit of land), Over 10 Years Increase Less than double Double More than double Total Decrease No change Arsenic, depleted 17 25 51 (88) (1.5) (10.5) Arsenic, non-depleted 15 12 17 44 (75) 15 (25) Non-arsenic, depleted (12) 51 (88) Non-arsenic, non-depleted 13 23 (40) 34 (60) * figures in (.) are percentages Source: Author’s calculationsFrom simple descriptive statistics, it can be seen that the percentage of farmers who increased fertiliser use per hectare of crop land is substantially higher in the two arsenic blocks Increasing fertiliser application seems to be one of the main adaptation strategies adopted by farmers to deal with the problem of irrigating with arsenic-concentrated water Table 19: Perception about change in soil fertility over 10 years Increase Decrease No change Arsenic, depleted 52 (90) (10) Arsenic, non-depleted 35 (60) 24 (40) Non-arsenic, depleted 11 (19) 18 (31) 29 (50) Non-arsenic, non-depleted (10) 30 (53) 21 (37) * figures in (.) are percentages In terms of perception about soil fertility, 90 percent of farmers in arsenic and depleted block reported that soil fertility has deteriorated over the years Similarly, the same percentage was 60 % for farmers in the arsenic and non-depleted blocks Again, the figures are lower for the non-arsenic blocks, 23   pointing towards the fact that arsenic contaminated irrigation over the long-run leads to deterioration in soil fertility Conclusion and Policy Implications Arsenic can be considered as a drain on farmers’ health and wealth, as it entangles them into a vicious circle of poor health and low incomes The already impoverished households in these areas with poor nutritional status succumb to the disease burden due to arsenic ingestion, rendering them incapable of serving as labour at an earlier than normal age While medical costs increase, incomes reduce due to lower earning capability and lower profits from agriculture (Indu, Krishnan and Shah, 2007) With increasing dependence on groundwater for irrigation purposes in the Bengal delta, use of this resource which is replete with arsenic is likely to have severe implications for agricultural sustainability in the region and at large, for India’s food security which crucially depends on agricultural production in the eastern states While on the one hand, arsenic in irrigation water has been an entry point into the food chain, on the other hand, it is known to deteriorate fertility of soil and reduce crop yields over the long-run.In this light, institutional measures towards enhancing use of arsenic contaminated water for irrigation are likely to worsen the problem The study provides evidence of stark differences in the agricultural situation between arsenic affected areas vis-à-vis areas not affected by arsenic Farmers in arsenic affected areas seem have adopted a cropping pattern that is relatively less sensitive to arsenic impact on crop yield, i.e mustard and wheat instead of potato in rabi season and jute instead of paddy during summer Furthermore, for the similar crop grown across arsenic and non-arsenic sample blocks, i.e kharif paddy, we clearly observe a statistically significant difference in the crop yield between the two areas based on MannWhitney U test These two factors substantially squeeze the gross returns to farmers in the arsenic affected areas An additional burden for farming in arsenic areas arises on account of the extra costs, particularly, in terms of fertilisers, owing to deterioration in soil fertility perceived by a huge majority of our sample households in the arsenic blocks From a policy viewpoint, steps need to be taken to ensure improved livelihoods for people in arsenic affected areas, particularly those dependent on agriculture Even though arsenic removal plants of different sizes and capacities, which can provide water at the cost of paise per litre developed by IIT-Madras have been developed, the capital cost of such plants is prohibitively high for private investors While these have been installed through government initiatives in some arsenic affected areas, they were for the purpose of supplying clean drinking water It seems unlikely that the Centre or State government would be willing to invest at the scale for supplying irrigation water when it is still reeling under the pressure of ensuring safe drinking water to affected populations Similarly, deep tube wells which are free from arsenic or surface water where available can be utilised for providing water for drinking and domestic uses, but are unlikely to be adequate, reliable sources of irrigation water For agricultural sustainability, it is the farmers’ own adaptation measures which hold the pathway for tackling the adverse impacts of arsenic contamination These include measures such as usage of arsenic-resistant crop varieties, shift to cultivation of less arsenic sensitive and less water 24   intensive crops, deficit irrigation, fertilization and bio-remediation Each of these measures curtails the various impacts of arsenic affected irrigation water on agriculture ie uptake by crops, accumulation in soil and reduction in crop yields, to varying degrees Depending on the specific impact of interest, one or the other measure may be applied Appendix Table 1A: Two Sample Wilcoxon Rank-Sum Test (Mann-Whitney test) for Profit Per Hectare in Kharif Dummy_arsenic Observations 107 Rank sum 15395.5 Expected 11556 108 7824.5 11664 Combined 215 23220 23220 Unadjusted variance 208008.00 Adjustment for ties -0.50 Adjusted variance 208007.50 H0: median (dummy_arsenic=0) = median (dummy_arsenic=1) z = 8.419 Prob > |z| = 0.0000 Table 2A: Two Sample Wilcoxon Rank-Sum Test (Mann-Whitney test) for Profit Per Hectare in Rabi Dummy_arsenic Observations 113 Rank sum 16446 Expected 13051.5 117 10119 13513.5 Combined 230 26565 26565 Unadjusted variance 254504.25 Adjustment for ties -0.75 Adjusted variance 254503.50 H0: median (dummy_arsenic=0) = median (dummy_arsenic=1) z = 6.729 Prob > |z| = 0.0000 Table 3A: Two Sample Wilcoxon Rank-Sum Test (Mann-Whitney test) for Profit Per Hectare in Summer Dummy_arsenic Observations Rank sum Expected 85 10221.5 8202.5 107 8306.5 10325.5 Combined 192 18528 18528 Unadjusted variance 146277.92 Adjustment for ties -6.82 Adjusted variance 146271.10 H0: median (dummy_arsenic=0) = median (dummy_arsenic=1) z = 5.279 Prob > |z| = 0.0000 Test results confirm that rank sums of profit per hectare in Kharif, Rabi and Summer for nonarsenic blocks are statistically significantly higher than the arsenic blocks 25   Table 4A: Two Sample Wilcoxon Rank-Sum Test (Mann-Whitney test) for Paddy Yield in Kharif Dummy_arsenic Observations Rank sum Expected 108 15208 11718 108 8228 11718 Combined 216 23436 23436 Unadjusted variance 210924.00 Adjustment for ties -4149.96 Adjusted variance 206774.04 H0: median (dummy_arsenic=0) = median (dummy_arsenic=1) z = 7.675 Prob > |z| = 0.0000 Test results confirm that rank sums of paddy yield (kg per hectare) for non-arsenic blocks are statistically significantly higher than the arsenic blocks Table 5A: Two Sample Wilcoxon Rank-Sum test (Mann-Whitney test) for Paddy Output Price in Kharif Dummy_arsenic Observations Rank sum Expected 108 13852 11664 107 9368 11556 Combined 215 23220 23220 Unadjusted variance 208008.00 Adjustment for ties -12819.73 Adjusted variance 195188.27 H0: median (dummy_arsenic=0) = median (dummy_arsenic=1) z = 4.952 Prob > |z| = 0.0000 Test results confirm that rank sums of the output price for paddy (Rs per kg) for non-arsenic blocks are somewhat higher than the arsenic blocks; results being statistically significant at percent level Table 6A: Two Sample Wilcoxon Rank-Sum test (Mann-Whitney test) for Fertiliser Per Hectare in Kharif Dummy_arsenic Observations Expected 107 8542.5 11556 108 14677.5 11664 Combined 215 23220 23220 Unadjusted variance 208008.00 Adjustment for ties -232.83 Adjusted variance 207775.17 H0: median (dummy_arsenic=0) = median (dummy_arsenic=1) z = -6.611 Prob > |z| = 0.0000 26   Rank sum Table 7A: Two Sample Wilcoxon Rank-Sum Test (Mann-Whitney test) for Fertiliser Per Hectare in Rabi Dummy_arsenic Observations Rank sum Expected 102 7998.5 10710 107 13946.5 11235 Combined 209 21945 21945 Unadjusted variance 190995.00 Adjustment for ties -4770.98 Adjusted variance 186224.02 H0: median (dummy_arsenic=0) = median (dummy_arsenic=1) z = -6.283 Prob > |z| = 0.0000 Table 8A: Two Sample Wilcoxon Rank-Sum Test (Mann-Whitney test) for Fertiliser Per Hectare in Summer Dummy_arsenic Observations Rank sum Expected 85 3962 8160 106 14374 10176 Combined 191 18336 18336 Unadjusted variance 144160.00 Adjustment for ties -6484.89 Adjusted variance 137675.11 H0: median (dummy_arsenic=0) = median (dummy_arsenic=1) z = -11.314 Prob > |z| = 0.0000 Test results confirm that rank sums of fertiliser cost per hectare in Kharif, Rabi and summer for arsenic blocks are statistically significantly higher than the non-arsenic blocks Table 9A: Cost concepts as per CACP methodology Cost notation Definition Cost A1 All actual expenses in cash and kind incurred in production by owner/operator Cost A2 Cost A1+ rent paid for leased-in-land Cost B1 Cost A1 + interest on value of owned capital assets (excluding land) Cost B2 Cost B1 + rental value of owned land (net of land revenue) and rentpaid for leased-in-land Cost C1 Cost B1+ imputed value of family labour Cost C2 Cost B2 +imputed value of family labour 27   References Abedin, Md Joinal, J Cotter-Howells and Andy A Meharg (2002) Arsenic Uptake and Accumulation in Rice (Oryza sativa L.) Irrigated with Contaminated Water Plant and Soil, 240: 311-19 Asadullah, M Niaz and Nazmul Chaudhury (2011) Poisoning the Mind: Arsenic Contamination of Drinking Water Wells and Children’s Educational Achievement in Rural Bangladesh Discussion Paper No 5716 Institute for the Study of Labor, Bonn, Germany Azad, M A K, A H M F K Mondal, M I Hossain and M Moniruzzaman (2012) Effect of Arsenic Amended Irrigation Water on Growth and Yield of BR-11 (Orgyza sativa L.) Grown in Open Field Gangetic Soil Condition in Rajshahi J Environ Sci & Natural Resources, (1): 55-59 Brammer, Hugh and Peter Ravenscroft (2009) Arsenic in Groundwater: A Threat to Sustainable Agriculture in South and South-East Asia, Environment International, 35: 647-54 Central Ground Water Board (2013) Report on the Dynamic Ground Water Resources of West Bengal Government of India Chaturvedi, I (2006) Effects of Arsenic Concentrations and Forms on Growth and Arsenic Uptake and Accumulation by Indian Mustard (Brassica juncea L) Genotypes Journal of Central European Agriculture, 7: 31-40 Das, Abhijit, Joyashree Roy and Sayantan Chakrabarti (2016) Socio-Economic Analysis of Arsenic Contamination of Groundwater in West Bengal, India Studies in Business and Economics, Springer Das, D K, P Sur and K Das (2008) Mobilization of Arsenic in Soils and in Rice (Oryza sativa L) Plant Affected by Organic Matter and Zinc Application in Irrigation Water Contaminated with Arsenic Plant Soil Environment, 54 (1): 30-37 Das, Indranil, Koushik Ghosh, D K Das and S K Sanyal (2013) Assessment of Arsenic Toxicity in Rice Plants in areas of West Bengal Chemical Speciation and Bioavailability, Taylor and Francis, 25 (3): 201-8 Dittmar J, A Voegelin, L C Roberts, S J Hug, G C Saha, M A Ali, B M Badruzzaman and R Kretzschmar (2007) Spatial Distribution and Temporal Variability of Arsenic in Irrigated Rice Fields in Bangladesh Paddy Soil Environ Sci Technol, 41: 5967-72 Duxbury, J M, A B Mayer, J G Lauren, N Hassan (2003) Food chain aspects of arsenic contamination in Bangladesh: Effects on Quality and Productivity of Rice J Environ Sci Health, Part A: Toxic/Hazard Subst Environ Eng., 38: 61-69 Duxbury, J M, G Panaullah and S K Oshima (2007) Remediation of Arsenic for Agriculture Sustainability, Food Security and Health in Bangladesh Working Paper Rome: Food and Agriculture Organization (FAO) Government of India (n.a.) Bringing Green Revolution to Eastern India, Ministry of Agriculture http://bgrei-rkvy.nic.in/, accessed on 10/05/2016 ————— (2011) Census of India 2010-11 Office of Registrar General and Census Commissioner, Ministry of Home Affairs, Government of India http://censusindia.gov.in/, accessed on 30/04/2018 28   ————— (2015) Report of the Inter-Ministerial Group for ‘Arsenic Mitigation’ Ministry of Water Resources, River Development and Ganga Rejuvenation, Government of India, July 2015 ————— (2015) Green Revolution, Press Information Bureau, Ministry of Agriculture and Farmers’ Welfare http://pib.nic.in/newsite/PrintRelease.aspx?relid=115699, accessed on 10/05/2016 Government of West Bengal (2005) Report of the Task Force on Formulating Action Plan for Removal of Arsenic Contamination in West Bengal Ghosh, N C and R D Singh (2009) Groundwater Arsenic Contamination in India: Vulnerability and Scope for Remedy Roorkee: National Institute of Hydrology Gibbons, Jean Dickinson (1993) Nonparametric Statistics: An Introduction Series: Quantitative Applications in the Social Sciences, a Sage University Paper, Sage Publications Haque, Md Nazmul, Md Hazrat Ali, TuhinSuvra Roy, Sheikh Muhammad Masum and Imtiaz Faruk Chowdhury (2015) Yield Reduction and Arsenic Accumulation in Potatoes (Solanum tuberosum L.) in an Arsenic Contaminated Soil AgronomiaColombiana, 33 (3): 315-21 Heikens, Alex (2006) Arsenic Contamination of Irrigation Water, Soil and Crops in Bangladesh: Risk Implications for Sustainable Agriculture and Food Safety in Asia Food and Agriculture Organization of the United Nations Hossain, M F (2005) Arsenic Contamination in Bangladesh: An Overview Agriculture, Ecosystems and Environment, 113 (1–4): 1-16 Hossain M B, Jahiruddin M, Panaullah G M, Loeppert R H, Islam M R, Duxbury J M (2008) Spatial Variability of Arsenic Concentration in Soils and Plants, and Its Relationship with Iron, Manganese and Phosphorus, Environmental Pollution, 156: 739-44 Indu, Rajnarayan, SunderrajanKrishnan and Tushaar Shah (2007) Impacts of Groundwater Contamination with Fluoride and Arsenic: Affliction Severity, Medical Cost and Wage Loss in Some Villages of India International Journal of Rural Management, (1): 69-93 Kamal, A S M and P Parkpian (2003) Arsenic Severity in Bangladesh and Suitability of Arsenic Removal Techniques Water Resources Journal, 214: 1-15 Khan, A, M Rafiqul Islam, G M Panaullah, J M, Duxbury, Jahiruddin and R H Leoppert (2010) Accumulation of Arsenic in Soil and Rice under Wetland Condition in Bangladesh Plant and Soil, 333 (1-2): 263-74 Liu, Q J, C M Zheng, C X Hu, Q L Tan, X C Sun and J J Su (2012) Effects of High Concentrations of Soil Arsenic on the Growth of Winter Wheat (Triticum aestivum L) and Rape (Brassica napus) Plant, Soil and Environment, 58: 22-27 Mandal, B K, T R Chowdhury, G Swinterta, D P Mukherjee, C R Chanda, K C Saha, D Chakraborti (1998) Impact of Safe Water for Drinking and Cooking on Five Arsenic-Affected Families for Years in West Bengal, India The Science of Total Environment, 218: 185-201 Mazumder, D N Guha, Aloke Ghosh, Kunal Kanti Majumdar, Nilima Ghosh, Chandan Saha and R N Guha Mazumder (2010) Arsenic Contamination of Ground Water and its Health Impact on Population of District of Nadia, West Bengal, India Indian Journal of Community Medicine, 35 (2): 331-38 29   Meharg, A A and M Rahman (2003) Arsenic Contamination of Bangladesh Paddy Field Soils: Implications for Rice Contribution to Arsenic Consumption Environmental Science and Technology, 37: 229-34 Mukherji, Aditi, Tushaar Shah and Parthasarathi Banerjee (2012) Kickstarting a Second Green Revolution in Bengal Economic and Political Weekly, 47 (18): 27-30 Nickson, R T, J M McArthur, P Ravenscroft, W G Burgess and K M Ahmed (2000) Mechanism of Arsenic Release to Groundwater, Bangladesh and West Bengal Applied Geochemistry, 15: 403-13 Norra, S, Z ABerner, P Agarwala, F Wagner, D Chandarsekharam and D Stuben (2005) Arsenic Intake Via Water and Food by a Population Living in an Arsenic Affected Area of Bangladesh Science of the Total Environment, 381: 68-76 Panaullah, G M, T Alam, M B Hossain, R H Loeppert, J G Lauren, C A Meisner, Z U Ahmed and J M Duxbury (2009) Arsenic Toxicity to Rice (Oryza sativa L.) in Bangladesh Plant and Soil, 317: 31-39 Pal, A, U K Chowdhury, D Mondal, B Das, B Nayak, A Ghosh, S Maity and D Chakraborti (2009) Arsenic Burden from Cooked Rice in the Populations of Arsenic Affected and Nonaffected Areas and Kolkata City in West-Bengal, India Environmental Science & Technology, 43 (9): 3349-55 Pigna, M, V Cozzolino, A Violante and A A Meharg (2009) Influence of Phosphate on the Arsenic Uptake by Wheat (Triticum durum L.) Irrigated with Arsenic Solutions at Three Different Concentrations Water Air & Soil Pollution, 197: 371-80 Pitt, Mark M, Mark R Rosenzweig and Nazmul Hassan (2012) Identifying the Hidden Costs of a Public Health Success: Arsenic Well Water Contamination and Productivity in Bangladesh PSTC Working Paper Series 2012-02, July 2012, Population Studies and Training Center, Brown University Public Health Engineering Department, Government of West Bengal (2014) http://maps.wbphed.gov.in/arsenic/index.html, accessed on 18/12/2018 Roychowdhury, Tarit, Hiroshi Tokunaga, Tadashi Uchino and Masanori Ando (2005) Effect of ArsenicContaminated Irrigation Water on Agricultural Land Soil and Plants in West Bengal, India Chemophere, 58: 799-810 Sarkar, S, B Basu, C K Kundu and P K Patra (2012) Deficit Irrigation: An Option to Mitigate Arsenic Load of Rice Grain in West Bengal, India Agriculture, Ecosystems & Environment, 146 (1): 147-52 Shrivastava, Anamika, Devanta Ghosh, Ayusman Dash and Suatapa Bose (2015) Arsenic Contamination in Soil and Sediment in India: Sources, Effects and Remediation Current Pollution Reports, Springer, 1: 35-46 Sikdar, P K and S Chakraborty (2008) Genesis of Arsenic in Groundwater of North Bengal Plain using PCA: A Case Study of English Bazar Block, Malda District, West Bengal, India Hydrological Process, 22: 1796-1809 Smedley, P L and D G Kinniburgh (2002) A Review of the Source, Behaviour and Distribution of Arsenic in Natural Waters, Applied Geochemistry, 17: 517-68 30   Williams, P N, M R Islam, E E Adomako, A Raab, S A Hossain, Y G Zhu, J Feldman and A A Meharg (2006) Increase in Rice Grain Arsenic for Regions of Bangladesh Irrigating Paddies with Elevated Arsenic in Ground Waters Environmental Science and Technology, 40: 4903-08 Xu, X Y, S P McGrath, A A Meharg and F J Zhao (2008) Growing Rice Aerobically Markedly Decreases Arsenic Accumulation Environmental Science & Technology, 42 (15): 5574-79 Zhong L, C Hu, Q Tan, J Liu and X Sun (2011) Effects of Sulfur Application on Sulfur and Arsenic Absorption by Rapeseed in Arsenic-Contaminated Soil Plant, Soil and Environment, 57: 42934 31   ... assessment of potential threats of contamination in unexploited areas or of worsening arsenic levels in already exploited regions (UNICEF, 2008) ™ Agronomic impacts of irrigating with arsenic contaminated. .. Socio-Economic Analysis of Arsenic Contamination of Groundwater in West Bengal, India Studies in Business and Economics, Springer Das, D K, P Sur and K Das (2008) Mobilization of Arsenic in Soils... (i) mineral dissolution, (ii) desorption of arsenic under alkaline and oxidizing conditions, (iii) desorption and dissolution of arsenic under reducing conditions, (iv) reduction of oxide mineral