Environ Sci Pollut Res DOI 10.1007/s11356-014-2531-6
RESEARCH ARTICLE
Persistence, variance and toxic levels of organochlorine pesticides in fluvial sediments and the role of black carbon in their retention Musarrat Parween & AL. Ramanathan & P. S. Khillare & N. J. Raju
Received: 1 August 2013 / Accepted: 7 January 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract The present study assesses the persistence and variation of organochlorine pesticides (OCPs) and their regulation by total organic carbon (TOC) and black carbon (BC) in freshwater sediment. Sediment samples from the Yamuna River, a major tributary of the Ganges (one of the most populated and intensively used rivers in Asia), had high levels of Σ20OCPs (21.41 to 139.95 ng g−1). β-Hexachlorocyclohexane (β-HCH) was the most predominant component. ΣHCH and Σdichlorodiphenyl-trichloroethane (DDT) constituted ~86 % of Σ20OCPs. Isomer ratios indicated fresh usage of lindane, DDT and technical-grade HCH. Toxicological comparison with freshwater sediment quality guidelines showed γ-HCH and DDT at high levels of concern. β-HCH, α-HCH, endrin, heptachlor epoxide, dichloro-diphenyl-dichloroethane (DDD), dichlorodiphenyl-dichloroethylene and chlordane were above some of the guideline levels. TOC and BC had mean concentrations of 1.37±0.51 % and 0.46±0.23 mg g−1, respectively. BC constituted 1.25 to 10.56 % of TOC. We observed low to moderate correlations of BC with isomers of HCH, p,p′-DDT and methoxychlor while of TOC with Σ20OCPs, γ-HCH, endosulfan sulfate and methoxychlor. Principal component analysis enabled correlating and clustering of various OCPs, BC and TOC. OCP distribution was related with pH, electrical conductivity, soil moisture and finer fractions of sediment. OCPs with similarity in properties that determine their interactions with carbonaceous components of sediment clustered together. A number of factors may, thus, be involved in the regulation of interactive forces Responsible editor: Leif Kronberg Electronic supplementary material The online version of this article (doi:10.1007/s11356-014-2531-6) contains supplementary material, which is available to authorized users. M. Parween : A. Ramanathan (*) : P. S. Khillare : N. J. Raju School of Environmental Sciences, Jawaharlal Nehru University, New Delhi 110067, India e-mail:
[email protected]
between BC and OCPs. BC in this study may be more important than TOC in the retention of some OCPs into fluvial sediments, thereby reducing their bioavailability. The finding is probably the first of its kind to report and emphasises the role of BC in the persistence of OCPs in fluvial sediments. Keywords Organochlorine pesticide . Black carbon . Ecotoxicology . Principal component analysis . Yamuna River . Sediment . Persistent . Correlation
Introduction Organochlorine pesticides (OCPs) and black carbon (BC) are among the most attractive global contaminants. As per the Stockholm Convention, persistent organic pollutants (POPs) included nine OCPs which were produced internationally and used extensively on agricultural crops or for vector control. These compounds are highly toxic and potentially carcinogenic (USEPA (US Environmental Protection Agency) 1980). They possess a cumulative capacity and slow degradation rate, and their residues are present in the soil, water and sediment even after several years of their application. Despite bans imposed on most of them several years ago, numerous investigations have reported the continued and ubiquitous presence of OCPs in the global atmosphere (Hung and Thiemann 2002). These biomagnify and accumulate in living organisms, including animal tissue, human blood, adipose tissue and breast milk (Malik et al. 2009). BC consists of an entire continuum ranging from partly charred plant material through charcoal to graphite and soot particles recondensed from the gas phase with no unambiguously acceptable boundaries (Seiler and Crutzen 1980). Incomplete combustion of fossil fuels or biomass leads to the formation of carbon-rich (>60 %) aromatic residues (char) and condensates (soot) (Akhter et al.
Environ Sci Pollut Res
1985; Goldberg 1985; Novakov 1984). They remain ubiquitous in the atmosphere, sediment, soil and water and influence a wide range of biogeochemical processes (Schmidt and Noack 2000; Watson et al. 2005). More than 80 % of BC produced ends up in soil/sediment, where it can reside for hundreds to thousands of years without much biological and chemical breakdown (Hammes et al. 2007; Forbes et al. 2006; Preston and Schmidt 2006). Studying the accumulation of soot BC is specifically attractive due to the properties of radiative forcing and high recalcitrance in soil and sediment (Ramanathan and Carmichael 2008; Goldberg 1985). Its accumulation or burial in soil/sediment therefore makes it attractive for attaining a reduction in both global warming and bioavailability of organic pollutants to living organisms (Gustafsson and Gschwend 1997). The deposition of sediments to which OCPs are sorbed leads to substantial increases in the residence time of these compounds in aquatic ecosystems, thereby facilitating bioaccumulation. It is well known that OCPs bind mainly with organic matter present in the sediments, owing to their hydrophobicity. Total organic carbon (TOC) has been regarded as a key factor influencing concentrations of OCPs in sediments (Malik et al. 2009). Recent studies have demonstrated that BC is capable of adsorbing several types of POPs (including dioxins, polycyclic aromatic hydrocarbons, polychlorinated biphenyls and various pesticides) much more strongly than bulk organic carbon (Buckley et al. 2004; Gustafsson et al. 1997; Hung et al. 2010; Jonker and Smedes 2000; Jonker and Koelmans 2002). However, such relationships in context to OCPs are not yet well established. A few related studies were conducted earlier (Hung et al. 2007; Hung et al. 2010; Tan et al. 2009), but no relationship could be observed. Hung et al. (2010) observed a poor correlation of OCPs (ranging from 0.46 to 4.02 nanograms per gram (ng g−1)) with both particulate organic matter and BC in fluvial sediments of the Danshui River (at 12 stations). They suggested that pesticides were not being primarily controlled by sediment organic carbon content, probably due to variability in the sources of the two and hypothesised that short residence time of organic carbon and BC might have led to insufficient adsorption of OCPs. Tan et al. (2009) too observed poor correlation of OCPs with TOC and BC in estuarine sediments of the Daliao River. They too hypothesised the non-attainment of equilibrium to be the reason behind such poor correlations. Hung et al. (2007) found the correlation of OCPs (not detectable to 23 ng g−1) with TOC and BC to be significant in coastal sediments of the Danshui River. However, the same was not found to be significant in the case of fluvial sediments (number of samples, n=6). This led to a speculation of BC being less important for OCP distribution in fluvial sediments. The authors attributed this to the diversity in sources of BC and TOC, with BC being mostly atmospheric and emission products of fossil fuels and industrial combustion. The study by Tan et al. (2009) was conducted on an
estuary which may have considerably different biogeochemistry as compared to a fluvial ecosystem. Therefore, the study is not comparable to the present study. The other two studies were limited to a substantially low number of sampling locations, which may not be sufficient enough for representing actual levels and patterns of contaminants. Moreover, the levels of OCPs observed in all the three studies were very low as compared to that of the current study, which might have led to poor correlations. The relationships may be complex and inexplicit, or correlations may exist within a different set of conditions, such as higher levels of contaminants (Tan et al. 2009). The study of heavily polluted sediments with regard to such relationships has not yet been conducted. The subject needs further and extensive investigations employing strong statistical tools for revealing implicit associations. The present study attempt is therefore dedicated with extensive study of OCPs and carbonaceous materials and their variations along with the employment of strong tools, like principal component analysis (PCA) for investigating their relationship. The study is therefore, unique of its kind, attempted for a tropical river system. Riverine sediments of Yamuna, a major tributary of one of Asia’s largest rivers (the Ganges), were analysed for the purpose. The Ganges, which rises in the western Himalaya, flows east to the Bay of Bengal, through the immeasurable plains of northern India into Bangladesh and supports one of world’s densest populations. It is, thus, a good representative of a polluted tropical river. Despite being hardly 2 % of the entire stretch, Delhi’s contribution of pollution load into the Yamuna River is around 80 % (Singh et al. 2007). The river receives a fair amount of pesticides as run-off from the agricultural fields around it. Under the confluence of numerous industrial activities, domestic and agricultural burning, thermal power plants and high traffic loads, sediment in this region is also a potential sink for carbonaceous components. Moreover, lack of data pertaining to BC levels in Asian freshwater sediment promoted the selection of this region as the study area. Asia is the largest source for BC emission with China being the dominant emitter of BC, accounting for 61 % of all emissions in 2006, followed by India (12 %) and Indonesia (6 %) (Kroeger 2010). China and India together account for 25 to 35 % of global BC emission. BC resides in the atmosphere for several days to few weeks only, as compared to CO2 which has a lifetime of over 100 years (Ramanathan and Carmichael 2008). It provides a repository of carbon in sediment. It is, thus, highly pertinent to undertake a study of this kind.
Methods and materials Study area Delhi (28°37′N, 77°12′E, population ~20 million) is spread over an area of 1,483 km2 and has an average elevation of
Environ Sci Pollut Res
~216 m above sea level. The region is characterised by intense hot summers (mean temperatures of 32–34 °C from May to June) and cold winters (mean temperatures of 12–14 °C in December–January). The mean annual rainfall is 714 mm, of which around 80 % is received during the monsoon season (July to October). The predominant wind direction is from the north and northwest, except during the monsoon season that is characterised by easterly or southeasterly winds (Srinivas 1999). The study area lies in the semi-arid region of northern India and is characterised by loosely bound mineral dusts. Dust storms, known as Aandhi, are common during summer. Yamuna is polluted with domestic waste, silt and industrial waste. The water quality is critical and severely degraded within the stretch of the Delhi region (Singh and Ghosh 2003a). Tapping of water at Wazirabad and Okhla barrages for drinking water supply to Delhi turns the water downstream merely to untreated or partially treated domestic and industrial wastewater from several drains during the dry season (Kumar et al. 2012). Eighteen drains, including Najafgarh and Shahdara Drains (the largest drains), fall into the river.
Sampling locations Selection of locations (Fig. 1) for the collection of samples was based on the drainage system of the Yamuna River in the Delhi region. Most of them are in receipt of effluent from major drains and canals and thus predicted to be the most polluted locations in the study area. In view of the water quality status, the river was distinguished as sections in this study and two locations were selected for sample collection within each section of the river. 1 and 2 which are rural agricultural locations in the village Jagatpur, which are relatively remote to urban Delhi and unpolluted. These were considered as background locations and within section I. The water quality at these locations seemed to be fairly better as compared to those lying downstream to them. 3 and 4 (section II) are urban industrial locations. Locations 5 and 6 in section III are urban industrial locations which receive discharges from the Najafgarh Drain and the supplementary drain. Locations 7 and 8 of section IV are urban agricultural locations. They receive discharges from the Grand Trunk Road Drain and Eastern Yamuna Canal. Locations 9 and 10 of section V receive discharges from 10 major drains joining the river.
Fig. 1 Map of Delhi showing the sampling locations and major sources of pollution (most polluted drains, sewage treatment plants, treated effluent discharge points, industrial areas, power plants) in the Yamuna River
Environ Sci Pollut Res
These are high-traffic sites lying downwind to the Rajghat power plant. Locations 11 and 12 (section VI) are urban industrial and residential locations downwind to the Pragati power plant. Locations 13 and 14 within section VII receive discharges from the Shahdara Outfall Drain and agricultural runoff from Uttar Pradesh (UP). These are urban residential locations amidst high-traffic zone in Delhi. Apart from the major drains mentioned above, several other drains, effluents and agricultural discharges join the river upstream to each location selected on either side of the river for the present study. Sample collection Eighty-four samples of surface sediment (~10 cm) from bank and flood plain including either side of the river were collected at 14 locations with the help of a stainless steel auger, transferred into polythene bags, sealed and transported to laboratory. Each sample was air-dried followed by removal of debris and pebbles with the help of a 2-mm sieve. The samples were mixed thoroughly and segregated by the quartering and coning method (Ingram 1971) to obtain representative samples at each location. Samples were collected once in each season (post-monsoon/winter, pre-monsoon/summer and monsoon) during the months of February, June and September in 2010, respectively. A total of 28 samples were collected in each season. All samples were reduced to fine powders lesser than 200 μm before extraction for OCP and carbon content. Extraction and analysis The sediment samples were extracted by the Soxhlet extraction method (Kumar et al. 2011, 2012; USEPA 1995) for 24 h. Preheated (450 °C) and purified sodium sulfate (Merck, Germany) was used for the removal of moisture from the extracted samples. Activated copper was added for desulfurisation. Clean-up of extracts by standard column chromatography was carried out with silica gel (Silica gel 60, particle size 0.0630– 0.200 mm, 70–230 mesh ASTM purchased from Merck KgaA, Darmstadt, Germany) as per Method 3630C of USEPA (Pandey et al. 2011). The samples were finally concentrated and transferred to hexane with the help of a rotary evaporator (Buchi, Switzerland) before analysis of OCPs with GC-ECD (Shimadzu 2010, Japan). Samples were run in splitless mode using an Rtx®-5MS column (30 m×0.25 mm×0.25 μm). The column temperature was initially maintained at 190 °C for 2 min and then ramped to 230 °C (1.5 °C/min) and held for 1 min and further ramped to 280 °C (20 °C/min) and held for 5 min. A standard mixture of OCPs containing 22 compounds specified in EPA method 8081B (Sigma-Aldrich, USA) was used for the quantification of OCPs through the external standard quantification method. The instrument was calibrated with standards to obtain regression coefficients R2 >0.99 before analysis. Confirmation for the identification of
compounds was done with the help of GC/MS (QP2010 Plus; Shimadzu, Japan). Samples for TOC analysis were prepared as described by Hu et al. (2009), while BC was determined by the chemothermal oxidation method at 375 °C (Gelinas et al. 2001), suitable for detecting soot BC. BC and TOC were analysed on a CHNSO analyser (Euro Elemental Analyzer, EuroVector EA3000) by combusting the samples at 1,050 °C. A PTFE packed GC column (2 m×8 mm) was used with He as a carrier gas at a flow rate of 100 ml/min at a pressure of 89 kPa. Purge flow was 80 ml/min. Four-point calibrations (R2 >0.98) using synthetic mixtures for soil #2 and soil #4 (EuroVector, E11036-A and E11038-A) were done before analysis of samples. Slurries were prepared by mixing sediment with deionised water in the ratio 1:2.5 (solid/liquid) for the determination of pH and allowed to stand for 30 min before electrical conductivity (EC) determination. Moisture and carbonate contents were determined by the LOI method (Heiri et al. 2001). Size fraction of sediment was analysed using a laser particle size analyser (Microtrac S3500, USA). The samples were preprocessed before analysis according to the method described by Konert and Vandenberghe (1997). Quality assurance With each set of 15 samples, a procedural blank and a spiked matrix sample with known amounts of standards were run to check for contamination, peak identification and quantification. The limit of detection for all OCPs was 0.01 ng g−1. Recovery percentages ranged between 84 and 95 %. Procedural blanks were included with each sample batch, and analysed values obtained in the blanks were subtracted from the values found in the samples. The relative standard deviations of the replicate samples were α. In cultivated sandy soil, they dissipate in the order α>γ>δ>β (Suzuki et al. 1975). Moreover α-HCH and γ-HCH are slowly transformed to β-HCH with the age of the environmental sample (Hu et al. 2009). α-HCH also has a high log octanolwater partition coefficient (Kow) and Henry’s law coefficient, so it partitions more readily into water than solid particles and predominate in air. The variations in the distribution of β- and γ-HCH were significant within different sections and seasons (ANOVA, p
