Golam M. Panaullah
Department of Crop and Soil Sciences, Cornell University, 904 Bradfield Hall, Ithaca, NY 14853, USA
Tariqul Alam
Bangladesh Institute of Nuclear Agriculture, Mymensingh 2000, Bangladesh
M. Baktear Hossain
Bangladesh Institute of Nuclear Agriculture, Mymensingh 2000, Bangladesh
Richard H. Loeppert
Department of Soil and Crop Sciences, Texas A&M University, College Station, TX 77843, USA
Julie G. Lauren
Department of Crop and Soil Sciences, Cornell University, 904 Bradfield Hall, Ithaca, NY 14853, USA
Craig A. Meisner
Department of Crop and Soil Sciences, Cornell University, 904 Bradfield Hall, Ithaca, NY 14853, USA
Zia U. Ahmed
Department of Crop and Soil Sciences, Cornell University, 904 Bradfield Hall, Ithaca, NY 14853, USA
John M. Duxbury
Department of Crop and Soil Sciences, Cornell University, 904 Bradfield Hall, Ithaca, NY 14853, USA
Arsenic phytotoxicity, Food security, Irrigation, Soil contamination, Yield reduction
Village of Paranpur of the Sadar Upazilla in Faridpur
Crop-Soil-Water Management
Experimental site and design The study site was a STW command area of about 8 ha in farmer fields located at 23°36.246′ N and 89°45.738′E in the village of Paranpur of the Sadar Upazilla in Faridpur, a highly arsenic-affected district of the Gangetic Floodplain in central Bangladesh. The soil was a fine-textured (silty clay) Aquept (pH 7.5–7.9, organic C 13–18 g kg−1). The predominant cropping pattern at the study site was rain-fed monsoon rice (transplanted Aman) followed byirrigated winter rice (Boro). The STW (35 m deep) has been used to irrigate Boro rice since installation in 1990. The As concentration in the water measured at different times from 2005–2007 ranged between 0.10–0.13 mg L−1, but notemporal trend was observed. A single irrigationchannel runs laterally in both directions from the STW, and individual fields are irrigated by allowingwater to flow from field to field. Soil samples were collected from 56 geo-referenced points following a 30-m grid spacing across the command area, and a map of total soil As was generated by ordinary kriging of the soilAs point data. In 2006, six 10×10 m treatmentplots (T1-T6) were established across a soil-Asgradient from 12 to 68 mg kg−1. Eachtreatment plot was divided into 4 replicate sub-plotsto assess variability. Before setting up the experiment, composite soil samples (0–15 cm) were collectedfrom each sub-plot and analyzed for total As, total Fe, oxalate-extractable Fe and Mn, and available P,Zn and Cu. Similar samples were collected immedi-ately after harvest from each sub-plot for soil-Asdetermination only. In 2007, four new treatment plots close tolocations T1, T3, T4 and T6 were established across the soil-As gradient and divided into 4 replicate sub-plots to assess the variability. Composite soil samples (0–15 cm) were collectedfrom each sub-plot before planting and after harvestand analyzed for total As. Rice seedlings (variety BRRI dhan 29) were raisedin a non As-contaminated soil and 35 day-old seedlingswere transplanted (2 seedlings per hill) with a hill tohill spacing of 20×20 cm on January 25, 2006 and2007. Fertilizers were applied to all plots at elemental rates of 140, 25, 50, 20 and 4 kg ha−1 for N, P, K, S and Zn, respectively. Full doses of all fertilizers except N were applied 2 days before transplanting and mixedthoroughly with the soil. Nitrogen was applied inthree equal splits: one-third 2 days before transplanting, one-third 40 days after transplanting (DAT) and one-third 7–10 days before panicle initiation (60–70 DAT). Continuous flooding to a depth of ~ 8 cmusing the STW water was maintained from trans-planting until 80% grain maturity. The plots were thendrained to prepare for harvest. At crop maturity, twenty randomly selected hillsfrom each sub-plot were cut at ground level and usedfor agronomic measurements and chemical analyses. A5m2area from the center portion of each sub-plotwas harvested to determine grain yield. Rice wasthreshed by hand immediately after harvest, andweight and grain-moisture content were recorded.Grain yield was adjusted to 14% moisture. Rice strawand husked grain samples were analyzed for total As, Fe, Mn and P in 2006 but only total As in 2007. Throughout the 2006 cropping season, pore-water samples were collected from the root zone of rice at 10-cm depth in one sub-plot at each soil As level. Thewater extraction device consisted of a porous ceramiccup attached to a 1-cm diameter, 30-cm long PVCpipe. A vacuum inside the PVC tube-ceramic cup unitwas created with the use of a hand suction pump. Thepore-water sampler was kept in the field for 2–3h to collect about 50 mL of pore water. The Eh of the porewater was measured immediately after collectionu sing a portable pH meter equipped with a glass electrode and a Pt electrode. A 20-mL aliquot of the pore water was transferred immediately after collection to a polyethylene vial containing 2 mL of 2MHNO3. The pore-water samples were kept in the darkand analyzed for total As, As (III) and As (V) on thesame day and for Fe and Mn within 2 days of collection. The procedure used should be adequate forchemical species preservation (Bednar et al. 2002). Chemical analyses Soil samples were digested with HNO3-H2O2 in a block digester at 110°C for total As and Fe determination (Tang and Miller1991). Total As inthe soil digest was measured by flow-injectionhydride-generation flame-atomic-absorption spec-trometry (FI-HG-FAAS) using NaBH4 to generate AsH3.(Samanta et al.1999) and a Perkin-ElmerAAnalyst 200 instrument. Soil samples were extracted with pH 3.0, 0.2M ammonium-oxalate inthe dark for determination of poorly crystalline (amorphous) Fe and Mn oxides (Loeppert andInskeep1996), and Fe and Mn in the extracts were measured by FAAS. For available P (“Olsen P”), soilsamples were extracted with 0.5M NaHCO3, and P in the extract was determined colorimetrically by the phosphomolybdate ascorbic acid method (Olsen and Sommers 1982; Watanabe and Olsen 1965). Several samples analysed using ICP gave results similar tothe molybdenum blue procedure, indicating little, if any, inter ference from arsenate. Available Zn and Cu in soil were determined by DTPA (diethylenetriamine pentaacetic acid) extraction (Lindsay and Norvell 1978) and FAAS. The concentrations of As (III), As (V) and total As inpore water were determined by FI-HG-FAAS using1.5% NaBH4in 0.5% NaOH as the reductant and aphosphate eluent (0.2M NaH2PO4,pH 3.0) for determination of As (III) and 5M HCl eluent for determination of total inorganic As (Samanta et al. 1999). Standard solutions of As (III) and As (V) were preparedfrom certified As2O3 and As2O5, respectively. Rice straw and husked grain samples were digested with HNO3-H2O2 in the same manner as for soil samples. Total As in the digest was determined by FI-HG-FAAS, Fe and Mn by FAAS and P colorimetrically.
Plant Soil
Journal