A LIMITED ROLE OF ARSENIC REDUCTION FOR ITS MOBILIZATION IN DEEP BANGLADESH AQUIFER

The role of microbial arsenate and iron reduction in the mobilization of As can be difficult to distinguish. In a setting of great practical importance, a set of incubation experiments was conducted over a period of 3 months using deep Pleistocene (orange) aquifer sands from Bangladesh that are typically associated with low groundwater As concentrations. Aliquots of 5 g of orange sands collected in 2001 and resuspended in 2005 in 10 ml of anaerobic artificial groundwater (Agw) were inoculated with wild-type Shewanella sp. ANA-3 and a mutant Shewanella sp. ANA-3 (AMR1). The wild-type ANA-3 can use Fe(III), Mn(IV), or As(V) as terminal electron acceptors. The AMR1 strain is deficient in As(V) reduction due to deletions in genes for As-respiration (arrA) and As-detoxification (arsC). The Pleistocene sands initially contained little Fe(II) in the acid-leachable Fe fraction (Fe(II)/Fe=0.05) and about 0.14 mg/kg phosphate-extractable As, all as arsenate. Aqueous As speciation in solution and in the P-extract was assayed by a voltammetric method that selectively detects arsenite with a detection limit of ~ 0.2 µg/L. Anaerobic incubations were conducted with and without additions of 0.02 M lactate or kanamycin. Kanamycin was used in an attempt to suppress the activity of indigenous microbes (both Shewanella strains contained plasmids with kanamycin resistance genes). Surprisingly, wild-type and mutant strains of Shewanella mobilized As and Fe to a similar extent in the presence of lactate. The equivalent of 0.035 and 0.031 mg/kg sediment of As was mobilized on day 23 of the incubations with the wild and mutant strain, respectively. However, 97% of the aqueous As was arsenite in the wild incubation, whereas only 37% of the aqueous As was arsenite in the mutant incubation. The solid phase in both inoculations contained the equivalent of <0.004 mg/kg As(III) on day 23. The controls with or without lactate and without either strain of Shewanella all released less than the equivalent of 0.003 mg/kg As from the solid phase. We conclude that reduction of As may not be required to release As from oxidized Pleistocene sands of Bangladesh. Microbial Fe reduction may therefore be the critical process to evaluate to assess the vulnerability of Pleistocene aquifers of Bangladesh.