A 65 km stretch of the San Joaquin River upstream of Mud Slough was delisted in 2010 . On the other hand, selenium loads in the 10 km stretch of Mud Slough through which selenium-rich drainage is being delivered from the San Luis Drain to the San Joaquin River have increased since the start of the GBP and usually exceed 5 µg/L . This rise in concentrations is likely to endanger local sensitive species including juvenile Chinook salmon and Steelhead trout . For the Chinook salmon in particular, seasonally elevated selenium concentrations in the stretch of the San Joaquin River between the confluence with Mud Slough and the confluence with the Merced River may prove problematic. Selenium concentrations have exceeded 10 µg/L in 6 of 24 months during the most recently published monitoring period, typically during the rainy season, i.e. between September and February . The seasonality of these peaks coincides with the emergence of the Chinook salmon’s sensitive juvenile live-stage, and the concentrations are in a range were increased mortality of up to 20% can be expected for juveniles . Thus selenium input through Mud Slough in this particular stretch of the San Joaquin River may represent an obstacle for ongoing efforts to restore salmon above the Merced, where they have been extirpated due to water diversions . Additionally, whereas selenium concentrations in most of the marshes have decreased, the 2 µg/L criterion was still exceeded in parts of the wetlands as recently as 2002, due to high flow input originating from the Delta-Mendota Canal , which was not captured by the bypass . As a result,procona London container the Grassland marshes listed in 1988 remain on California’s 303 list of impaired waters today .
First, caution needs to be exercised in preventing ecological damage in locations to which seleniferous runoff is being diverted. Second, circumvented locations may still receive selenium inputs due to other sources in the watershed. Consequently thorough monitoring of both circumvented and receiving water bodies is essential, especially during periods of high flow.In 2010, USGS scientists Theresa Presser and Samuel Luoma completed an ecosystem-scale selenium modeling effort in support of site-specific fish and wildlife criteria development for the San Francisco Bay and Delta . In brief, the model consists of three key components. First, the partitioning between dissolved selenium concentrations and the “particulate/planktonic” concentrations at the base of the food web is simulated using site-specific partitioning coefficients . Second, the local food web is resolved around target predator species of concern, including separate compartments for prey species or groups of species these predators feed on that differ significantly with respect to their selenium accumulation potential. Finally, the model comprises trophic transfer factors that correlate the concentration in the tissue of each species or trophic compartment with that of its diet. The model thus allows the calculation of tissue concentration estimates for all species that are part of the food web by simply multiplying dissolved concentrations by the Kd and then by the TTF for each lower connecting link in the food web. The model can also be used to convert tissue concentration limits for a target species to limits in solution. This approach is unique in that it captures critical trophic transfer steps relevant to the toxicological effects of selenium on individual species, while being generic enough to remain applicable to a wide array of aquatic ecosystems with diverse biogeochemical conditions. The main requirement is the availability of site-specific field data on the partitioning between dissolved concentrations and those at the local base of the food chain as well as information on the local food web, and on trophic transfer factors for key species present at a location.
The specific application of this approach to the San Francisco Bay and Delta lead to the capability of realistically translating tissue-based criteria for the protection of desired fish and bird species into dissolved or particulate concentration limits . In addition, the model was used to predict the ecosystem impacts of selenium in the Bay and Delta under various management scenarios . The California office of the EPA is now in the process of developing site-specific selenium criteria for the Bay and Delta based on the model results. In the future, the same modeling approach is to be used to develop site-specific criteria for other Californian ecosystem in which problems with selenium contamination occur . Such regulation would allow the protection of the most sensitive ecosystems without imposing an unnecessary regulatory burden in areas with less sensitive ecosystems and also the targeted protection of critical species. This approach would thus represent a landmark in the regulation of aquatic contaminants in the US and a significant improvement over regulation that traditionally has taken the form of state or nationwide criteria based on dissolved or acid soluble concentrations. Whereas the use of the model to determine site-specific dissolved selenium criteria or TMDLs is well warranted, its application to predict ecosystem impacts under changing environmental or management conditions may be limited to environments with relatively stable bio-geochemical conditions. The reason is that the model does not explicitly account for selenium speciation nor does it separate environmental compartments such as sediments and the water column. Transformation between chemical species and transfer between environmental compartments are far more dynamic and less linear than transfer through food webs and thus necessitate a dynamical model for adequate representation.
Such transfer could be of great importance in predicting selenium exposure under changing environmental conditions, especially for shallow stagnant water bodies with a high sediment-water interface to volume ratio, such as the Salton Sea. Fortunately, a dynamic model extension for water-sediment interactions could be integrated with Presser and Luoma’s approach, since mathematically this approach represents a simple multiplication of concentrations by partitioning and trophic transfer factors. Without such extensions, the models predictive power is limited to the geochemical steady-state and site-specific regulatory limits derived from it need to be coupled to ongoing monitoring and periodic revisions.The last three decades have seen significant progress with respect to the management and regulation of irrigation-induced selenium contamination in California. Among local remediation methods, sequential drainage reuse ending in well-designed evaporation facilities can be a viable option limited primarily by the scalability of the operation cost and the disposal of produced salts. Much can be learned from the integrated approach pursued as part of the Grassland Bypass Project . In particular, the project provides a blueprint of the framework necessary to establish and enforce load limits in an agricultural non-point context. As long as the means to track discharge quantities and concentrations are available, this approach can be translated to other agricultural sources of selenium or other pollutants . Given that a majority of selenium load reductions to date have been achieved by a reduction in drainage loads rather than selenium concentrations, there appear to be opportunities for additional reductions through management practices that enhance selenium retention in the source soils. Recent research suggests that unexplored options remain in this area, such as the management of soil structure to enhance microbial selenium reduction , or the addition of organic matter amendments to enhance reduction, retention, and volatilization . A better understanding of the factors controlling selenium speciation in soils would also help evaluate the long-term sustainability of drainage reuse schemes. The site-specific regulation currently under development for the San Francisco Bay and Delta represents an appropriate and timely update to federal selenium water quality criteria, which as discussed in the background section,cut flower transport bucket have proven inadequate in light of scientific findings over the last two decades. In this context, scientists and resource managers should think ahead about the needs that will arise as the approach is expanded to other sites of concern with respect to selenium contamination. California’s 303 list of impaired waters currently includes more than 60 water bodies polluted by selenium . By far the largest of these is the Salton Sea with an estimated 944 km2 affected . Like the Kesterson Reservoir in the 1980s, the Salton Sea has long been the receiving body of seleniferous irrigation drainage . It is also one of the most important bird habitats in the American Southwest, used by hundreds of thousands of waterfowl pertaining to resident and migratory bird species, including endangered ones like the brown pelican .
Whereas the Salton Sea is an obvious target for the expansion of the site-specific regulatory approach, matters may be complicated by the Sea’s uncertain management future . For example, management choices that expose sediment to oxic conditions may lead to the local release of reduced selenium accumulated in sediments, creating ecological hazard beyond that due to ongoing seleniferous irrigation-drainage inputs. In this case, the development of site-specific selenium criteria would need to be coupled to a detailed understanding not only of the trophic transfer processes in the local food web, but also the local bio-geochemical transformations in the shallow basin. Thus, the promise of the site-specific approach to the regulation of selenium as contaminant creates renewed urgency for the improvement of biogeochemical models of selenium cycling and the acquisition of field data at sites of concern.Within the physically complex matrix of a soil, microbial reduction is dictated by the local chemical conditions and thus subject to the soil’s physical, chemical and biological heterogeneity. Aggregates, which are mm to cm sized structural units of clay, silt and sand particles bound by roots, hyphae, and organic matter , represent the smallest systems in which the spatial coupling of transport with biogeochemical reactions can be studied on a well-defined scale . While advective solute transport is prevalent in the inter-aggregate macropores, transport in the intra-aggregate micropores is dominated by diffusion . In conjunction with local microbial metabolic activity this often leads to the formation of strong chemical gradients within aggregates. The importance of aggregate-scale heterogeneity in particular for local redox levels has long been recognized . Full anoxic to oxic gradients have been observed within aggregates as small as 4 mm in diameter . Tokunaga et al. showed that anoxic microzones within flat synthetic soil aggregates are likely to support localized sites of Se reduction and documented transport-controlled reduction of soluble Cr to solid Cr taking place exclusively within the surface layer of natural soil aggregates immerged in a Cr solution . Pallud et al. recently investigated ferrihydrite reduction in anoxic flow-through experiments utilizing novel artificial aggregate systems that closely mimic field transport conditions in structured soils and found strong radial gradients in secondary mineralization products as a result of mass-transfer limitations. Utilizing the same artificial aggregate systems, Masue-Slowey et al. investigated arsenic reduction and release in artificial aggregates surrounded by oxic solution and found through reactive transport modeling that the development of an anoxic region within the aggregate best described their experimental results. Given the valuable insights that these novel aggregate reactors systems have shed on the dynamics of iron and arsenic redox chemistry at the aggregate-scale an application to selenium reduction appears consequential. The dynamics of selenium cycling at the aggregate-scale are expected to differ drastically from those investigated so far in these systems, since unlike arsenic, selenium is an example of a contaminant that can be reductively immobilized from solution in soils. In this study I present data on selenium reduction from a series of flow-through reactor experiments utilizing these novel aggregate reactor systems that mimic the dual porosity of structured soils with a microporous artificial soil aggregate contained in a flow-through reactor macropore. Our guiding hypothesis was that aggregate-scale transport coupled to microbial selenium reduction will lead to systematic spatial concentration gradients within aggregates. Similarly to what has been observed for iron minerals and arsenic , we expected the Se reduction rates and emergent gradients to depend on the bulk chemical concentrations of carbon source and electron acceptor, aeration conditions, microbial activity, and the presence of sorptive phases in the solid matrix of aggregates. Our objective was thus to assess the impact of these factors on aggregate-scale selenium reduction and transport as well as to characterize emergent chemical gradients. Aggregates were made of sand or ferrihydrite-coated sand, to assess the impact of sorption on selenium reduction and transport, and initially contained a homogenous distribution of either Thauera selenatis or Enterobacter cloacae SLD1a-1 as model selenium-reducer.