ABSTRACT
Changes in groundwater discharge to streams caused by irrigation practices can influence stream temperature. Observations along two currently flood-irrigated reaches in the 640-square-kilometer upper Smith River watershed, an important agricultural and recreational fishing area in west-central Montana, showed a downstream temperature decrease resulting from groundwater discharge to the stream. A watershed-scale coupled surface water and groundwater flow model was used to examine changes in streamflow, groundwater discharge to the stream and stream temperature resulting from irrigation practices. The upper Smith River watershed was used to develop the model framework including watershed climate, topography, hydrography, vegetation, soil properties and current irrigation practices. Model results were used to compare watershed streamflow, groundwater recharge, and groundwater discharge to the stream for three scenarios: natural, pre-irrigation conditions (PreIrr); current irrigation practices involving mainly stream diversion for flood and sprinkler irrigation (IrrCurrent); and a hypothetical scenario with only groundwater supplying sprinkler irrigation (IrrGW). Irrigation increased groundwater recharge relative to natural PreIrr conditions because not all applied water was removed by crop evapotranspiration. Groundwater storage and groundwater discharge to the stream increased relative to natural PreIrr conditions when the source of irrigation water was mainly stream diversion as in the IrrCurrent scenario. The hypothetical IrrGW scenario, in which groundwater withdrawals were the sole source of irrigation water, resulted in widespread lowering of the water table and associated decreases in groundwater storage and groundwater discharge to the stream. A mixing analysis using model predicted groundwater discharge along the reaches suggests that stream diversion and flood irrigation, represented in the IrrCurrent scenario, has led to cooling of stream temperatures relative to natural PreIrr conditions improving fish thermal habitat. However, the decrease in groundwater discharge in the IrrGW scenario resulting from large-scale groundwater withdrawal for irrigation led to warmer than natural stream temperatures and possible degradation of fish habitat.
ABSTRACT
Mercury is a worldwide contaminant derived from natural and anthropogenic sources. River systems play a key role in the transport and fate of Hg because they drain widespread areas affected by aerial Hg deposition, transport Hg away from point sources, and are sites of Hg biogeochemical cycling and bioaccumulation. The Madison and Missouri Rivers provide a natural laboratory for studying the fate and transport of Hg contributed by geothermal discharge in Yellowstone National Park and from the atmosphere for a large drainage basin in Montana and Wyoming, United States of America (USA). Assessing Hg in these rivers also is important because they support fishery-based recreation and irrigated agriculture. During 2002 to 2006, Hg concentrations were measured in water, sediment, and fish from the main stem, 7 tributaries, and 6 lakes. Using these data, the geothermal Hg load to the Madison River and overall fate of Hg along 378 km of the Missouri River system were assessed. Geothermal Hg was the primary source of elevated total Hg concentrations in unfiltered water (6.2-31.2 ng/L), sediment (148-1100 ng/g), and brown and rainbow trout (0.12-1.23 µg total Hg/g wet weight skinless filet) upstream from Hebgen Lake (the uppermost impoundment). Approximately 7.0 kg/y of geothermal Hg was discharged from the park via the Madison River, and an estimated 87% of that load was lost to sedimentation in and volatilization from Hebgen Lake. Consequently, Hg concentrations in water, sediment, and fish from main-stem sites downstream from Hebgen Lake were not elevated and were comparable to concentrations reported for other areas affected solely by atmospheric Hg deposition. Some Hg was sequestered in sediment in the downstream lakes. Bioaccumulation of Hg in fish along the river system was strongly correlated (r(2)=0.76-0.86) with unfiltered total and methyl Hg concentrations in water and total Hg in sediment.
