Robert E. Stauffer

Arsenic and antimony in geothermal waters of Yellowstone National Park, Wyoming, USA

Abstract A total of 268 thermal spring samples were analyzed for total soluble As using reduced molybdenum-blue; 27 of these samples were also analyzed for total Sb using flame atomic absorption spectrometry. At Yellowstone the Cl As atomic ratio is nearly constant among neutral-alkaline springs with Cl > 100 mg L −1 , and within restricted geographic areas, indicating no differential effects of adiabatic vs . conductive cooling on arsenic. The Cl As ratio increases with silica and decreases with decreasing Cl ΣCO 3 ; the latter relationship is best exemplified for springs along the extensively sampled SE-NW trend within the Lone Star-Upper-Midway Basin region. The relationship between Cl As and Cl ΣCO 3 at Yellowstone suggests a possible rock leaching rather than magmatic origin for much of the Parks total As flux. Condensed vapor springs are low in both As and Cl. Very high Cl As ratios ( > 1000) are associated exclusively with highly diluted (Cl −1 ) mixed springs in the Norris and Shoshone Basins and in the Upper White Creek and Firehole Lake areas of Lower Basin. The high ratios are associated with acidity and/or oxygen and iron; they indicate precipitation of As following massive dilution of the Asbearing high-Cl parent water. Yellowstone Sb ranged from 0.009 at Mammoth to 0.166 mg L −1 at Josephs Coat Spring. Within basins, the Cl Sb ratio increases as the Cl ΣCO 3 ratio decreases, in marked contrast to As. Mixed springs also have elevated Cl Sb ratios. White (1967) and Weissberg (1969) previously reported stibnite (Sb 2 S 3 ), but not orpiment (As 2 S 3 ), precipitating in the near surface zone of alkaline geothermal systems.

Cycling of iron, manganese, silica, phosphorus, calcium and potassium in two stratified basins of Shagawa Lake, Minnesota

Geochemical studies of Fe, Mn, Si, P, Ca, and K were made in two basins of Shagawa Lake, Minnesota during summers 1977–1979. Within the metalimnion, upward and laterally diffusing Fe(II) oxidizes and co-precipitates a large fraction of soluble P. Nevertheless, the soluble FeP ratio decreases while progressing upward through the oxidation zone. The mean paniculate FeSi ratio is 3.0 ± 0.5 in the thermocline, but because of the stoichiometric excess of dissolved Si over Fe, and preferential associations between Fe and P, the oxidation of Fe has a negligible effect on the dissolved SiO2 profiles. The lake sediments are net sources of Ca, K, and P during summer stratification. These elements are transported upward through the thermocline and exported from the lake. Despite the very active sediment release of dissolved SiO2, and upward transport through the thermocline, the lake acts as a net sink for biogenic opal in summer. The prominent soluble Mn maxima in the thermocline of the east basin result from antecedent sediment release (Mn2+) and vertical transport (primarily in the west basin), followed by dispersion and oxidation in the epilimnion, and finally downward settling and dissolution within the east basin thermocline. Rhodochrosite solubility does not influence Mn2+ profile development. Very large percentage increases in epilimnetic Fe, Mn, Si, and P occur simultaneously during episodes of wind-activated entrainment of metalimnion water. These increases are in stoichiometric agreement with the entrained water mass, provided account is also taken of the preformed slurries overlying the metalimnetic shelf sediments in mid-summer. The bathylimnetic soluble FeP ratio remains low since the cultural inputs of P were terminated in early 1973. This geochemical stability is explained by the large reserves of potentially mobile P in the upper post-cultural sediments. Phosphorus release from these sediments and transport phenomena account for the persistent eutrophic condition of the lake during mid-late summer, despite the 80% reduction in external P loading.

Vertical nutrient transport and its effects on epilimnetic phosphorus in four calcareous lakes

The influence of vertical nutrient transport on epilimnetic phosphorus was studied comparatively in four calcareous Wisconsin lakes during 1972. In the two lakes with steep metalimnetic nutrient gradients (Mendota, Delavan), upward fluxes by entrainment and eddy diffusion exceeded all other influxes combined; here epilimnetic total-P concentrations increased during period of high windpower and thermocline migration, and decreased during comparatively stagnant intervals. In the two lakes lacking upper metalimnetic P gradients (Green, Fish), higher windpower had little or no effect on epilimnetic phosphorus.In each of the four lakes epilimnetic P declined in early summer until a quasi steady-state was achieved between P influxes and P sedimentation. In Mendota and Delavan steady-state featured higher concentrations of total-P, and much higher epilimnetic concentrations of particulate-P and chlorophyll, than in Green and Fish Lakes — mainly on account of the high fluxes of molybdate-reactive, biologically-available P through the seasonal thermocline. The flux analysis illustrates why mean lake TP concentration after ice-out is an inconsistent measure (not sufficient statistic) for estimating nutrient potential and chlorophyll standing crops during the following summer in stratified lakes.

Hydrology and alkalinity regulation of soft Florida waters: An integrated assessment

Natural waters in ridge provinces of Florida and southeast Georgia were classified geographically, by degrees of cultural disturbance, and according to the dominant hydrologic and biogeochemical processes controlling chemistry. The ionic composition of lakes, upland streams, and surficial aquifer (water table) springs in relatively undeveloped catchments reflects the geographic variations in bulk deposition corrected for evapotranspiration (Na, Cl), plus a slight gain (net watershed mobilization) of Mg, and partial to nearly complete losses (net retention) of nitrate, sulfate, Ca, and K. Recharge to the Floridan aquifer in infertile, forested, sandy ridge provinces of northern Florida contains 360–580 μmol CO2. On the basis of indirect geochemical evidence, sulfate retention appears less important in lake sediments than in the regions highly weathered, ferruginous, kaolinitic, sand soils. Silica concentrations in upland streams and water table springs closely reflect the predicted equilibrium between kaolinite and gibbsite. Along with other evidence, the Si concentrations in ridge lakes indicate that seepage inflow is much more important than assumed in Baker et al.s (1988) regional model. Lakes and streams are acidified either by humic acids or nonmarine sulfate but rarely by both, as reflected by the significant inverse correlation between these two components. Contrary to previous reports, there is no significant difference in alkalinity for culturally undisturbed lakes in the northern Trail versus southern Highlands Ridge areas.

Testing lake energy budget models under varying atmospheric stability conditions

Abstract Several variations of a basic energy budget model were tested under varying atmospheric stability conditions using temperature data from Lake Mendota and meteorologic data from nearby Truax Airfield, Madison, Wisconsin. The flux enhancements predicted by Model 1 (Deardorff) for prevailing over-lake lapse conditions were inconsistent with measured changes in lake heat storage. When using the Budyko formulation for sky emissivity, e, satisfactory energy residuals were obtained by eliminating the lapse enhancement while retaining the suppressive effects of over-lake inversions (Model 2 = modified Deardorff). Here, the best estimate of the 10 m Dalton coefficient was 1.45 × 10−3 (cgs units; neutral profile). However, when the Koberg model for e was adopted instead, satisfactory energy residuals were obtained only after eliminating all adjustments for non-neutral conditions, and increasing the 10 m Dalton coefficient to approximately 1.55 × 10−3 (Model 3). Models 2 and 3 predicted annual evaporation from Lake Mendota of 71 and 86 cm during 1977 (including approximately 6 cm of sublimation during ice cover), bracketing earlier estimates for this locality (76 cm) based on U.S. Weather Bureau Class A pan data, and consistent with intensive studies of Pretty Lake in northern Indiana. My estimates of the 10 m Dalton coefficient (1.45 × 10−3 to 1.55 × 10−3) also narrowly bracket values for Lake Hefner and Lake Mead (1.49 × 10−3), obtained in studies by the U.S. Geological Survey. Moreover, as in the case of the Will Rogers airport near Lake Hefner, meteorologic data collected at Truax Airfield, Wisconsin, proved adequate for estimating sensible and latent heat transfer over the adjacent lake.

A comparative analysis of iron, manganese, silica, phosphorus, and sulfur in the hypolimnia of calcareous lakes

Abstract The chemistry of the hypolimnion was studied in late summer in 23 individual basins of 19 calcareous lakes situated in the Midwestern U.S.A. The basins represented a broad range in morphometry, and drainage basin influences. All but two of the basins were anoxic by early September. The concentrations of Fe were low in the anoxic basins because of precipitation of FeS. Excess sulfide accumulated in the hypolimnion (up to 7 mg 1−1) making these basins sulfuretums. The concentrations of total-P ranged widely—from 0.03 mg 1−1 in several seepage lakes to > 0.5 mg 1−1 in drainage lakes with very short τw, or influenced by agricultural or urban drainage, or municipal waste. Neither apatite nor vivianite limited the hypolimnetic build-up of this key nutrient. Hypolimnetic Mn was almost entirely soluble if the basin was anoxic. The concentrations indirectly reflected external inputs, not the thermodynamics of rhodochrosite. Concentrations of dissolved silica were generally much lower than in inflowing waters. The hypolimnetic Si/P ratio was low unless the lake lacked surface inflows and also experienced high rates of groundwater inflow rich in silica and very low in P. Because of the very low hypolimnetic Fe/P and Fe/Si ratios, phosphate and silica are liberated from significant chemical interactions with Fe accompanying their eddy diffusion transport upward within the hypolimnion, and across the metalimnetic oxycline. Because of external P inputs and the efficiency of internal P recycling, Si is potentially a limiting nutrient for diatoms in calcareous lakes.

Fluorapatite and fluorite solubility controls on geothermal waters in Yellowstone National Park

Abstract Potential solubility controls on phosphorus in Yellowstone National Park geothermal waters were investigated using the analytical phosphate estimates of Stauffer and Thompson (1978), the computer program, WATEQF, and adopting the equilibrium constant: log K 25° = −61.4 for fluorapatite (FAP = Ca5(PO4)3F) dissolution. The near-boiling high-Cl geyser and spring effluents are at or near saturation with respect to (with) FAP. The sixteen representative springs in this category had FAP saturation indices ( S.I. = log IAP/K r ) ranging from − 3.2 to +4.9 and averaging +0.9. The strongly positive indices were all associated with the highly alkaline conditions resulting from adiabatic cooling in the near surface environment. Hot spring waters indicating extensive dilution (reduced Cl) by meteoric water have lower pHs, and despite PO4 and Ca concentrations an order of magnitude higher than the geysers, are still frequently undersaturated with FAP. The travertine-depositing “Mixed-water” springs are invariably supersaturated with FAP at ground surface and at or near saturation with hydroxylapatite. Supersaturation may result from kinetic inhibition of apatite crystallization by the elevated Mg2+, HCO3−, and lower temperatures in these springs. The phosphates may be residuals of the meteoric dilution water. Separately, if Strubels temperature-dependent estimates of fluorite (CaF2) solubility are adopted, the high-Cl geysers and springs on “Geyser Hill” and at Norris are consistently undersaturated with CaF2 at the 90–100° orifice temperatures. The apparent disequilibrium may reflect fluorite equilibration at the much higher temperatures (> 200°C) in the deeper enthalpy reservoirs.

Linkage between the phosphorus and silica cycles in lake Mendota, Wisconsin

Abstract The phosphorus and silica cycles in Lake Mendota were studied during 1970–1973, bracketing the diversion of all municipal wastewater inputs to the lake, and again in 1977–1980, following 1 to built;1 2 lake hydraulic residence times. Because of the high concentrations of soluble phosphate and inorganic nitrogen, the vernal diatom bloom is limited by silica. The low nutrient uptake ratio ( mean ΔSi ΔP = 5.5 ± 0.2, by atoms ) reflects nutrient availability and the dominance of low silica diatoms (especially Stephanodiscus species) in this eutrophic calcareous lake. The lower ratio, ΔSi ΔP , in the hypolimnion during summer stratification (mean 2.8 ± 0.1) illustrates that phosphorus recycling proceeds more efficiently than in the case for Si. The vertical fluxes of biologically available P and Si into the epilimnion averaged 9.5 ± 1.0 and 33 ± 6 mg m −2 ( A 0 ) day −1 during summers 1977–1979. Despite the low Si P flux ratio, soluble silica accumulated each summer in the epilimnion whereas epilimnetic P declined. This difference is probably a consequence of seasonal blue-green algal dominance in the epilimnion. The available Si P in the epilimnion shifts downwards by 3 orders of magnitude during September and October, because of entrainment of metalimnetic and hypolimnetic waters with a lower ratio, and the selective uptake of Si by the fall diatom crop. Because of the seasonal demands by diatoms, coupled with the comparative inefficiency of Si recycling (vs P), Lake Mendotas annual retention coefficient for soluble silica is 0.98 ± 0.01, as compared to 0.40 ± 0.05 for biologically available P. Based on various nutrient and hypolimnetic redox indices, Lake Mendota failed to respond during the eight years following nutrient diversion. Nor has the lake changed significantly in the twentieth century.

Effects of an early fall cold front on heat, phosphorus, silica, and manganese distributions in the hypolimnion of Lake Mendota, Wisconsin

Abstract I examined the effects of an early fall (8–10 September 1977) cold front on eutrophic Lake Mendota, Wisconsin, during which ‘Lake Number’ L N (dimensionless) attained a minimum 3 h value of 0.45, and remained below 1.0 for a total of 39 h. The front accelerated heat transport downward through the main thermocline, resulting in hypolimnetic warming; the vertical eddy conductivity K z increased monotonically with depth from a minimum 0.8 m 2 day −1 at 16 m to 1.9 m 2 day −1 at 21 m. The front also affected significant lateral and vertical redistribution of three passive solute ‘tracers’: phosphate (P), silicic acid (Si) and soluble reduced manganese (Mn 2+ ). Lateral concentration gradients along isopycnals predating the storm were erased. For Mn, especially, the vertical redistribution within the hypolimnion agreed with modeled changes based on vertical gradients and the K z profile. Mn and P behaved quasi-conservatively within the anoxic hypolimnion (no net sources or sinks), but additional supplies of Si appeared in the hypolimnion (8.0 ± 1.5 mg m −3 day −1 ). This rate over 4 days agreed with the mean rate of hypolimnetic Si accumulation over the 80 day interval preceding 7 September 1977 (8.7 mg m −3 day −1 ). These results corroborate other environmental evidence that Mn and P both enter Lake Mendotas hypolimnion mainly during the first half of the summer, prior to the onset of bulk hypolimnetic anoxia, whereas Si is released by hypolimnetic sediments continuing into the late summer.

Environmental tracer analysis of winter profile development in two basins of Shagawa Lake, Minnesota

Abstract Winter profile development was studied comparatively in two basins ( z max = 14 m) separated by a 6 m sill in Shagawa Lake, Minnesota. Based on physical arguments and profile measurements, heat and solutes released from shelf sediments in the western basin cause convective instability at 8–11 m. After a mid-winter interval of instability, radon profiles indicate very low rates of vertical mixing in the benthic boundary layers following their stabilization by ferrous iron. Because of hydrographic features, two types of plunging inflows sometimes enter the western basin in late winter; one anthropogenic inflow alters the lower profiles by adding sodium, radon, and oxygen; the second modifies the mid-depth region by adding ‘weathering tracers’ (especially calcium and silicon) derived from the Ely greenstone. The oxidation of Fe 2+ diffusing upward and laterally from the sediment-water interface is accompanied by near-complete coprecipitation of phosphorus. The oxidation of Fe 2+ has only a minor effect on silicic acid, probably because of unfavorably large Si:P and preferential association of Fe(III) with phosphorus. Despite low pH and temperature, detectable amounts of Mn 2+ oxidize and precipitate above iron in the water column. Solute profiles differentially register the complex effects of hydrodynamic vs. geochemical processes, and as a consequence, affect the estimation of solute budgets in ice-covered lakes.