Applied Geochemistry (v.20, #12)

Subject Index (XXIII-LII).

Author Index (LIII-LVIII).

Data are presented for trace element concentrations determined in the <63 μm fraction of streambed sediment samples collected at 24 sites on the island of O‘ahu, Hawai‘i. Sampling sites were classified as urban, agricultural, mixed (urban/agricultural), or forested based on their dominant land use, although the mixed land use at selected sampling sites consisted of either urban and agricultural or forested and agricultural land uses. Forest dominated sites were used as reference sites for calculating enrichment factors.Trace element concentrations were compared to concentrations from studies conducted in the conterminous United States using identical methods and to aquatic-life guidelines provided by the Canadian Council of Ministers of the Environment. A variety of elements including Pb, Cr, Cu and Zn exceeded the aquatic-life guidelines in selected samples. All of the Cr and Zn values and 16 of 24 Cu values exceeded their respective guidelines. The potential toxicity of elements exceeding guidelines, however, should be considered in the context of strong enrichments of selected trace elements attributable to source rocks in Hawai‘i, as well as in the context of the abundance of fine-grained sediment in the streambed of O‘ahu streams.Statistical methods including cluster analysis, Kruskal–Wallis non-parametric test, correlation analysis, and principal component analysis (PCA) were used to evaluate differences and elucidate relationships between trace elements and sites. Overall, trace element distributions and abundances can be correlated to three principal sources of elements. These include basaltic rocks of the volcanic edifice (Fe, Al, Ni, Co, Cr, V and Cu), carbonate/seawater derived elements (Mg, Ca, Na and Sr), and elements enriched owing to anthropogenic activity (P, Sn, Cd, Sn, Ba and Pb). Anthropogenic enrichment gradients were observed for Ba, Cd, Pb, Sn and Zn in the four streams in which sediments were collected upstream and downstream.The findings of this study are generally similar to but differ slightly from previous work on sediments and suspended particulate matter in streams, from two urban watersheds of O‘ahu, Hawai‘i. Inter-element associations in the latter were often stronger and indicated a mixture of anthropogenic, agricultural and basaltic sources of trace elements. Some elements fell into different statistical categories in the two studies, owing in part to differences in study design and the hydrogeological constraints on the respective study areas.

Four karstic springs in southern France, along with rainwater in the same area, were monitored during two hydrological cycles. Stable isotopic ratios (δ18O and δD) allow the contribution of the rain waters from the previous periods (discretised as winter or summer period) to the spring water to be estimated. These computations indicate that heavy rainfall events during the autumn season contribute for 56 ± 7% and 68 ± 9% of the spring water discharge during the following winter and summer seasons, respectively. During the low flow period, residence time does not exceed 1 hydrological cycle, for a large part of the water.Stable isotopic ratios (δ18O and δD) combined with Cl concentration allow the evapotranspiration coefficients to be estimated although the recharge surface is not known and hydrological balance can not be computed. Except for one spring, the evapotranspiration coefficients range from 0.54 to 0.38 (46–62% volume reduction), even during the high flow period.The short residence times, and the strong evapotranspiration coefficients whatever the period (winter or summer) are interpreted as the result of the major role of the epikarst reservoir in the karst recharge functioning over direct or diffuse infiltration.

Ochre is an unwanted waste product that accumulates in wetlands and streams draining abandoned coal and metal mines. A potential commercial use for ochre is to remediate As contaminated soil. Arsenic contaminated soil (605 mg kg−1) was mixed with different ochres (A, B and C) in a mass ratio of 1:1 and shaken in 20 mL of deionised water. After 72 h As concentration in solution was ca. 500 μg kg−1 in the control and 1–2.5 μg kg−1 in the ochre treated experiments. In a second experiment soil:ochre mixtures of 0.05–1:1 were shaken in 20 mL of deionised water for 24 h. For Ochres A and C, as solution concentration was reduced to ca. 1 μg kg−1 by 0.2–1:1 ochre:soil mixtures. For Ochre B, as concentration only reached ca. 1 μg kg−1 in the 1:1 ochre:soil mix. Sorption of As was best modelled by a Freundlich isotherm using As sorption per mass of goethite in the ochre (log  K  = 1.64, n  = 0.79, R 2  = 0.76, p  ⩽ 0.001). Efficiency of ochre in removing As from solution increased with increasing total Fe, goethite, citrate dithionite extractable Fe and surface area.

The role of “excess” CO2 in the formation of trona deposits by Sam Earman; Fred M. Phillips; Brian J.O.L. McPherson (2217-2232).
The prevailing theory for the formation of trona [Na3(CO3)(HCO3) · 2(H2O)] relies on evaporative concentration of Na + – HCO 3 - water produced by silicate hydrolysis of volcanic rock or volcaniclastic sediments. Given the abundance of closed drainage basins dominated by volcanics, it is puzzling that there are so few trona deposits and present-day lakes that would yield dominantly Na–CO3 minerals upon evaporation. Groundwater in the San Bernardino Basin (southeastern Arizona, USA and northeastern Sonora, Mexico) would yield mainly Na–CO3 minerals upon evaporation, but waters in the surrounding basins would not. Analysis of the chemical evolution of this groundwater shows that the critical difference from the surrounding basins is not lithology, but the injection of magmatic CO2. Many major deposits of trona and Na–CO3-type lakes appear to have had “excess” CO2 input, either from magmatic sources or from the decay of organic matter. It is proposed that, along with the presence of volcanics, addition of “excess” CO2 is an important pre-condition for the formation of trona deposits.

Sequential Factor Analysis (seqFA) is presented here as an enhanced alternative to multivariate factorial techniques including robust and classical Factor Analysis (FA) or Principal Component Analysis (PCA). A geochemical data set of 145 sediment samples from very heterogeneous, mainly riverine, deposits of the Rhine-Meuse delta (The Netherlands) analyzed for 27 bulk parameters was used as a test case. The innovative approach explicitly addresses the priority issues when performing PCA or FA: heterogeneity and overall integrity of the data, the number of factors to be extracted, and which optimum minimal set of key variables to be included in the model. The stepwise decision process is based on quantitative and objectively derived statistical criteria, yet also permitting arguments based on geochemical expertize. The results show that seqFA, preferably in combination with robust methods, yields a highly consistent factor model, and is favorable over classical methods when dealing with heterogeneous data sets. It optimizes rotation of the factors, and allows the extraction of less distinct factors supported by only a few variables, thus uncovering additional geochemical processes and properties that would easily be missed with other approaches. The identification of key variables simplifies the geochemical interpretation of the factors, and greatly facilitates the construction of a geochemical conceptual model. For the case of the fluvial deposits, the conceptual model effectively describes their bulk chemical variation in terms of a limited number of governing processes.

Thermochemically induced transformations in Al-smectites: A Spanish natural analogue of the bentonite barrier behaviour in a radwaste disposal by L. Pérez del Villar; A. Delgado; E. Reyes; M. Pelayo; J.M. Fernández-Soler; J.S. Cózar; M. Tsige; A.J. Quejido (2252-2282).
The thermal effect induced by the Morrón de Mateo volcanic dome (Cabo de Gata volcanic region, Spain) on the adjacent bentonitised tuffaceous beds has been studied as a natural analogue of the thermal behaviour of the bentonite-engineered barrier of a geological radwaste repository. These bentonites consist mainly of Fe-rich smectites and were formed in equilibrium with seawater at temperatures between 75 and 95 °C, according to the δ 18O and δD values. In contrast, bentonites from other localities in the region consist mainly of Al-smectites, formed in equilibrium with meteoric water below 25 °C.This investigation is focussed on the detection of the chemical differences between smectites from proximal and distal zones to the dome, as well as to test whether the temperatures calculated based on the O and H isotopic values correspond to their formation or transformation. The initial hypothesis was that the chosen smectites could be formed under marine conditions, being later transformed and isotopically re-equilibrated as a result of the intrusion. To check this hypothesis, a detailed mineralogical, chemical, geochemical and isotopic study has been performed on the smectitised tuffaceous materials and the overlaying biocalcarenites outcropping near and far from the dome.The results show that distal smectites are dioctahedral Al-smectites, similar to those from other deposits in the region, while proximal smectites are Fe- and Mg-rich smectites, showing two evolutionary trends on a Fe–Mg–Al ternary diagram. Similar features are observed when their structural formulae are plotted on the muscovite–celadonite–pyrophylite diagram. Thus, they plot in the smectite domain with interlayer charge less than 1, which is mainly due to octahedral substitution for distal smectites, while for proximal ones it is caused by both octahedral and tetrahedral substitutions. In this ternary diagram, the domains of both proximal and distal smectites are partially overlapped. The coexistence of di- and trioctahedral smectites was only detected in one proximal sample. Further, proximal biocalcarenites are enriched in Fe-rich dolomite in relation to the distal ones.The 87Sr/86Sr and δ 13C values in carbonates and δD in smectites indicate equilibrium with seawater. In contrast, δ 18O values of carbonates and smectites indicate that they were transformed and re-equilibrated between 40 and 90 °C, and between 55 and 66 °C, respectively, independently of their location with respect to the dome.These data suggest that the transformation of calcite into Mg–Fe-carbonates and the occurrence of Fe- and Mg-rich smectites near to the dome resulted from a chemically induced process at similar temperatures. The compositional differences among samples suggests that Fe, Mg and minor Mn were supplied by a contaminant plume originated from the dome, migrating through the sediments and becoming more diluted away from the source. The absence of a well-defined thermal gradient in the system could be due to the small size, semi-closed and shallow character of the basin, as well as to its high underlying volcanic activity.Finally, the results are discussed in terms of analogue processes that can be expected in the bentonite barrier of a radwaste geological repository.

Hydrochemical baseline condition of groundwater at the Mizunami underground research laboratory (MIU) by T. Iwatsuki; R. Furue; H. Mie; S. Ioka; T. Mizuno (2283-2302).
Hydrochemical conditions up to depths of 1000 m below ground level around the Mizunami Underground Research Laboratory were investigated to construct a “baseline condition model” describing the undisturbed hydrochemical environment prior to excavation of the underground facilities at Mizunami, Gifu, Japan. Groundwater chemistry in this area was classified into a Na–Ca–HCO3 type of groundwater in the upper part of sedimentary rock sequence and a Na–(Ca)–Cl type of groundwater in the deeper part of the sedimentary rock sequence and basement granite. The residence time of the groundwaters was estimated from their 14C contents to be approximately 9.3 ka in the middle part of the sedimentary rock and older than 50 ka in the deep part of the granite. The evolution processes of these groundwaters were inferred to be water–rock interactions such as weathering of plagioclase, dissolution of marine sulphate/sulphide minerals and carbonate minerals in the Na–Ca–HCO3 type of groundwater, and mixing between “low-salinity water” in the shallow part and “higher-salinity water” in the deeper part of the granite in the Na–(Ca)–Cl type of groundwater. The source of salinity in the deeper part of the granite was possibly a palaeo-hydrothermal water or a fossil seawater that recharged in the Miocene, subsequently being modified by long-term water–rock interaction. The Cl-depth trend in granitic groundwater changes at a depth of −400 m below sea level. The hydrogeological properties controlling the groundwater flow and/or mixing processes such as advection and diffusion were inferred to be different at this depth in the granite. This hydrochemical conceptual model is indispensable not only when constructing the numerical model for evaluating the hydrochemical disturbance during construction and operation of the MIU facility, but also when confirming a hydrogeological model.

Arsenic mobility in alteration products of sulfide-rich, arsenopyrite-bearing mine wastes, Snow Lake, Manitoba, Canada by Kristin A. Salzsauler; Nikolay V. Sidenko; Barbara L. Sherriff (2303-2314).
The Arsenopyrite Residue Stockpile (ARS) in Snow Lake, Manitoba contains approximately 250,000 tons of cyanide treated, refractory arsenopyrite ore concentrate. The residue was deposited between 1950 and 1959 in an open waste rock impoundment, and remained exposed until 2000, when the pile was capped with layers of waste rock and clay. During the time when the ARS was exposed to the atmosphere, arsenopyrite, pyrrhotite, pyrite and chalcopyrite were oxidized producing scorodite, jarosite and two generations of amorphous Fe sulfo-arsenates (AISA). These secondary phases attenuated some of the As released to pore water during oxidation in the upper layers of the ARS. The imposition of the cap prevented further oxidation. The secondary As minerals are not stable in the reduced environment that currently dominates the pile. Therefore, As currently is being released into the groundwater. Water in an adjacent monitoring well has concentrations of >20 mg/L total As with relative predominance of As(III).

Historical backcasting of metal concentrations in the Chattahoochee River, Georgia: Population growth and environmental policy by Klaus Neumann; W. Berry Lyons; Elizabeth Y. Graham; Edward Callender (2315-2324).
The impact of increasing urbanization on the quality of a river system has been investigated by examining the current concentration of trace metals in the Chattahoochee River south of Atlanta, GA, and comparing these to previously published historical sediment data from reservoirs along the river. The lack of historical data for dissolved metal concentrations prior to ∼1980 requires an approach using these historic metal data from sediment cores. Core data are combined with current suspended load and dissolved metal data to “backcast” dissolved metal concentrations in the metro-Atlanta portion of the Chattahoochee River. The data suggest that the per capita input of dissolved trace metals have actually decreased since the 1920s, but anthropogenic inputs of metal are still a substantial water quality issue.

Grasslands of north-central Kansas are underlain by carbonate aquifers and shale aquitards. Chemical weathering rates in carbonates are poorly known, and, because large areas are underlain by these rocks, solute fluxes are important to estimating global weathering rates. Grasslands exist where the amount of precipitation is extremely variable, both within and between years, so studies in grasslands must account for changes in weathering that accompany changes in precipitation. This study: (1) identifies phases that dominate chemical fluxes at Konza Prairie Biological Station (KPBS) and Long-Term Ecological Research Site, and (2) addresses the impact of variable precipitation on mineral weathering. The study site is a remnant tallgrass prairie in the central USA, representing baseline weathering in a mid-temperate climate grassland.Groundwater chemistry and hydrology in the 1.2 km2 watershed used for this study suggest close connections between groundwater and surface water. Water levels fluctuate seasonally. High water levels coincide with periods of precipitation plus low evapotranspiration rather than during precipitation peaks during the growing season. Precipitation is concentrated before recharging aquifers, suggesting an as yet unquantified residence time in the thin soils.Groundwater and surface water are oversaturated with respect to calcite within limitations of available data. Water is more dilute in more permeable aquifers, and water from one aquifer (Morrill) is indistinguishable from surface water. Cations other than Ca co-vary with each other, especially Sr and Mg. Potassium and Si co-vary in all aquifers and surface water, and increases in concentrations of these elements are the best indicators of silicate weathering at this study site. Silicate-weathering indices correlate inversely to aquifer hydraulic conductivity. 87Sr/86Sr in water ranges from 0.70838 to 0.70901, and it decreases with increasing Sr concentration and with increasing silicate-weathering index. Carbonate extracted from aquifer materials, shales, soil, and tufa has Sr ranging from about 240 (soil) to 880 ppm (Paleozoic limestone). 87Sr/86Sr ranges from 0.70834 ± 0.00006 (limestone) to 0.70904 ± 0.00019 (soil). In all cases, 87Sr/86Sr of aquifer limestone is lower than 87Sr/86Sr of groundwater, indicating a phase in addition to aquifer carbonate is contributing solutes to water.Aquifer recharge controls weathering: during periods of reduced recharge, increased residence time increases the total amount of all phases dissolved. Mixing analysis using 87Sr/86Sr shows that two end members are sufficient to explain sources of dissolved Sr. It is proposed that the less radiogenic end member is a solution derived from dissolving aquifer material; longer residence time increases its contribution. The more radiogenic end member solution probably results from reaction with soil carbonate or eolian dust. This solution dominates solute flux in all but the least permeable aquifer and demonstrates the importance that land-surface and soil-zone reactions have on groundwater chemistry in a carbonate terrain.

Lake Chapala is in the Citala Rift of western Mexico, which in association with the Tepic-Zacoalco and Colima Rifts, form the well-known neotectonic Jalisco continental triple junction. The rifts are characterized by evidence for both paleo- and active hydrothermal activity. At the south shore of the lake, near the Los Gorgos sublacustrine hydrothermal field, there are two tar emanations that appear as small islands composed of solid, viscous and black bitumen. Aliquots of tar were analyzed by GC-MS and the mixtures are comprised of geologically mature biomarkers and an UCM. PAH and n-alkanes are not detectable. The biomarkers consist mainly of hopanes, gammacerane, tricyclic terpanes, carotane and its cracking products, steranes, and drimanes. The biomarker composition and bulk C isotope composition (δ 13C = −21.4%) indicate an organic matter source from bacteria and algae, typical of lacustrine ecosystems. The overall composition of these tars indicates that they are hydrothermal petroleum formed from lacustrine organic matter in the deeper sediments of Lake Chapala exceeding 40 ka (14C) in age and then forced to the lakebed by tectonic activity. The absence of alkanes and the presence of an UCM with mature biomarkers are consistent with rapid hydrothermal oil generation and expulsion at temperatures of 200–250 °C. The occurrence of hydrothermal petroleum in continental rift systems is now well known and should be considered in future energy resource exploration in such regions.

The retention of 22Na and 85Sr on Opalinus clay from Benken and the Mont Terri Rock Laboratory (Switzerland) was measured by through- and out-diffusion measurements on intact rock samples, and by batch sorption on crushed material. The sorption values obtained from these measurements were compared with one another and with ones calculated from a selectivity coefficient/geochemical modelling approach. One of the main aims of the work was to test to what extent sorption values deduced from diffusion experiments on intact rock were compatible with those measured in batch tests on crushed material and with sorption values calculated from geochemical modelling. This is an important consideration in repository performance assessment studies because apparent diffusion coefficients are often calculated from batch R d values and effective diffusion coefficients measured using tritiated water.In general, there was excellent agreement between the distribution ratios for Na obtained by all 3 approaches for both sources of Opalinus clay. For Sr the agreement was also good, but the calculations based on selectivity coefficients in the case of Benken, and the batch sorption data in the case of Mont Terri material were a factor of ∼2 different from the values found in the other two cases. Nevertheless, a factor of ∼2 difference in the worst case is not considered to be significant when the errors associated with the different methods are considered realistically.The uptake mechanism for both elements in the Opalinus clay system is cation exchange and the results strongly indicate that there is no significant difference between the exchange capacity available in the dispersed and in the intact rock systems i.e. reduced site accessibility in intact Opalinus clay is not an issue where sorption by cation exchange is occurring.