Applied Geochemistry (v.22, #8)

Metal interactions with natural organic matter by D. Scott Smith; Fengchang Wu (1567).

Organo-colloidal control on major- and trace-element partitioning in shallow groundwaters: Confronting ultrafiltration and modelling by Olivier Pourret; Aline Dia; Mélanie Davranche; Gérard Gruau; Odile Hénin; Maxime Angée (1568-1582).
Ultrafiltration experiments using new small ultracentifugal filter devices were performed at different pore size cut-offs to allow the study of organo-colloidal control on metal partitioning in water samples. Two shallow, circumneutral pH waters from the Mercy site wetland (western France) were sampled: one dissolved organic carbon (DOC)- and Fe-rich and a second DOC-rich and Fe-poor. Major- and trace-element cations and DOC concentrations were analysed and data treated using an ascendant hierarchical classification method. This reveals the presence of three groups: (i) a “truly” dissolved group (Na, K, Rb, Ca, Mg, Ba, Sr, Si and Ni); (ii) an inorganic colloidal group carrying Fe, Al and Th; and (iii) an organic colloidal group enriched in Cr, Mn, Co, Cu and U. However, REE and V have an ambivalent behaviour, being alternatively in the organic pool and in the inorganic pool depending on sample. Moreover, organic speciation calculation using Model VI were performed on both samples for elements for which binding constants were available (Ca, Mg, Ni, Fe, Al, Th, Cr, Cu, Dy, Eu). Calculation shows relatively the same partitioning of these elements as ultrafiltration does. However, some limitations appear such as (i) a direct use of ultrafiltration results which tends to overestimate the fraction of elements bound to humic material in the inorganic pool as regards to model calculations as well as, (ii) a direct use of speciation calculation results which tends to overestimate the fraction of elements bound to humic material in the organic pool with regard to ultrafiltration results. Beside these limitations, one can consider that both techniques, ultrafiltration and speciation calculation, give complementary information, especially for more complex samples where inorganic and organic colloids compete.

Three experimental techniques – ion exchange, liquid–liquid extraction with competitive ligand exchange, and solid-phase extraction with competitive ligand exchange (CLE–SPE) – were evaluated as methods for determining conditional stability constants (K) for the binding of mercury (Hg2+) to dissolved organic matter (DOM). To determine the utility of a given method to measure stability constants at environmentally relevant experimental conditions, experimental results should meet three criteria: (1) the data must be experimentally valid, in that they were acquired under conditions that meet all the requirements of the experimental method, (2) the Hg:DOM ratio should be determined and it should fall within levels that are consistent with environmental conditions, and (3) the stability constants must fall within the detection window of the method. The ion exchange method was found to be limited by its detection window, which constrains the method to stability constants with log  K values less than about 14. The liquid–liquid extraction method was found to be complicated by the ability of Hg–DOM complexes to partition into the organic phase. The CLE–SPE method was found to be the most suitable of these methods for the measurement of Hg–DOM stability constants. Stability constants for DOM isolates measured using the CLE–SPE method at environmentally relevant Hg:DOM ratios were log  K  = 25–30 (M−1). These values are consistent with the strong Hg2+ binding expected for reduced S-containing binding sites.

Rapid quantification of humic and fulvic acids by HPLC in natural waters by F.C. Wu; R.D. Evans; P.J. Dillon; Y.R. Cai (1598-1605).
A simple method based on high-performance size-exclusion chromatography (HPSEC) has been developed for rapid quantification of humic and fulvic acids (HA and FA) in stream waters. A Tsk-gel column was used to separate natural dissolved organic matter (DOM) into two components: peak A and B. In terms of HPSEC chromatograms and fluorescence patterns, peak A and B were similar to the corresponding XAD-extracted HA and FA, respectively. It is suggested that peak A fraction mainly consisted of HA, and peak B fraction FA. The similar separation of HA and FA using HPSEC and a conventional XAD method suggests the consistency of molecular size distribution and physical–chemical properties of DOM. HPSEC offers a simple and rapid method for the quantification of HA and FA instead of tedious extractions of humic substances. Analyses of natural water samples show that the calculation of HA/FA based on UV absorbance was under- or over-estimated, the calibration using the extracted HS allows a more accurate quantification. The fast screening of HA and FA provides useful quantitative and qualitative information that can be used in environmental or monitoring studies.

Determination of sulfide ligands and association with natural organic matter by James R. Kramer; Russell A. Bell; D. Scott Smith (1606-1611).
Group B metals, such as Hg, Cu, Ag, Pb and Cd bind strongly to reduced inorganic and organic S(II−) ligands. These S(II−) ligands, stable in oxic waters for significant periods of time, occur at the <1–100 s nM concentrations. It is hypothesized that S(II−) ligands are stabilized as Cu–S molecules associated with organic matter by multi-ligand binding or in nano-pore encapsulations in organic matter. S(II−) ligands are estimated by two methods: purge/trap analysis as Cr-reducible sulfide (CRS), and strong ligand (SLT) from a competitive ligand titration with Ag(I). The CRS/SLT ratio is nearly one for selected samples. CRS correlates reasonably well (r 2  ∼ 0.5) with organic C with a slope of 14.6 nM per mg C. The conditional binding constant of Ag–SL is 11.3 for effluent associated with waste-water and decreases for river waters from about 12–8.8 as the strong sites are occupied with Ag(I).

The results of a comprehensive 113Cd NMR study of complexation of Cd2+ by Suwannee River natural organic matter (NOM) are presented and compared with the more familiar behavior of simple systems containing Cd2+ and a single ligand (e.g., ethylenediamine). pH-dependent trends in the chemical speciation of Cd(II), chemical shift of 113Cd, and line width of resonance peaks lead to the conclusion that the contribution of N donor atoms to the primary coordination sphere of Cd2+ increases steadily with increasing pH, so much so that relatively strong downfield resonances can be observed at alkaline pH. Those observations also indicate that there is no accessible, unique combination of magnetic field strength and solution chemistry for which the exchange rates of all forms of Cd are simultaneously fast, intermediate, or slow.

WHAM, incorporating Humic Ion Binding Model VI, was used to analyse published data describing the binding of Hg(II) and methylmercury (CH3Hg) by isolated humic substances. For Hg(II), the data covered wide ranges of pH and levels of metal binding, whereas for CH3Hg the range of metal binding was relatively narrow. Data were fitted by adjustment of a single model parameter, log K MA, the intrinsic equilibrium constant characterising, in the standard version of the model, the binding of metal ions and their first hydrolysis products to humic carboxylic acid groups. Other model parameters, including those characterising the tendency of metal ions to interact with “softer” ligand atoms (N and S), were held at their default values. The importance of the first hydrolysis products in binding was considered, and also the possible influence of competition by residual Fe(III), bound to the humic matter.Of the 11 data sets for Hg(II), eight gave results reasonably consistent with one another, and with the previously-estimated default values of log K MA. There was no consistent indication that assuming the presence or absence of competing Fe(III) gave superior fits; neither did the inclusion or exclusion of HgOH+ binding provide consistently better results. The experimental data and the model show that apparent binding strength towards Hg(II) is highly dependent upon the metal loading, reflecting the high degree of heterogeneity in binding sites for the metal. Of the 24 metals to which WHAM/Model VI has now been applied, Hg(II) shows the strongest binding to humic substances, and the greatest range in binding affinities. The relatively few data characterising the interactions of CH3Hg with humic substances can be approximately fitted with the model. The results show that CH3Hg binding is appreciably weaker than that of Hg(II). New default values of log K MA are 3.6 for Hg(II)–HA binding, 3.1 for Hg(II)–FA and 0.3 for CH3Hg–HA and CH3–FA.

The ecological roles of dissolved organic matter (DOM) in seawater have not been well understood. One definite function of DOM stems from its complexation ability with trace metals under the conditions of seawater. A chemical complexation model of the marine system was introduced in order to clarify the ecological roles of strong organic ligands in DOM related to the acquisition of bioactive metals (Cu, Fe and Zn) by phytoplankton, assuming that two types of strong organic ligands coexist in oceanic DOM and complexes with bioactive metals. The results reveal that the weaker organic ligand, rather than the stronger one, plays a significant role in the reduction of Cu toxicity for phytoplankton growth. It is suggested that the presence of reactions with Cu that are competitive to the strong organic ligand causes extremely low Fe concentrations in seawater and leads to Fe deficiency for phytoplankton growth. Therefore, it is concluded that the strong ligands in DOM play a chemical role in controlling free ion concentration levels of bioactive metals in the marine environment.

An agent-based biogeochemical model has been developed which begins with biochemical precursor molecules and simulates the transformation and degradation of natural organic matter (NOM). This manuscript presents an empirical quantitative structure activity relationship (QSAR) which uses the numbers of ligand groups, charge density and heteroatom density of a molecule to estimate Cu-binding affinity ( K Cu ′ ) at pH 7.0 and ionic strength 0.10 for the molecules in this model. Calibration of this QSAR on a set of 41 model compounds gives a root mean square error of 0.88 log units and r 2  = 0.93. Two simulated NOM assemblages, one beginning with small molecules (tannins, terpenoids, flavonoids) and one with biopolymers (protein, lignin), give markedly different distributions of log K Cu ′ . However, calculations based on these log K Cu ′ distributions agree qualitatively with published experimental Cu(II) titration data from river and lake NOM samples.

Relationships between DOC concentration, molecular size and fluorescence properties of DOM in a stream by F.C. Wu; D.N. Kothawala; R.D. Evans; P.J. Dillon; Y.R. Cai (1659-1667).
The patterns of dissolved organic matter (DOM) fluorescence properties were examined in a Precambrian shield stream over a seven-month field study. Unique spatial and temporal patterns of simultaneous changes were observed in dissolved organic carbon concentration (DOC), humic-like fluorescence intensity, maximum excitation and emission wavelengths and fluorescence index (the ratio of the emission intensity at a wavelength of 450 nm to that at 500 nm at an excitation wavelength of 370 nm). The spatial change indicates the alteration of DOM along the length of the stream, and temporal change corresponded to a drought event in August. In contrast to humic-like fluorescence, the protein-like fluorescence shows considerable variability, suggesting its ephemeral nature. There were strong relationships between humic-like fluorescence intensity, fluorescence index, maximum Ex/Em wavelengths, DOC concentration and molecular size of DOM. This study has significant implications to the understanding of the nature and biogeochemical cycling of DOM.

Fluorescence characterization of dissolved organic matter in an urban river and its complexation with Hg(II) by Pingqing Fu; Fengchang Wu; Congqiang Liu; Feiyue Wang; Wen Li; Lanxiu Yue; Qingjun Guo (1668-1679).
Dissolved organic matter (DOM) in the Nanming River, an urban river in Guiyang City in SW China, and its complexation behavior with Hg(II) were investigated using fluorescence spectroscopy and the quenching titration technique. Three major fluorophores, two humic-like and one protein-like fluorescence, were observed in most of the DOM samples. Significant correlations were observed between the humic-like and protein-like fluorescence intensities, as well as, between them and other water quality parameters such as dissolved organic carbon, PO 4 3 - , chemical oxygen demand and NH 4 + concentrations, suggesting that agricultural and municipal wastewaters may be the source for both protein-like and humic-like fluorescence materials in the river. The fluorescence quenching titration resulted in similar values for the conditional stability constants for Hg(II) complexes with the humic-like and protein-like fluorophores, likely due to the dominance of Hg binding with O-containing function groups at the high Hg(II) concentrations used in the titration. Effects of Cl, Ca2+, Mg2+ and Cu2+ ions on the binding between Hg(II) and three different fluorophores were also studied. The fluorescence index from the Nanming River was further found to be controlled by pH and Hg(II), cautioning its use in discriminating the sources of DOM.

Watershed-scale geochemistry by W. Berry Lyons; Russell S. Harmon; David T. Long (1680-1681).

Solute-based geochemical mass balance methods are commonly used in small-watershed studies to estimate rates of a variety of geochemical processes at the Earth’s surface, including primary-mineral weathering and soil formation, and the quantitative contribution of these elemental transfer processes to cation budgets, nutrient cycling, and landscape susceptibility to acid deposition. Weathering rates of individual minerals in watershed mass-balance studies are determined by solving a system of simultaneous linear geochemical mass-balance equations with constant (stoichiometric) coefficients. These equations relate the measured net fluxes to the (known) stoichiometries and (unknown) rates of weathering reactions for multiple minerals in the weathering profiles. Solving the system of equations requires petrologic, mineralogic, hydrologic, botanical, and aqueous geochemical data. The number of mineral-weathering rates that can be determined is limited by the number of elements for which solute mass-balance equations can be written. In addition to calculating mineral weathering rates, elemental transfer into or out of the biomass may also be calculated. Elemental uptake by aggrading forest vegetation can act as an intrawatershed sink for at least some mineral-derived cations, producing mineral weathering rates higher than would be estimated from solute fluxes alone; similarly, element release from decaying forest biomass can result in higher solute fluxes than are produced by weathering alone. The mathematics of, significant contributions from, role of biomass in, and recent advances in, watershed geochemical mass-balance methods are discussed using examples from the Appalachian headwaters watersheds of the Coweeta Hydrologic Laboratory in the southern Blue Ridge Physiographic Province of North Carolina, USA.

Taiwan is a typical active orogenic belt situated at the collision boundary between the Eurasian Continental Plate and the Philippine Sea Plate. Dissolved major and trace constituents, as well as Sr and Sr isotopes in river waters collected from the Danshuei River basin in northern Taiwan have been studied to evaluate chemical weathering processes. The results of principal component analysis show that the ion sources in these river waters can be categorized into 3 major components: chemical weathering, seasalt contribution and local anthropogenic input. The chemical weathering is the most dominant factor that contributes about 85% of total variances. Significantly increased Na/Cl and Ca/Cl, as well 87Sr/86Sr, were observed in most upper stream samples. The Na/Cl and Ca/Cl ratios in the Dahan Stream, however, are much higher than the Shindien Stream. Even though average rainfall is stronger in the Shindien drainage basin, chemical evidence from river waters supports less intense weathering in the region. Selective dissolution of secondary calcites explains the observed high Ca/Cl, Sr/Cl and Ca/Na in the Dahan Stream. These results highlight the potential importance of tectonic factors, such as uplift and physical erosion in studying chemical weathering in an active orogenic belt.The variations of 87Sr/86Sr in the Danshuei River are quite large, reflecting some strata that released Sr. Most of the upstream waters exhibit more radiogenic 87Sr/86Sr, 0.713243–0.714338, due to weathering of ambient low-grade metamorphic rocks, 0.71678–0.72216. The distributions of Cl, Sr and 87Sr/86Sr in the main stream were affected by somewhat conservative mixing with sources varing between ambient rocks and seasalt. In the upper Dahan Stream, heavy 87Sr/86Sr ratios were coincident with large deviations of Na/Cl from the average seawater value, as high as 40. The degree of chemical weathering in ambient rocks plays a dominant role in affecting the distribution of Sr and 87Sr/86Sr in the Danshuei River. This isotopic characteristic makes 87Sr/86Sr an invaluable tracer for studying source mixing, migration pathways and chemical weathering in an active orogenic belt.

The Haast and Clutha rivers drain opposing flanks of New Zealand’s Southern Alps. Major element analyses of grain size fractions (2–1 mm, 1 mm–355 μm, 355–63 μm, and <63 μm) from bedload sediments collected throughout the reach of each river suggest that weathering is strongly partitioned between the chemical weathering of carbonates and the physical weathering of silicates. Sand size fractions from both rivers are depleted in CaO (∼0.2–2.1 wt%) relative to source schists (∼3 wt% CaO), while silt fraction CaO concentrations range from 2–5 wt%. The depletion of CaO in the sediments is interpreted to be due at least in part to removal of carbonate during chemical weathering of the schist protolith in the soil zone. The observed covariance of CaO and P2O5 concentrations in all river sediment suggests that most CaO is bound in a combination of phosphate-bearing minerals such as apatite along with other heavy mineral phases with similar hydrodynamic properties (e.g. epidote). Chemical index of alteration (CIA) values for grain size fractions from both rivers are similar (Haast: 54–63, Clutha: 49–61) and do not systematically vary with grain size or sample location. Al2O3–CaO  + Na2O–K2O (A–CN–K) relationships suggest that CIA values are controlled by albite–muscovite mixing rather than feldspar weathering. Both A–CN–K relationships and modal mineralogical calculations from Clutha river samples indicate progressive downstream attrition of muscovite from coarser to finer grain size fractions. In contrast, Haast river sediments display less variable normative muscovite concentrations and no downstream enrichment/depletion trends.The Haast and Clutha watersheds have drastically different sediment yields, but the similarity of sediments from both rivers indicates that there is minimal climatic control on the weathering intensity of fluvial sediments. Rather, bedload geochemistry is controlled primarily by mechanical breakdown of lithic fragments and subsequent preferential attrition of muscovite > albite > quartz. The geochemical signature of mechanical attrition and hydrodynamic winnowing is more developed in Clutha river samples because of longer sediment residence time within its fluvial system. These findings suggest that high standing island (HSI) fluvial sedimentary evolution is characterized by the dominance of physical weathering processes and the absence of silicate chemical weathering signatures.

Pandoh is a beautiful Lake surrounded by Shivalik sequences (Lesser Himalaya) in the Mandi District of Himachal Pradesh. A comprehensive and systematic study on the seasonal pattern of major ions (Cl, PO 4 3 - , HCO 3 - , NO 3 - , SO 4 2 - , Na+, K+, Ca2+, Mg2+) and SiO2 was carried out to understand the geochemical processes controlling water quality. There is marked seasonal variation for almost all ions. Carbonate weathering and atmospheric precipitation are strong factors controlling the chemistry of major ion such as Ca2+, Mg2+ and HCO 3 - in the winter and summer seasons. In the monsoon season, relatively low concentrations of all the ions were observed due to dilution effects. Ca2+, Mg2+ and HCO 3 - generally account for about 70% of the total ions throughout the year. The ratios Ca2+  + Mg2+/Na+  + Cl, HCO 3 - /(Ca2+  + Mg2+), (Ca2+  + Mg2+)/ T z + , ( Na + + K + ) / T z + indicate that carbonate weathering is the main control on water chemistry in contrast to other Himalayan lakes. The low concentrations of Na+ and K+ indicate a minimal contribution from the weathering of silicate minerals to Pandoh Lake. Correlation matrix and principal component analysis were used to identify various factors influencing the ionic strength of Pandoh Lake waters. The overall water quality variation (in space and time) as well as comparison with the quality of other Himalayan lake water indicates that this lake is non polluted, however there are indications of possible enrichment of nutrients (eutrophication) in the near future in the Himalayan region of the Indian subcontinent. Hence Pandoh Lake deserves further detailed attention for management of the water resource to preserve its eco-hydrological status.

Stable isotopic compositions of waters in the karst environments of China: Climatic implications by Hong-Chun Li; Teh-Lung Ku; Dao-Xian Yuan; Nai-Jung Wan; Zhi-Bang Ma; Ping-Zhong Zhang; Miryam Bar-Matthews; Avner Ayalon; Zai-Hua Liu; Mei-Liang Zhang; Zhao-Yu Zhu; Ruo-Mei Wang (1748-1763).
A total of 117 water samples, including cave water, ground water, spring water and river water, collected from the monsoonal area of China have been analyzed for their H- and O-isotope composition. Overall, a δ 18O–δD correlation is observed of δD = −4.45 + 6.6δ 18O (R 2  = 0.90) and a significant evaporation effect observed for the southern sites. Average δ 18O and δD site values generally correspond to those of precipitation in nearby cities, with correlations of δD = 2.18 + 7.23δ 18O (R 2  = 0.95) for the sample sites and δD = 11.05 + 7.95δ 18O (R 2  = 0.95) for the cities. The effects of rainfall amount and temperature on precipitation δ 18O were calculated using a simplified theoretical model derived from the Rayleigh distillation equation, which demonstrated that the sign of δ 18Op vs. T correlation is dependent on precipitation intensity. The mean δ 18O value of cave waters exhibit decreasing trends with increasing latitude and reveal a spatial pattern of positive correlation with annual mean temperature and precipitation, mainly reflecting isotopic fractionations in the moisture source traveling from the ocean side to the inland continent. This spatial pattern implies that the δ 18O values recorded in the proxy climate records derived from speleothems might be influenced by shifts in monsoon boundary during the past, especially between glacial and interglacial intervals.

New data on major element, trace and rare-earth element (REE) contents of groundwaters and sedimentary bedrock in the Lastochka Spa in the Primorye region of far eastern Russia, together with previous stable isotope data (δD, δ 13C(TIC), and δ 18O), allow elucidation of the origin and evolution of groundwater from the spa. The sedimentary bedrock in this area is mainly highly-permeable sandstone. Dominant minerals are K-feldspar, quartz and plagioclase, and secondary minerals are calcite, limonite, sericite and kaolinite. Bedrock samples are enriched in light REE (LREE) and depleted in heavy REE (HREE) and samples where calcite fills the fractures have positive Eu anomalies. Two types of groundwater issue from spring and wells in the study area: fresh water with low mineralization (TDS up to 0.4 g/L) and high pCO2 water with high mineralization (TDS up to 4.7 g/L). Isotopic data indicate that both types of groundwater are meteoric in origin having a short residence time. Groundwaters are characterized by enrichment of HREE in comparison with LREE and a positive Eu anomaly. Mass balance calculations are consistent with albite dissolution, with or without CO2 of deep-seated mantle origin as the dominant factor controlling the chemical composition of the groundwater.

Impact of storm runoff from tropical watersheds on coastal water quality and productivity by E. Heinen De Carlo; Daniel J. Hoover; Charles W. Young; Rebecca S. Hoover; Fred T. Mackenzie (1777-1797).
Storm runoff in the steep watersheds in Hawaii leads to sediment and freshwater pulses to coastal waters that quickly affect nearshore water quality. This is particularly true in semi-enclosed embayments, such as Kaneohe Bay, Oahu, where water has a relatively long residence time compared to more open coastal areas of the islands. In this paper the authors discuss water quality and productivity in Kaneohe Bay after back-to-back rain events in late November and early December 2003, following a particularly dry summer. The short-term biogeochemical response of coastal waters and the ecosystem to runoff and physical forcing was evaluated through a combination of continuous in situ measurements and adaptive synoptic sampling carried out on a variety of temporal and spatial scales.Dissolved N:P ratios in Kaneohe Bay, which normally range from 2 to 4, consistent with a previously documented N-limited system, increase to as high as >25 during storm runoff. Order of magnitude increases in nutrients and chlorophyll in the bay shortly after the first storm and subsequent changes in the plankton community structure reflect an evolving biological response stimulated by storm inputs to the bay. Phytoplankton blooms did not draw nutrients down to limiting levels, likely due to grazing pressure by zooplankton, yet phytoplankton were not grazed to limiting levels. As a result, a slow but steady increase of the phytoplankton standing stock was observed over time.Low phosphate levels (<0.2 μM) combined with very high N:P values are typical in Kaneohe Bay waters after most storms and P often becomes the ultimate limiting nutrient. Prior to and after the November–December 2003 storms, however, dissolved P remained near or above 0.3 μM, implying that the system never became P limited and suggesting particle buffering of P concentrations. Furthermore, concentrations of NH3 became elevated (8–16 μM) following the initial storm, first in deep and subsequently in surface waters, and remained high for several months. Remineralization of organic matter transported into southern Kaneohe Bay during the storm possibly contributes nutrients that sustain phytoplankton productivity for extended periods.

The Pecos River, situated in eastern New Mexico and western Texas, receives water from a drainage area of 91 000 km2. There are primarily two major water inputs, namely snowmelt from winter storms in the headwater region of the southern Rocky Mountains and runoff from warm-season monsoonal rainfall in the lower valley. The Pecos River suffers from high levels of total dissolved solids (TDS >5000 mg L−1) under normal flow conditions. This not only poses serious problems for agricultural irrigation and safe drinking water supply, but also results in a permanent loss of biodiversity. This study examines changes in stream flow and water chemistry of the Pecos River over the last 70 a to better understand the long-term variability in stream salinity and the role of agricultural practices in salt transfer. A TDS record from the lower Pecos River near Langtry (Texas) back to 1935 was extracted to show a distinct pattern of decadal variability similar to the Pacific Decadal Oscillation (PDO), in which stream salinity is overall above average when the PDO is in positive (warm) phase and below average when the PDO is in negative (cold) phase. This is due to: (1) the dissolved salts contributed to the river are largely from dissolution of NaCl and CaSO4-bearing minerals (e.g., halite and gypsum) in the upper basin, (2) the amount of the dissolved salts that reach the lower basin is mainly determined by the stream flow yield in the upper basin and (3) the stream flow yield from the upper basin is positively correlated with the PDO index. This further attests that large-scale climatic oscillation is the major source of long-term changes in stream flow and salinity of the Pecos River. On the other hand, there is also a strong indication that the rate of salt export has been affected by reservoir operations and water diversions for agricultural practices.

It is essential to have suitable tools able to trace the fate of manure organic matter in the environment to assess whether manure disposal on the soils of catchments could affect the organic quality of rivers. Sterol compounds – mainly expressed as C29 + 28/C27 and 5β/C27 ratios – have been shown to be specific molecular tracers of pig, dairy and poultry manures in soils. The objective of this study was to measure C29 + 28/C27 and 5β/C27 ratios in five Brittany rivers (Elorn, Yar, Léguer, Min Ran and Couesnon) draining agricultural catchments receiving massive annual inputs of pig, poultry and dairy manures and compare these ratios with ratios published for enriched soils and manure samples. The particulate organic fractions from the studied rivers yielded steroid signatures typical of animal manures. More specifically, a stanol compound diagnostic of pig slurry – the 5β-stanol known as coprostanol – was found to be very widespread, with particularly high concentrations in one of the rivers (Elorn). The C29 + 28/C27 and 5β/C27 ratios of the particulate fractions of the rivers were compared with ratios measured directly in pig, dairy and poultry manure samples, as well as with the breeding activities on river catchments. These comparisons show that the steroid profiles of the five investigated rivers correlate closely with the types of manure (i.e. pig, poultry or dairy) spread on soils in their catchments. For instance, the C29 + 28/C27 and 5β/C27 ratios in the Elorn river are similar to the values typical of pig slurry (e.g. 5β/C27 > 4); compared with other catchments, the soils in this area receive by far the largest amount of pig slurry. By contrast, the Yar river drains a catchment receiving only poultry and dairy manures, and its soils exhibit C29 + 28/C27 and 5β/C27 ratios similar to those of dairy and poultry manures (e.g. 5β/C27 ∼ 1). Thus, this study indicates that the organic quality of rivers is modified in catchments where there is intense manure spreading on soils. It also provides evidence that rivers draining areas receiving different manure types may exhibit differences in the long-term evolution of their OM content. Indeed, two of the investigated river catchments receive dominantly dairy and poultry manure, and exhibit clear long-term upward trends in OM. On the other hand, one catchment receiving high proportions of pig slurry clearly shows a long-term downward trend in OM contents. A survey of the literature shows that the relative amount of OM and N received by soils in agricultural catchments could be the key parameter in determining the direction of the long-term OM trend of the river, rather than the absolute amount and/or type of manure that is applied to the soil. In any case, the present study suggests that sterol/stanol compounds may be of diagnostic value in determining whether a stream or a river is undergoing contamination by manure-derived organic matter.

In this paper, the influence of anthropogenic activity on surface water chemistry is investigated. Base flow samples from dominant land use streams in the Muskegon River Watershed, Michigan, USA, were analyzed for nutrients, major ions, and trace elements. Principal component and hierarchical cluster analysis were used to investigate the processes controlling the effects of land use on the biogeochemistry of streams in this Watershed. Biogeochemical fingerprints of land use were developed based on the results of the cluster analysis. Spider diagrams which referenced the natural environment aided the identification of the individual contributions of urban and agricultural land uses to surface water chemistry. The biogeochemical fingerprints quantified support previous findings (Urb: Na, K, Cl/Ag: Ca, Mg) and produced new insights into the effects of land use on the behavior of nutrients (Ag: N and Urb: N, P) and trace elements (Urb: V, Cr, Co, Cu, Se, Rb, Mo, Sr, Cd, Pb and Ba/Ag: U and As). The higher correlations of urban than agricultural land uses with nutrients, specifically P, were not expected and may reflect the effects of fertilization and wastewater or the season the samples were taken. More study in different geologic and urban settings is needed to help refine these fingerprints, but it is becoming apparent that despite obscuring factors, land use produces consistent, quantifiable associations between biogeochemical analytes.