Ronald W. Harvey

Role of physical heterogeneity in the interpretation of small-scale laboratory and field observations of bacteria, microbial-sized microsphere, and bromide transport through aquifer sediments

The effect of physical variability upon the relative transport behavior of microbial-sized microspheres, indigenous bacteria, and bromide was examined in field and flow-through column studies for a layered, but relatively well sorted, sandy glaciofluvial aquifer. These investigations involved repacked, sieved, and undisturbed aquifer sediments. In the field, peak abundance of labeled bacteria traveling laterally with groundwater flow 6 m downgradient from point of injection was coincident with the retarded peak of carboxylated microspheres (retardation factor, RF = 1.7) at the 8.8 m depth, but preceded the bromide peak and the retarded microsphere peak (RF = 1.5) at the 9.0 m depth. At the 9.5 m depth, the bacterial peak was coincident with both the bromide and the microsphere peaks. Although sorption appeared to be a predominant mechanism responsible for immobilization of microbial-sized microspheres in the aquifer, straining appeared to be primarily responsible for their removal in 0.6-m-long columns of repacked, unsieved aquifer sediments. The manner in which the columns were packed also affected optimal size for microsphere transport, which in one experiment was near the size of the small (∼2 μm) groundwater protozoa (flagellates). These data suggest that variability in aquifer sediment structure can be important in interpretation of both small-scale field and laboratory experiments examining microbial transport behavior.

The reversibility of virus attachment to mineral surfaces

Virus transport through groundwater is limited by attachment to mineral surfaces and inactivation. Current virus transport models do not consider the implications of the reversibility of virus attachment to minerals. To explore the reversibility of virus attachment to mineral surfaces, we attached PRD1, a bacteriophage considered to be a good model of enteric viruses, to quartz and ferric oxyhydroxide-coated quartz surfaces over a range of pH values in equilibrium “static columns.” Following attachment, we detached the viruses by replacing the pore solution with solutions of equal and higher pH. The extent of virus attachment followed an attachment “edge” that occurred at a pH value about 2.5–3.5 pH units above the pHiep of the mineral surfaces. Viruses attached below this edge were irreversibly attached until the pH of the detachment solution exceeded the pH value of the attachment edge. Viruses attached above this edge were reversibly attached. Derjaguin-Landau-Verwey-Overbeek (DLVO) potential energy calculations showed that the attachment edge occurred at the pH at which the potential energy of the primary minimum was near zero, implying that the position of the primary minimum (attractive or repulsive) controlled the equilibrium distribution of the viruses. The results suggest that the reversibility of virus attachment must be considered in virus transport models for accurate predictions of virus travel time.

Importance of closely spaced vertical sampling in delineating chemical and microbiological gradients in groundwater studies

Abstract Vertical gradients of selected chemical constituents, bacterial populations, bacterial activity and electron acceptors were investigated for an unconfined aquifer contaminated with nitrate and organic compounds on Cape Cod, Massachusetts, U.S.A. Fifteen-port multilevel sampling devices (MLSs) were installed within the contaminant plume at the source of the contamination, and at 250 and 2100 m downgradient from the source. Depth profiles of specific conductance and dissolved oxygen at the downgradient sites exhibited vertical gradients that were both steep and inversely related. Narrow zones (2–4 m thick) of high N 2 O and NH 4 + concentrations were also detected within the contaminant plume. A 27-fold change in bacterial abundance; a 35-fold change in frequency of dividing cells (FDC), an indicator of bacterial growth; a 23-fold change in 3 H-glucose uptake, a measure of heterotrophic activity; and substantial changes in overall cell morphology were evident within a 9-m vertical interval at 250 m downgradient. The existence of these gradients argues for the need for closely spaced vertical sampling in groundwater studies because small differences in the vertical placement of a well screen can lead to incorrect conclusions about the chemical and microbiological processes within an aquifer.

Transport and Fate of Microbial Pathogens in Agricultural Settings

An understanding of the transport and survival of microbial pathogens (pathogens hereafter) in agricultural settings is needed to assess the risk of pathogen contamination to water and food resources, and to develop control strategies and treatment options. However, many knowledge gaps still remain in predicting the fate and transport of pathogens in runoff water, and then through the shallow vadose zone and groundwater. A number of transport pathways, processes, factors, and mathematical models often are needed to describe pathogen fate in agricultural settings. The level of complexity is dramatically enhanced by soil heterogeneity, as well as by temporal variability in temperature, water inputs, and pathogen sources. There is substantial variability in pathogen migration pathways, leading to changes in the dominant processes that control pathogen transport over different spatial and temporal scales. For example, intense rainfall events can generate runoff and preferential flow that can rapidly transport pathogens. Pathogens that survive for extended periods of time have a greatly enhanced probability of remaining viable when subjected to such rapid-transport events. Conversely, in dry seasons, pathogen transport depends more strongly on retention at diverse environmental surfaces controlled by a multitude of coupled physical, chemical, and microbiological factors. These interactions are incompletely characterized, leading to a lack of consensus on the proper mathematical framework to model pathogen transport even at the column scale. In addition, little is known about how to quantify transport and survival parameters at the scale of agricultural fields or watersheds. This review summarizes current conceptual and quantitative models for pathogen transport and fate in agricultural settings over a wide range of spatial and temporal scales. The authors also discuss the benefits that can be realized by improved modeling, and potential treatments to mitigate the risk of waterborne disease transmission.

Humic Acid Facilitates the Transport of ARS-Labeled Hydroxyapatite Nanoparticles in Iron Oxyhydroxide-Coated Sand

Hydroxyapatite nanoparticles (nHAP) have been widely used to remediate soil and wastewater contaminated with metals and radionuclides. However, our understanding of nHAP transport and fate is limited in natural environments that exhibit significant variability in solid and solution chemistry. The transport and retention kinetics of Alizarin red S (ARS)-labeled nHAP were investigated in water-saturated packed columns that encompassed a range of humic acid concentrations (HA, 0-10 mg L(-1)), fractional surface coverage of iron oxyhydroxide coatings on sand grains (λ, 0-0.75), and pH (6.0-10.5). HA was found to have a marked effect on the electrokinetic properties of ARS-nHAP, and on the transport and retention of ARS-nHAP in granular media. The transport of ARS-nHAP was found to increase with increasing HA concentration because of enhanced colloidal stability and the reduced aggregate size. When HA = 10 mg L(-1), greater ARS-nHAP attachment occurred with increasing λ because of increased electrostatic attraction between negatively charged nanoparticles and positively charged iron oxyhydroxides, although alkaline conditions (pH 8.0 and 10.5) reversed the surface charge of the iron oxyhydroxides and therefore decreased deposition. The retention profiles of ARS-nHAP exhibited a hyperexponential shape for all test conditions, suggesting some unfavorable attachment conditions. Retarded breakthrough curves occurred in sands with iron oxyhydroxide coatings because of time-dependent occupation of favorable deposition sites. Consideration of the above effects is necessary to improve remediation efficiency of nHAP for metals and actinides in soils and subsurface environments.

STOCHASTIC ANALYSIS OF VIRUS TRANSPORT IN AQUIFERS

A large-scale model of virus transport in aquifers is derived using spectral perturbation analysis. The effects of spatial variability in aquifer hydraulic conductivity and virus transport (attachment, detachment, and inactivation) parameters on large-scale virus transport are evaluated. A stochastic mean model of virus transport is developed by linking a simple system of local-scale free-virus transport and attached-virus conservation equations from the current literature with a random-field representation of aquifer and virus transport properties. The resultant mean equations for free and attached viruses are found to differ considerably from the local-scale equations on which they are based and include effects such as a free-virus effective velocity that is a function of aquifer heterogeneity as well as virus transport parameters. Stochastic mean free-virus breakthrough curves are compared with local model output in order to observe the effects of spatial variability on mean one-dimensional virus transport in three-dimensionally heterogeneous porous media. Significant findings from this theoretical analysis include the following: (1) Stochastic model breakthrough occurs earlier than local model breakthrough, and this effect is most pronounced for the least conductive aquifers studied. (2) A high degree of aquifer heterogeneity can lead to virus breakthrough actually preceding that of a conservative tracer. (3) As the mean hydraulic conductivity is increased, the mean model shows less sensitivity to the variance of the natural-logarithm hydraulic conductivity and mean virus diameter. (4) Incorporation of a heterogeneous colloid filtration term results in higher predicted concentrations than a simple first-order adsorption term for a given mean attachment rate. (5) Incorporation of aquifer heterogeneity leads to a greater range of virus diameters for which significant breakthrough occurs. (6) The mean model is more sensitive to the inactivation rate of viruses associated with solid surfaces than to the inactivation rate of viruses in solution.

Use of PRD1 bacteriophage in groundwater viral transport, inactivation, and attachment studies

PRD1, an icosahedra-shaped, 62 nm (diameter), double-stranded DNA bacteriophage with an internal membrane, has emerged as an important model virus for studying the manner in which microorganisms are transported through a variety of groundwater environments. The popularity of this phage for use in transport studies involving geologic media is due, in part, to its relative stability over a range of temperatures and low degree of attachment in aquifer sediments. Laboratory and field investigations employing PRD1 are leading to a better understanding of viral attachment and transport behaviors in saturated geologic media and to improved methods for describing mathematically subsurface microbial transport at environmentally significant field scales. Radioisotopic labeling of PRD1 is facilitating additional information about the nature of viral interactions with solid surfaces in geologic media, the importance of iron oxide surfaces, and allowing differentiation between inactivation and attachment in field-scale tracer tests.

What Makes a Natural Clay Antibacterial

Natural clays have been used in ancient and modern medicine, but the mechanism(s) that make certain clays lethal against bacterial pathogens has not been identified. We have compared the depositional environments, mineralogies, and chemistries of clays that exhibit antibacterial effects on a broad spectrum of human pathogens including antibiotic resistant strains. Natural antibacterial clays contain nanoscale (<200 nm), illite-smectite and reduced iron phases. The role of clay minerals in the bactericidal process is to buffer the aqueous pH and oxidation state to conditions that promote Fe(2+) solubility. Chemical analyses of E. coli killed by aqueous leachates of an antibacterial clay show that intracellular concentrations of Fe and P are elevated relative to controls. Phosphorus uptake by the cells supports a regulatory role of polyphosphate or phospholipids in controlling Fe(2+). Fenton reaction products can degrade critical cell components, but we deduce that extracellular processes do not cause cell death. Rather, Fe(2+) overwhelms outer membrane regulatory proteins and is oxidized when it enters the cell, precipitating Fe(3+) and producing lethal hydroxyl radicals.

Influence of organic matter on the transport of Cryptosporidium parvum oocysts in a ferric oxyhydroxide-coated quartz sand saturated porous medium

To assess the effect of organic matter on the transport of Cryptosporidium parvum oocysts in a geochemically heterogeneous saturated porous medium, we measured the breakthrough and collision efficiencies of oocysts as a function of dissolved organic matter concentration in a flow-through column containing ferric oxyhydroxide-coated sand. We characterized the surface properties of the oocysts and ferric oxyhydroxide-coated sand using microelectrophoresis and streaming potential, respectively, and the amount of organic matter adsorbed on the ferric oxyhydroxide-coated sand as a function of the concentration of dissolved organic matter (a fulvic acid isolated from Florida Everglades water). The dissolved organic matter had no significant effect on the zeta potential of the oocysts. Low concentrations of dissolved organic matter were responsible for reversing the charge of the ferric oxyhydroxide-coated sand surface from positive to negative. The charge reversal and accumulation of negative charge on the ferric oxyhydroxide-coated sand led to increases in oocyst breakthrough and decreases in oocyst collision efficiency with increasing dissolved organic matter concentration. The increase in dissolved organic matter concentration from 0 to 20mg L(-1) resulted in a two-fold decrease in the collision efficiency.

Associations of free-living bacteria and dissolved organic compounds in a plume of contaminated groundwater

Associations of free-living bacteria (FLB) and dissolved organic contaminants in a 4-km-long plume of sewage-contaminated groundwater were investigated. Abundance of FLB in the core of the plume (as delineated by maximum specific conductance) steadily decreased in the direction of flow from a point 0.25 km downgradient from the source to the toe of the plume. At 0.25 km downgradient, FLB comprised up to 31% of the total bacterial population, but constituted <7% of the population at 2 km downgradient. Abundance of FLB correlated strongly (r = 0.80, n = 23) with total dissolved organic carbon (DOC) in contaminated groundwater between 0.64 and 2.1 km downgradient, although distributions of individual contaminants such as di-, tri- and tetrachloroethene were highly variable, and their association with FLB less clear. Numbers of FLB in the downgradient portion of the plume which is contaminated with branched-chain alkylbenzenesulfonate (ABS) surfactants were low (<5 · 108/L) in spite of relatively high levels of DOC (up to 4 mg/L). However, abundance of FLB correlated strongly with non-surfactant DOC along vertical transects through the plume. The ratio of FLB to DOC and the ratio of FLB to attached bacteria generally decreased in the direction of flow and, consequently, with the age of the organic contaminants.