Micro-scale mineralogic controls on microbial attachment to silicate surfaces: the influence of iron and phosphate mineral inclusions

Microorganisms are ubiquitous in the shallow subsurface often attached to sediment surfaces where they alter the geochemical microenvironment and mediate weathering reactions. Microbial attachment on sediment grains, however, is heterogeneous and the variability of microbial occurrence cannot be easily explained. This study examined how phosphate mineral and iron oxide inclusions in feldspars influence microbial attachment and distribution on the silicate surface using field colonization experiments in an anaerobic groundwater. Magnetite inclusions were preferentially colonized compared to the silicate groundmass while there was no discernable difference between colonization of phosphate mineral inclusions and the silicate surface. These differences are likely due to electrostatic interactions between the inclusion surfaces and cells as a function of the pH of the groundwater. In this study we used silicate surfaces that do not have surface coatings, but contain mineral inclusions of iron and phosphate minerals that possess different surface characteristics than the silicate groundmass. The surface expression of these inclusions is relatively small, ~10-50 um in diameter, but collectively they may function like macroscopic charge heterogeneity; therefore, enabling models of the inclusions with a bulk uniform electrostatic potential (Song et al. 1994). These inclusions are also potential sources of essential nutrients or terminal electron acceptors that may be otherwise scarce in solution and limiting to the indigenous microbial population. 1.2 Silicates as nutrient sources Because of their positive charge at neutral pH and their ubiquitous nature as coatings in natural sediments, the role of Fe oxides in colloid deposition has been studied extensively. Fe can be scarce in many groundwaters because of the low solubility of Fe oxyhydroxides at neutral pH but it is still necessary for microorganisms in cellular electron transport. Additionally, Fe-oxidizing bacteria derive energy from the oxidation of Fe to Fe, while facultative and obligate anaerobes derive energy from the Fe/Fe redox couple (+0.77 V) using iron as a terminal electron acceptor (TEA). Many dissimilatory iron reducing bacteria (DIRB) require contact with Fe to achieve reduction (e.g. Lovley & Phillips 1988) and may use flagella to detect and attach to Fe-rich surfaces (e.g. Caccavo & Das 2002). Few if any studies, however, have investigated the role of phosphate minerals in colloid deposition. Like the Fe oxides, phosphate minerals are sparingly soluble at neutral pH and exhibit pHzpc ranging from 6.4 to 8.5 (Stumm & Morgan 1996). Bioavailable P is commonly lacking in many subsurface environments and can diminish cell growth and metabolic efficiency. P is a fundamental macronutrient needed by microorganisms for synthesis of nucleic acids, nucleotides, phosphoproteins, and phospholipids (e.g. Ehrlich 2002), compounds used as the energy source for driving biosynthetic reactions in the cell. Because P plays a fundamental role in microbial life functions, microorganisms significantly impact the distribution and cycling of P in subsurface environments. The major reservoir for inorganic P in these geologic settings are minerals such as apatite, and microorganisms may access this mineral-bound P from both detrital (e.g. Goldstein 1986) and igneous sources (Taunton et al. 2000). We hypothesize that microorganisms will preferentially attach to surface exposures of Feor P-bearing mineral inclusions rather than to the nutrient-poor silicate groundmass in which they are included. Attachment is likely promoted by the positive surface charge of the included minerals interacting with negatively charged cell membranes. 2 EXPERIMENTAL APPROACH 2.1 Mineral characterization and preparation A suite of silicates rocks containing varying amounts of P and Fe, including anorthoclase (Wards # 46E0575), two different microclines (Wards # 46E5125, Keystone, South Dakota and Wards #46E5120, Ontario, Canada), and quartz (Wards # 46E6605) were used to investigate the role of spatial compositional heterogeneity as it relates to microbial attachment. Silicate rock specimens have been characterized previously using light microscopy, trace metal and whole rock analysis. Rocks were prepared as thin sections with four rock types on each slide. Anorthoclase, S.D. microcline, quartz, and O. microcline were mounted in ~1 cm chunks, sectioned to ~35 um and probepolished. Thin sections were utilized in all field and laboratory experiments to minimize surface roughness or microtopography as attachment variables, and to provide an ideal surface for spatial analysis of the trace and major element geochemistry (LA-ICPMS), and microbial attachment (SEM). Thin sections were analyzed for the initial spatial distribution of trace and major elements in the mineral phases via LA-ICP-MS. A ThermoElemental X7 ICP-MS was utilized coupled with Nd:YAG (266 nm) laser ablation setup which has been described in detail elsewhere (e.g. Crowe et al. 2003). Laser transects (7 micron spot size) were conducted across apatite, biotite, and magnetite phases in the host rock. Trace element concentrations in the apatite phases were calculated based on the stoichiometry of CaO in fluroapatite as an internal standard and a NIST 610 glass as an external standard. 2.2 Field colonization studies Thin sections were incubated in situ in a carbonrich, anaerobic groundwater using a stainless-steel holder that was placed in the water well and left undisturbed for twelve months. The sectioned rocks were exposed to the native microbial consortium, which is dominated by DIRB, methanogens, and fermenting bacteria (Bekins et al. 1999). The samples were retrieved and tissues were fixed in the field using a chemical critical point drying method (Vandevivere & Bevaye 1993; Nation 1983). The slides were stub-mounted, gold sputter coated and imaged on a LEO 1550 field emission scanning electron microscope at 20 kV. Samples were imaged using secondary electron (SED) and backscatter electron detectors (BSD) to optimize for imaging of attached microorganisms and mineralogy, respectively. Fe and phosphate mineral inclusions were located using BSD and inclusion composition was confirmed using EDAX and LA-ICP-MS. The inclusions and the surrounding silicate groundmass were then imaged at a range of magnifications to discern Figure 1. Reflected light photomicrograph of mineral assemblage in anorthoclase, including apatite (near center, hexagonal), magnetite, biotite, and amphibole. Scale bar is 250 μm. Figure 2. SEM photomicrograph of anorthoclase surface after reaction in anaerobic groundwater for twelve months. An apatite inclusion intergrown with magnetite is shown. Scale bar is 10 μm. changes in attachment density as a function of distance away from the inclusion. These images were then analyzed using Opti Analysis imaging software for quantification of attached cells as a function of surface mineralogy. 3 RESULTS AND DISCUSSION 3.1 Mineralogy and inclusion chemistry Quartz and the O. microcline were free of inclusions and therefore served as colonization controls for S.D. microcline and anorthoclase, both of which exhibited abundant mineral inclusions. S.D. microcline contained phosphate mineral inclusions, which occur as elongate terminating prisms. Anorthoclase contained assemblages of magnetite, biotite, amphibole, and apatites. Laser ablation of phosphate inclusions in the S.D. microcline indicate a shift away from stoichiometric apatite, with enrichment of Pb ~1,000 times chrondrite. Apatite inclusions in anorthoclase are enriched in REE, U, and Th (light REE ~1,000 times chondrite; heavy REE ~10,000 times chondrite). Preliminary analyses of the magnetite inclusions indicate they have incorporated some Zn and alkali metals. REE element enrichment is limited to the apatite. Figure 1 shows an inclusion-rich zone within anorthoclase. Smaller apatites are commonly included within amphibole while larger apatite inclusions occur in close association with magnetite inclusions. 3.2 Microbial colonization of mineral inclusions Groundwater in the study zone is anaerobic with a pH of 6.74, 3.4 mM DOC, 0.66 mM Fe, 0.74 mM CH4, and 1.2 mM Si. The water also contains 4.58 mM Ca, 1.35 mM Mg, 12.4 mM HCO3 and <0.01 mM of Al, K, Na, SO4, NO3 and PO4. After twelve months in the groundwater, only magnetite inclusions showed substantial colonization by the native microbial consortium. Magnetite inclusions are heavily colonized with an average of 2x10 cells mm compared to the silicate groundmass which supports ~2x10 cells mm, likely due to favorable electrostatic interactions between negatively charged cells and slightly positively charged oxide surface (pHzpc ~7; magnetite). The phosphate mineral and apatite inclusions, however, did not exhibit colonization densities that were significantly different from the silicate groundmass. Figure 2 shows an SEM photomicrograph of an apatite inclusion adjacent to magnetite. Few cells are visible on the apatite surface, while several cells can be seen on the magnetite. Preferential attachment to both inclusion types was expected because the groundwater pH was below that of the pHzpc of the minerals. The pHzpc values used for apatite were reported for hydroxyapatite but LA-ICP-MS analyses revealed that the inclusions are not hydroxyapatite, but rather REEenriched apatites (anorthoclase) and Pb phosphates (S.D. microcline). These compositional differences likely impact the surface charge characteristics of the minerals and it is possible that the inclusions are negatively charged at the groundwater pH. These compositional differences may also be a deterrent to microbial attachment by serving as a source of toxic elements (Pb) or radionuclides (U, Th) which inhibit microbial activity. The close association between apatite and magnetite on the anorthoclase surface (Figs. 1 & 2) presents another possible explanation for the lack of colonization on apatite. If the apatites are dissolving, which is likely with less than 1 μm dissolved PO4 in solution, then PO4 may adsorb to the proximal oxide surface. Figure 3 is a speciation diagram of phosphorus sorption on hydrous ferric oxide (HFO) as a function of pH, showing substantial sorption of PO4 in the pH range for the groundwater. The diagram was generated using the geochemical speciation program J-Chess 2.0 (van der Lee 1998) using input parameters of 10 g/L hydrous ferric oxide, 1 umol kg Figure 3. Speciation diagram of phosphorus speciation on hydrous ferric oxide surface as a function of pH. Phosphate is adsorbed onto the oxide surface in the pH range of the field conditions. PO4, surface area of 300 m/kg, and surface complexation constants from Dzombak & Morel 1990. Sorption of P to the magnetite surfaces not only alleviates the need for microorganisms to attach to phosphate minerals, but also provides this essential nutrient in an available form on the Fe oxide and therefore may increase attachment on that surface. Phosphate sorption may also catalyze apatite dissolution, especially if there is active uptake by the microbial population. Previous batch dissolution experiments of anorthoclase support the assertion that PO4 is adsorbed to Fe oxide inclusions. Substantial release of PO4 occurred only in the presence of a low affinity iron chelator (3,4 dihydroxybenzoic acid, βFe(III) = 10) and concomitant with release of substantial Fe (Rogers & Bennett 2004).