The interplay of hydrological, chemical and microbial processes in … · 2015-11-10 · Limnetica,...

Limnetica, 29 (2): x-xx (2011)Limnetica, 34 (2): 365-380 (2015)c© Asociación Ibérica de Limnología, Madrid. Spain. ISSN: 0213-8409

The interplay of hydrological, chemical and microbial processes in theformation of iron-rich floating films in aquatic environments at acircumneutral pH

Marta Reina1, M. Carmen Portillo2, Laura Serrano∗,1, Esther C.H.E.T. Lucassen3, Jan G.M.Roelofs4, Antonio Romero5 and Juan M. González2

1 Department of Plant Biology and Ecology, University of Sevilla, P.O. Box 1095, 41080-Sevilla, Spain.2 Institute of Natural Resources and Agrobiology, CSIC, Sevilla, Spain.3 B-Ware Research Centre, Radboud University Nijmegen, Nijmegen, the Netherlands.4 Environmental Biology, Department of Ecology, Radboud University Nijmegen, Nijmegen, The Netherlands.5 Department of Crystallography, Mineralogy and Agricultural Chemistry, University of Sevilla, Sevilla, Spain.

∗ Corresponding author: [email protected]

Received: 04/11/2014 Accepted: 27/04/2015

ABSTRACT

The interplay of hydrological, chemical and microbial processes in the formation of iron-rich floating films in aquaticenvironments at a circumneutral pH

The direct contribution of microbial activity to the formation of iron-oxide minerals is difficult to prove in wetlands due tothe high reactivity of solid iron phases with different compounds and the variety of redox processes that may occur at eachoxic-anoxic boundary. Here, we propose an explanation for the formation of iron-oxide films in wetlands and groundwaterseepage areas fed by sandy aquifers based on the interaction of hydrological, chemical and microbiological processes undercircumneutral conditions. The presence of a floating iron-oxide film was found to create a boundary at the air-water interfacethat maintains a suboxic and slightly acidic environment below the film compared with the environments obtained in otherfree-film wetland areas. The water trapped below this film had an average pH of 6.1, was particularly poor in O 2, HCO–

3,Na+, Ca2+, Mg2+, K+, and Tot-S, and has high concentrations of Tot-P, Tot-Fe, NH+

4 and Zn. The formation of a floatingiron-oxide film was reproduced under anaerobic conditions after progressive enrichment through the incubation of naturalsediment samples in the laboratory. Heterotrophic bacteria belonging to the genus Enterobacter were the dominant bacteriain the enrichments that resulted in the formation of a floating iron-oxide film. The X-ray diffraction patterns showed thatthe presence of two-line ferrihydrite was common to the iron-oxide films collected in both the natural environment and thelaboratory cultures, whereas other iron-oxides (goethite and low-crystalline lepidocrocite) were observed only in the naturalenvironment. This study highlights the role of ubiquitous bacteria, which are generally considered unimportant participants iniron-transformation processes in the environment, and the contribution of both biological and non-biological processes to ironoxidation in natural systems under circumneutral conditions.

Key words: Iron surface complexes, heterotrophic bacteria, groundwater, Doñana.

RESUMEN

La interacción de procesos hidrológicos, químicos y microbiológicos en la formación de películas flotantes ricas en hierroen ambientes acuáticos de pH circumneutro

En los humedales, es difícil probar que la actividad microbiana sea la responsable de la formación de óxido de hierro mineraldebido, tanto a la gran reactividad del hierro en fase sólida con diferentes sustancias, como a la variedad de procesos redoxque pueden ocurrir en cada interfase óxica-anóxica. El presente trabajo propone una explicación, basada en la interacciónde procesos hidrológicos, químicos y microbiológicos en condiciones circumneutras, para explicar la formación de un filmde óxido de hierro en humedales y manaderos donde aflora agua subterránea proveniente de acuíferos de arenas silíceas.Además, la presencia de un film de óxido de hierro que flota sobre la interfase agua-aire genera condiciones subóxicas yligeramente ácidas en el agua que queda atrapada debajo, y que son muy distintas a otras zonas libres de film en el mismo

16159_Limnetica 34(2), pàgina 365, 30/10/2015

366 Reina et al.

humedal. Este agua atrapada bajo el film se caracterizó por presentar un pH medio de 6.1, una menor concentración de O 2,HCO–

3, Na+, Ca2+, Mg2+, K+, y S total, pero una mayor riqueza en P total, Fe total, NH+4 y Zn. La formación de un film

flotante de óxido de hierro se reprodujo en el laboratorio, en condiciones anaeróbicas, tras el enriquecimiento progresivode las muestras del sedimento natural que habían sido incubadas. En dichos enriquecimientos, donde se produjeron filmsflotantes de óxido de hierro, la bacteria dominante perteneció al género Enterobacter. Mediante difracción por rayos X, seencontró ferrihidrita con estructura en doble cadena, tanto en el film de muestras naturales como de cultivos de laboratorio.Además se encontraron otros tipos de óxidos minerales (goetita y lepidocrocita de pobre cristalización) sólo en las muestrasnaturales de film. El presente estudio muestra la relevancia de bacterias ubicuas, hasta ahora consideradas sin importancia enprocesos naturales de transformación del hierro, y la participación tanto de procesos bióticos como abióticos en la oxidacióndel hierro en sistemas naturales sometidos a condiciones circumneutras.

Palabras clave: Hierro acomplejado en superficies, bacterias heterótrofas, agua subterránea, Doñana.

INTRODUCTION

Iron-cycling is a central issue in aquatic systems.Iron not only is essential for living organisms butalso plays an important role in the redox transfor-mations of numerous compounds, the bioavail-ability of nutrients and trace metals, and the bind-ing processes of organic and inorganic ligands.The rate of the oxidation of ferrous iron by O2

as a function of pH and other solution variablesis well known (Stumm & Sulzberger, 1992), al-though a large variety of iron oxides, hydrox-ides and oxyhydroxides can be formed due to thehigh reactivity of solid iron phases with differentcompounds at different oxic-anoxic boundaries.Hence, more studies are needed to understandthe mechanism underlying the formation of iron-oxide films and their biogeochemical role (Klejaet al., 2012). Despite the wide-spread presence ofiron-oxide films floating on the water surface ofwetlands all over the world, these films are of-ten overlooked, mistaken for oil due to their oilyappearance, or misidentified as biofilms strictlygenerated by biological activity (Grathoff et al.,2007). Therefore, the relative contribution of mi-croorganisms in the formation of ferrous/ferriciron complexes and minerals in non-acidic en-vironments has not yet been quantified becausethe oxidation of ferrous iron can also proceedalong purely chemical routes (Châtellier et al.,2004; Trolard, 2006). In this sense, Rentz et al.(2007) used cyclic voltammetry in short experi-mental times (<30 min) to capture the rapid ki-

netics of the natural iron-oxidation rates undercircumneutral conditions and evidenced that mi-croaerobic conditions alone can account for theoxidation of ferrous iron even in the presenceof iron-oxidizing bacteria. Nevertheless, bacteriacan play both a metabolic and non-metabolic rolein iron oxidation because ferrous iron can adsorbreversibly onto bacterial cell surfaces and reachan adsorption equilibrium within minutes underanoxic conditions at an effective pH range of 3to 7 (Châtellier & Fortin, 2004). This type of bi-otic surface-mediated iron complexation will, inturn, promote abiotic iron oxidation on the cellsurface in oxic systems. Furthermore, recent find-ings suggest that biotic and abiotic iron oxidationco-occur at the same temporal and spatial scalesin natural systems and thus cannot be separatedbecause even biotically oxidised ferrous iron willserve as a nucleation matrix for abiotic ferric ironprecipitation at circumneutral pH (Ionescu et al.,2015).

The significance of microbial metabolism inthe weathering and formation of iron mineralsis widely recognized. Specific microorganisms,such as iron-oxidizing bacteria, iron-reducingbacteria and bacteria forming iron-complexingagents, mediate major transformations thatchange iron from a soluble to an insoluble formunder different redox and pH conditions (Berthe-lin et al., 2006). Studies of 16S rRNA genesequences have indicated that virtually everymajor group of prokaryotes can be associatedwith aerobic ferrous iron oxidation in natural


Iron-oxide films 367

environments under both acidic and neutrophilicconditions (Neubauer et al., 2002; James &Ferris, 2004, Sobolev & Roden, 2004).

Wetlands may provide a suitable environmentfor determining the extent to which microorgan-isms actively mediate the oxidation of dissolvedferrous iron at a variety of aerobic/anaerobic in-terfaces on the water surface, sediment layers andrhizosphere of aquatic macrophytes (Emerson& Weiss, 2004). Wetlands and bogs fed by thegroundwater seepage of sandy aquifers providethe perfect neutrophilic environment for studyingthis process because the alternation of the redoxpotential due to fluctuations in the water tableproduces heterogeneous accumulations of ironoxides and hydroxides, which eventually crystal-lize due to further diagenetic processes (Knappet al., 1998). In this type of environment, ironcan be transported several kilometres by the flowof groundwater through a combined sequence ofbiotic and abiotic processes that often leads to amodification of the mineralogical compositionof the original iron oxides (Herbillon, 2006).The oxidation of dissolved ferrous iron, whichoccurs as soon as the seepage water reachesthe air surface, is a common feature of theseaquatic environments that generally leads to theprecipitation of lepidocrocite at neutral pH andferrihydrite in the presence of aqueous silicaat redox boundaries and surfaces (Châtellieret al., 2004), and the crystallization of theseprecipitates will lead to the formation of goethiteand hematite (Cornell & Schwertmann, 1996).

The Iberian Pyrite Belt (SW Spain) is oneof the largest massive sulphide deposits in theworld. It was formed in the early Carboniferousand later transformed during the HercynianOrogeny. As a consequence of pyrite oxidationsince mining activity was initiated in this areaapproximately 5000 years ago, the Tinto andOdiel rivers are now acidic (pH< 3) and havedrained large amounts of sulphate and dissolvedmetals (Fe, Al, Mn, Cu, Zn, Cd, Pb) to theGulf of Cádiz (Davis et al., 2000). The Doñanaaquifer (∼ 3400 km2) borders the Tinto estuaryand extends westward to the Guadiamar andGuadalquivir rivers (Llamas, 1990). The aquiferis bottomed by impermeable marine blue marls

of Miocene age upon which sands, silts, andsilty-clay materials of estuarine, fluvial andeolian origins were later deposited. It holds anunconfined aquifer with a shallow water tableand several flow systems that feeds hundreds ofsmall ponds, wetlands and seepage areas (Ser-rano et al., 2006). A discontinuous iron hardpancomposed of goethite (Siljeström & Clemente1990) is commonly found at a soil depth ofless than 1.5 m, coincident with former ground-water seepage areas that have become buriedby younger dune generations (Muñoz-Reinoso,2001). Despite its complexity, the hydrologyof some of these wetlands is well established(Sacks et al., 1992), and given the presence ofshallow iron oxide deposits, these wetlands forman ideal environment for studying the formationof iron-oxide films. The present work describesthe environmental, microbiological and min-eralogical features of floating iron-oxide filmsat a circumneutral pH in several groundwaterseepage areas of the Doñana National Park (SWSpain). It also records the formation of a similarfilm in the laboratory after the enrichment andincubation of natural sediment samples, whichraises several questions: a) whether the forma-tion of this floating iron-oxide film is activelymediated by microorganisms, b) the identityof the microorganisms that are involved, andc) whether the process is mediated by light.Finally, a model that takes into account some ofthe hydrological, chemical and microbiologicalfeatures of one of the study sites is proposed toexplain both the formation of floating iron-oxidefilms on the SW shoreline of the study site ofinterest and the lack of a film on the opposite NEshore.

MATERIAL AND METHODS

Site description

The Doñana region (37◦N, 6◦W) extends alongthe coastal plain of the Gulf of Cádiz from theestuary of the Tinto River to the left bank of theGuadalquivir River and inland from the coast-line to the northern uplands bordering the Iberian


368 Reina et al.

Pyrite Belt (Fig. 1). This region has a Mediter-ranean climate with Atlantic influence and is gen-erally classified as dry subhumid. Doñana in-cludes several territories with different degreesof environmental protection covering more than100 000 ha. The Doñana National Park covershalf of this extension and is the most importantwetland area in Spain. It exhibits an extraordi-nary variety of aquatic systems, which are clas-sified into two broad groups according to the

hydro-geomorphology of their basins: the sanddunes and the marsh floodplain (Serrano et al.,2006).

Sample collection and environmental analyses

From October 2003 to June 2006, water andsediment samples were collected from severaltemporary ponds in the Doñana BiologicalReserve where groundwater seepage areas are

Figure 1. A: Map of the Doñana region within the Iberian Peninsula showing the location of the Iberian Pyrite Belt and the Doñanaaquifer. B: Detailed map of the study area with the temporary ponds. A: Mapa de Doñana en la Península Ibérica indicando laslocalizaciones de la Faja Pirítica Ibérica y el acuífero de Doñana. B: mapa detallado del área de estudio con las lagunas temporales.



common (Fig. 1): Las Verdes, Dulce, SantaOlalla, Zahíllo, and Taraje ponds. Water was col-lected from the surface with a 1-litre sample jar,and the top sediment was sampled with a handgauge. Additionally, Dulce pond was sampled atdistances of 0, 1 and 3 m to the SW shorelineand to the NE shoreline (three replicates at eachsite) in June 2006. The electrical conductivity(compensated at 25 ◦C), concentration of dis-solved O2 and pH in the water were measuredin situ with the corresponding electrodes. Threereplicate samples of the floating iron-oxidefilm covering the shoreline were also collectedusing a sterilized spatula, and each portion wasplaced in a sterile container. Pore-water sampleswere collected in situ to measure the sedimentpH in the laboratory. These were collectedanaerobically using Teflon rhizons (EijkelkampAqrisearch Equipment, The Netherlands) con-nected to vacuum serum bottles (2-4 replicates).The surface- and pore-water samples weredivided into two parts. One part was fixed with1 % nitric acid and analysed for Ca, Mg, Fe, Al,P, S, Si, Na, K and Zn using an Inductively Cou-pled Plasma Mass Spectrophotometer (ICP-MSX-series, Thermo, Waltham, MA, USA). Qualityassurance measures included blanks, replicateanalyses and matrix spikes. The recoveries frommatrix spikes ranged from 95 % to 107 %. Therepeated analyses did not reveal differencesgreater than 5 %. The other part of the sampleswas fixed with 0.125 g/L citric acid and analysedfor NO−3 (Kamphake et al., 1967) and NH+4(Grasshoff & Johansen, 1972) using an AutoAnalyser (AA 3, Bran + Luebbe, Norderstedt,Germany). The CO2 and HCO−3 concentrationswere analysed using a 0525 HR infrared carbonanalyser (Oceanography International). The con-centration of organic matter in the top sedimentsamples was estimated in 3-4 replicates by losson ignition (450 ◦C, 5 h). The samples of waterand ignited sediment were digested with 0.5 MH2SO4 and K2S2O8 at 120 ◦C. Then, the Tot-Pwas analysed as i-P (Murphy & Riley, 1962),and the Tot-Fe was analysed as dissolved ferrousiron using o-phenanthroline and ascorbic acidas the reducing agent (Golterman, 2004). TheP-fractionation of the sediment samples was per-

formed in 2-4 replicates with 10 mL of sedimentsuspension according to the EDTA method forsequential extraction (Golterman, 2004).

Mineralogical analysis

The mineralogical analysis was performed atthe Research General Services of the Universityof Seville (CITIUS) through X-ray diffraction(XRD) using a Bruker D8 Advance powder de-vice with CuKα radiation and a monochromator.Samples of the iron-oxide film collected in thenatural environment and the laboratory cultureswere analysed. Each film sample was powderedwith an agate mortar and scanned from 3 to 80◦

2θ (double theta) at a 0.02◦ step size and with a10-s count time per step.

Sediment profiles with microelectrodes

Measurements of the oxygen and sulphide con-centrations and pH within sediments of bothsides of the studied pond were performed usingmicroelectrodes attached to a motor-driven mi-cromanipulator (Unisense, Århus, Denmark)following previously described methods (Revs-bech, 2005). Measurements were collected every200 µ m for a depth of approximately 5 cm.Microelectrode calibrations were performed aspreviously described (Revsbech, 2005) followingthe manufacturer’s recommendations.

Enrichments from natural communities

The upper layer of sediment (0-2 cm) and shal-low water were sampled and incubated at 28 ◦Cunder aerobic and anaerobic conditions with il-lumination. Sterilized controls were conducted.If an iron-oxide film was formed, an aliquot ofthe enrichment was transferred to G medium.This transfer to G medium resulted in bacterialgrowth, which persisted through repeated cultiva-tions, always producing floating iron-oxide filmson the surface of the cultures. G medium (Lovleyet al., 1993) was composed (per litre) of 13.7 gFe(III) citrate, 2.5 g NaHCO3, 1.5 g NH4Cl, 0.6 gNaH2PO4, 0.1 g KCl, 2.5 g sodium acetate, and0.25 g Na2WO4 × 2H2O with 10 mL of trace el-


370 Reina et al.

Table 1. Average dissolved oxygen (mg/L) and chemical concentrations (µM) of surface water samples (n = 2 – 8) collected atdistances of 0, 1 and 3 m from the shoreline and of pore water samples collected at different depths in the sediment profiles ateach shoreline (NE and SW). The significant differences (one-way ANOVA + Holm-Sidak test, p < 0.05) between the surface watersamples collected from the SW shoreline are noted as follows: ab between the surface water samples collected at distances of 0 and1 m, ac between the surface water samples collected at distances of 0 and 3 m, ad between the surface water sample collected at adistance of 0 m and the pore water samples collected at depths of 0-2 cm, ae between the surface water sample collected at a distanceof 0 m and the pore water samples collected at depths of 2-5 cm, af between the surface water samples collected at a distance of 0m and the pore water samples collected at depths of 5-10 cm. In addition, significant differences between the different shorelines arenoted as follows: aa between the samples collected at a distance of 0 m from the SW shoreline and the samples collected at a distanceof 0 m from the NE shoreline. Concentraciones promedio de oxígeno disuelto (mg/L) y de disitntas sustancias químicas (μM) enlas muestras de agua superficial (n = 2 – 4) recogidas a 0, 1 y 3 m de distancia desde la orilla, y en el agua intersticial a diferentesprofundidades en los perfiles de sedimento de cada orilla (SW y NE). Se indican las diferencias significativas (ANOVA de una vía +test Holm-Sidak, p < 0.05) entre distintas muestras de agua superficial de la orilla SW de la siguiente forma: ab entre 0 y 1 m en aguasuperficial, ac entre 0 y 3 m en agua superficial, ad entre el agua superficial a 0 m y el agua intersticial a 0-2 cm de profundidad, ae

entre el agua superficial a 0 m y el agua intersticial a 2-5 cm, af entre el agua superficial a 0 m y el agua intersticial a 5-10 cm; asícomo aa entre el agua superficial de ambas orillas a 0 m.

surface water pore water at 0 m

O2 (mg/L) NESW

pH NESW

Cl− NESW

HCO−

3 NESW

Tot-S NESW

NO−

3 NESW

Tot-P NESW

Tot-Na+ NESW

Ca2+ NESW

Mg2+ NESW

Tot-Fe NESW

K+ NESW

NH+

4 NESW

Fe2+ NESW

Tot-Mn NESW

Zn NESW

0 m 1 m 3 m

6.90.2aa

7.26.1aa ab ac, ,

10712.11629.3aa ac,

2001.2704.5aa ab ad ae af, , , ,

981.048.4aa ac,

22.811.9aa ac af, ,

5.59.5ad ae af, ,

12156.01992.2aa ab ac, ,

930.1446.3aa ab ac ad ae af, , , , ,

1693.3249.6aa ab ac, ,

3.7617.8aa ac ad ae, , ,

210.778.7aa ac,

16.578.1ad ae af, ,

0.027.4aa ac,

14.613.2

0.12.1aa ab ac, ,

8.11.4

7.76.6

9305.92158.8

1614.21294.4

939.2132.6

20.212.5

6.69.2

11760.93065.4

848.6646.2

1631.1531.0

2.7462.3

196.1113.6

14.964.2

0.029.2

1.633.6

0.00.0

10.94.6

8.48.3

9623.68064.2

1632.71354.7

942.9918.8

22.018.1

3.24.4

11304.311022.8

828.0892.8

1591.01613.4

3.013.3

201.0165.8

13.214.3

0.00.0

0.71.5

0.00.0

0-2 cm 2-5 cm 5-10 cm

——

7.36.1

10620.71290.2

2701.91344.2

723.338.4

22.814.7

11.116.8

12094.61229.2

977.0596.0

1663.0306.5

19.1964.5

224.783.5

20.0340.1

——

54.614.4

0.32.3

——

7.26.1

10824.41235.5

2804.31307.5

594.034.0

20.513.5

14.428.4

12137.51185.5

951.1581.1

1649.3296.6

44.91078.9

249.880.8

19.5290.7

——

51.314.1

0.33.1

——

7.06.2

10148.41277.7

2261.01351.6

496.033.5

20.215.7

12.024.3

11552.21260.1

959.3590.4

1343.2304.9

654.8898.6

244.786.2

14.9249.6

——

61.614.3

0.53.0



ement solution and 10 mL of vitamin solution(Balch et al., 1979). The medium pH was ad-justed to 6.0, and the culture was incubated anaer-obically under 80 % N2 + 20 % CO2 gas. Bacte-rial films were collected using a sterile spatulaand containers. Once dried (24 hours at 70 ◦C),the films were digested with 4 mL of nitric acid(65 %) and 1 mL of hydrogen peroxide (35 %)using an ETHOS-D microwave labstation (Mile-stone, Sorisole, Italy). The digests were diluted(25×) and analysed with an ICP-MS as above.

Molecular analyses

Natural samples of water and sediment from theSW and NE pond shores were analysed for RNAto determine the metabolically active fraction ofthe bacterial community. Laboratory enrichmentswere analysed based on their DNA to identifythe bacteria forming these cultures. DNA andRNA were extracted using the Nucleospin FoodDNA extraction kit (Macherey-Nagel, Düren,Germany) and the RNAqueous4PCR Total RNAextraction kit (Ambion Inc., Austin, USA), re-spectively. The protocol for total RNA extractionincluded DNaseI treatment (37 ◦C for 1 h) toremove any DNA remaining in the final RNAextract.

A reverse transcriptase reaction was per-formed to obtain the complementary DNA(cDNA) to the RNA corresponding to the 16SrRNA genes to be amplified. The ThermoScriptreverse transcriptase (Invitrogen, Carlsbad, CA,USA) was used in this study with a 16S rRNAgene-specific primer, namely 518R (5′-ATT ACCGCG GCT GCT GG (Muyzer et al., 1993)), at anannealing temperature of 55 ◦C for 1 h. Controlslacking reverse transcriptase were also per-formed to check for the presence of DNA tracesin the extracted RNA. A standard amplificationreaction by PCR was performed after cDNApreparation. Amplification of 16S rRNA frag-ments from the cDNA was performed by PCRusing the forward primer 27bF (5′-AGA GTTTGA TYM TGG CTC AG (Portillo et al., 2008))and the reverse primer 518R. ExTaq (Takara,Shiga, Japan) was used as the DNA polymerasefor PCR, following the manufacturer’s recom-

mendations. The thermal conditions for theamplification reaction consisted of the followingsteps: 95 ◦C for 2 min; 30 cycles of 95 ◦C for15 s, 55 ◦C for 15 s and 72 ◦C for 1 min; and afinal incubation at 72 ◦C for 10 min. Amplifica-tions from DNA extracted from the enrichmentswere performed following a similar protocolusing the primers 27bF and 907R (5′-CCC CGTCAA TTC ATT TGA GTT T Lane, 1991).

DNA libraries were constructed from the am-plification products obtained from the DNA ex-tracted from the cultivated enrichments and fromthe cDNA prepared from reverse-transcribedRNA extracted from the natural samples. TheDNA libraries were constructed from the PCRproducts using the TA-TOPO cloning kit (Invit-rogen, CA, USA). The DNA library constructionand screening was performed as described byGonzález et al. (2003). Selected clones weresequenced. The sequences were manually editedand processed with the BLAST algorithm(Altschul et al., 1990) to search for the closestrelatives to the retrieved sequences using theGenBank database at NCBI (http://www.ncbi.nlm.nih.gov/blast). The sequences were checkedfor chimeras using the Ccode program as de-scribed by González et al. (2005). The bacteriaidentified based on RNA corresponded to themetabolically active fraction of the total micro-bial community because the RNA per bacterialcell is quantitatively proportional to cell growth(Mills et al., 2004). The microbial communitiesfrom the natural samples and the enrichmentswere characterized by molecular fingerprinting.The fingerprints were prepared using DGGE(Denaturing Gradient Gel Electrophoresis) aspreviously described (Muyzer et al., 1993) withthe modifications introduced by Portillo andGonzález (2008) to obtain relative quantitativefingerprints. To determine the grade of coverageof the detected sequences with respect to thebacterial community, rarefaction curves wereconstructed according to Hughes and Hellmann(2005). These curves represent the number ofprocessed clones against both the number ofdetected OTUs (Operational Taxonomic Units)and the number of different phyla detected in thisstudy. For practical purposes, we considered an


372 Reina et al.

OTU as a set of apparently identical sequencesshowing higher than 97 % identity (Huber et al.,2007).

Statistical analyses

We used one-way ANOVA followed by theHolm-Sidak method to examine the differencesin the concentration of the measured variablesbetween the site with the floating iron-oxide filmand the rest of the sites. The statistical analyseswere conducted using the SigmaPlot for Win-dows package (v 11.0). Significant differencesbetween the 16S rRNA gene fingerprints of thenatural bacterial communities were estimatedusing the fingshuf software according to Portilloand González (2008).

RESULTS

The top sediment samples of the groundwaterseepage areas analysed in this study were gener-ally rich in Tot-Fe because concentrations higher

than 180 µ mol/g d.w. were recorded in 10 out of12 sites. The organic matter and Tot-P concentra-tions in the sediment ranged widely (1.2-34.8 %and 2.3-121.7 µ mol/g d.w., respectively) andwere significantly correlated with each other(r = 0.739, p < 0.01). The concentrations ofTot-P and Tot-Fe in the sediment also showed asignificant correlation (r = 0.760, p < 0.01). Atthe sites where the sediment organic matter concen-tration was higher than 9 %, dissolved ferrous iron(Fe2+) was detected in the sediment pore water ata concentration ranging from 10.7 to 159.4µM.

The highest concentration of sediment Tot-Fe(1420 µ mol/g d.w.) was found on a narrow fringe(< 0.5 m) along the SW shoreline of Dulce pondcovered by patches of a fine floating metallic film(with a thickness < 1 mm). Its chemical compo-sition revealed a predominance of Fe (87.0 % ofdry weight in the digested sample) with tracesof other metals, such as Mn (0.32 %). The wa-ter collected at a distance of 1 m from the SWshoreline toward the centre of the pond did notcontain this floating iron-oxide film, but the pre-cipitation of colloidal Fe was visible in the water

Figure 2. Profiles of H2S (black line) and O2 (grey line) concentrations and pH (short-dashed black line) in sediment cores collectedfrom the SW and NE shores of Dulce pond (A and B, respectively). The air-water and water-sediment interfaces are indicated bylong-dashed horizontal lines. Perfiles de las concentraciones de H 2S (línea negra) y O2 (línea gris), y el pH (línea negra punteada)en muestras de sedimento tomadas de las orillas SW y NE de la laguna Dulce (A y B, respectivamente). Las interfases aire-agua yagua-sedimento se indican mediante líneas de trazos horizontales.



column. On the opposite NE shoreline, no iron-oxide film or colloidal Fe was visible, and theconcentration of Tot-Fe on the top sediment wasonly 22 µ mol/g d.w.

A detailed characterization of several en-vironmental features of both the SW and NEshorelines was then performed in June 2005(Table 1). The surface water trapped below thefloating iron-oxide film observed at a distanceof 0 m from the SW shoreline exhibited severalparticular features compared with the othersurface and pore water samples, and many ofthese differences were significant (Table 1). Thewater below this film showed the lowest pHand concentrations of O2, HCO−3 , Na+, Ca2+,Mg2+, K+, and Tot-S compared with all of theother surface sites and it presented the highestconcentration of Tot-P, Tot-Fe, NH+4 and Zn.Additionally, the concentrations of HCO−3 , NO−3 ,NH+4 , Ca2+ and Tot-P were significantly lowerin the surface water below this film than inthe deepest pore water samples collected onthe SW side (Table 1). The surface sedimentbelow the film was relatively rich in organicmatter (15.1 %), Tot-Fe and Tot-P (254.3 and29.9 µ mol/g d.w., respectively).

The profiles of the O2 concentration in a sedi-ment core determined by microelectrodes showedan initial decline of O2 just below the air-waterinterface and a second decline at the sedi-ment-water interface that resulted in the rapiddecrease in the concentration of O2 with increas-ing depth (Fig. 2). The H2S concentration showeda relatively constant profile with a low H2S pro-duction at the SW shore and a sharp H2S produc-tion layer at the NE shore just below the depth,

where O2 was depleted. The depth profiles of pHshowed relatively constant values at depths withmeasurable O2 concentrations and a slight acid-ification at increasing depths in the anoxic zoneof both the SW and NE shorelines (Fig. 2).

Anaerobic incubations of water and sedimentsamples collected from the natural environ-ment resulted in visible bacterial growth, asjudged by a turbidity analysis and phase-contrastmicroscopy. Floating iron-oxide films were ob-served in anaerobically incubated samples (Ta-ble 2). Aerobically incubated samples and anysample maintained in darkness did not result iniron-oxide film formation. Cultured enrichmentsobtained in G medium and incubated underanaerobic conditions (either in N2 or H2:CO2,80 %:20 %, atmosphere) showed both bacterialgrowth and formation of a floating iron-oxidefilm on the medium surface. During growth, achange in coloration of the medium could beobserved. The original colour of the medium wasdark brown, this colour cleared over the incu-bation to become transparent, and a black filmwas eventually observed floating on the surfaceof the cultures. After exposure to oxygen (airatmosphere), the colour of these floating filmsturned to ochre, and the films presented a similarappearance to natural iron-oxide films in thenatural environment. Further incubation underillumination conditions with exposure to oxygenresulted in a thickening of the film, and anaerobicconditions persisted under the film (Table 2).

A molecular survey of the microorganismspresent in the laboratory cultures was performed.The major bacterial components were detectedby DGGE profiling and identified through clone

Table 2. Formation of a floating film in laboratory incubations of the original water/sediment samples under different conditionsand their respective controls. Repeated G-medium enrichments where only performed for those incubations in which floating filmsformed. Formación del film flotante en las incubaciones de laboratorio provenientes de las muestras de agua/sedimento bajo diversascondiciones, y sus respectivos controles.

Anaerobic conditions Aerobic conditions O2 exposure afteranaerobic conditionsLight Darkness Light Darkness

Original sample incubations YES NO NO NO YES 2

G-medium enrichments YES NO 1 — — YES2

Sterilized controls NO NO NO NO NO

1Very poor film formation; 2It turns to ochre and thickens.


374 Reina et al.

Figure 3. A: DGGE fingerprints based on DNA extractedfrom the Fe-film formed through cultivated enrichment. B:DGGE fingerprints based on RNA extracted from the naturalcommunities of the SW and NE shores of Dulce pond. Thenumbered arrows indicate the migration of the different phy-lotypes identified in the cultivated enrichment: 1, Enterobacter;2, Clostridium, and 3, Anaeroarcus. The unnumbered arrow-heads indicate the location of migration markers correspondingto (from top to bottom) Pseudomonas aeruginosa, Escherichiacoli, Paenibacillus sp., and Streptomyces caviscabies. A: PerfilDGGE de los cultivos enriquecidos formadores de la películade Fe. B: Perfiles duplicados de las comunidades naturales dela orilla SW y NE de la laguna Dulce. Las flechas con númerocorresponden a la migración de los distintos filotipos identifi-cados en los cultivos enriquecidos: 1, Enterobacter; 2, Clostrid-ium, y 3, Anaeroarcus. Las puntas de flechas sin número corre-sponden a la migración de los siguientes marcadores (de arribaa abajo): Pseudomonas aeruginosa, Escherichia coli, Paeni-bacillus sp., y Streptomyces caviscabies.

screening and sequencing from the constructedDNA libraries (Fig. 3). Bacteria belonging to thegenus Enterobacter were the major representa-tives of the cultured enrichments forming thefloating iron-oxide films and represented only aminor component of the natural communities.This genus was present as several phylotypesshowing differential migration on the DGGE fin-gerprints. Firmicutes, which are closely relatedto the genera Anaeroarcus and Clostridium, werealso detected at lower abundances in the DNA li-braries and the DGGE analysis results.

The metabolically active microbial communi-ties from the SW and NE shores were analysedthrough DGGE fingerprinting and constructionof 16S rRNA gene libraries and sequencing.A total of 252 clones were analysed (131 and121 clones from the SW and NE shores, re-spectively), and the results revealed 90 differentOTUs (46 and 44 OTUs from the SW and NEshores, respectively). Among the microorgan-isms identified from these samples, typicaliron-reducing and oxidizing specialist bacteriawere scarce (1.6 % of the total number of clones)and thus represented a minor metabolicallyactive component of the microbial communitiesfrom both of the two studied shores. Thesepotentially iron-reducing bacteria were detectedat both shores and were related to the generaLeptothrix and Gallionella (Betaproteobacte-ria) and Geobacter (Deltaproteobacteria). Thephylum distribution of the metabolically activemicroorganisms detected by sequencing of the16S rRNA gene libraries is shown in Table 3.The nucleotide sequences obtained in this studycan be accessed under numbers from JN699913to JN700019. The major functional compo-nent of these communities were phototrophicmicroorganisms (approximately 29 % of pro-cessed clones), including both eukaryotic algae(chlorophytes and diatoms) and cyanobacteria.Cyanobacteria were present at a lower proportionin the metabolically active community at the SW(10.7 %) compared with the NE (38.0 %) shore.The presence of Betaproteobacteria (Leptothrixand Gallionella) among the metabolically activemicroorganisms was detected at a higher pro-portion in the film-covered SW shore (35.9 %)than in the NE shore (17.4 %), and most of theclones showed highest homology to sequencesfrom uncultured bacteria.

The level of coverage achieved during theanalysis of the 16S rRNA gene libraries per-formed in this study represented a fraction of0.75-0.79. The rarefaction curves presented forthe SW and NE shores, which were estimatedat 97 % similarity for OTU differentiation, in-dicated similar levels of diversity at both sidesof the pond. When phyla were considered, therarefaction curves showed that most of the major



Table 3. Proportion of metabolically active microorganisms classified into different phyla that were detected at the SW and NEshores of Dulce pond. Proporción de microorganismos metabólicamente activos clasificados en distintos filos que fueron detectaronen las orillas SW y NE de la laguna Dulce.

Percentage of total processed clonesPhylum SW NE Total

Eukaryotic algae 29.8 28.9 29.4Betaproteobacteria 35.9 17.4 27.0Cyanobacteria 10.7 38.0 23.8Deltaproteobacteria 5.3 3.3 4.4Bacteroidetes 3.0 5.8 4.4Alphaproteobacteria 2.3 4.1 3.2Verrucomicrobia 5.3 0.0 2.8Fusobacteria 3.0 0.0 1.6Chloroflexi 0.8 1.6 1.2Spirochaete 1.5 0.0 0.8Gammaproteobacteria 1.5 0.0 0.8Chlorobi 0.8 0.0 0.4Firmicutes 0.0 0.8 0.4

taxonomic categories present at these sites weredetected. A higher number of bacterial phylawere detected in the film-covered side of the pond(12 phyla) than at the film-free side (9 phyla).

The chemical composition of the natural float-ing film revealed a dominance of Fe as the mostabundant component (87 % of total). The propor-tion of Fe after digestion in the incubated filmwas very similar to that obtained in the natu-ral film (Table 4). Si was the second most com-mon element in the natural film, as expected bythe quartzitic nature of these sandy soils (average

concentration of Si in the water of the ground-water seepage area was 324 µ M). In contrast,in the film obtained at the laboratory, Na+ wasthe second most common element in abundance,probably due to the use of sodium salts in themedium composition. The X-ray diffraction pat-terns showed that the presence of two-line ferri-hydrite was common to the iron-oxide films col-lected in both the natural environment and thelaboratory cultures, whereas other oxides, such asgoethite and low-crystalline lepidocrocite, weredetected only in the natural environment.

Table 4. Concentrations of main elements after digestion of the film incubated in the laboratory and the natural film collected fromthe study pond. Concentraciones de los principales elementos tras la digestión del film obtenido tras incubación en el laboratorio, ydel film natural recogido en la laguna estudiada.

incubated film natural film

µ mol/g d.w. (%) µ mol/g d.w. (%)

Fe 1476.8 79.24 6218.8 87.00Na 178.3 9.56 129.1 1.81P 65.4 3.51 67.4 0.94Si 53.4 2.87 293.9 4.11S 33.2 1.78 96.1 1.34Ca 26.6 1.43 179.2 2.51Al 10.3 0.56 20.1 0.28K 6.5 0.35 33.1 0.46Mg 3.9 0.21 74.6 1.04Zn-H2 3.5 0.19 4.9 0.07W 2.2 0.12 0.3 < 0.01B 2.0 0.11 3.0 0.04Mn 0.4 0.02 22.7 0.32


376 Reina et al.

DISCUSSION

Dulce pond is a flow-through shallow systemwhere dilute groundwater out-seepages into thepond along the western margin (upgradient end)and recharges the ground water by in-seepage atits downgradient end on the eastern shore (Sackset al., 1992). Groundwater feeding remains rela-tively constant throughout the year, creating anin-seepage zone along the western side of thispond, where the water is slightly acidic and iso-topically lighter and has a more diluted concen-tration of major cations and anions than that atthe eastern margin; evaporation on the pond sur-face and calcite dissolution as water seeps backinto the shallow aquifer explain the differencesin the concentration of these ions (Sacks et al.,1992). Thus, this mechanism accounts for thehigher concentration of ions in both the surfaceand pore water samples collected from the NEside in the present study (Table 1). The concen-tration of NH+4 was significantly higher on theSW side and was maximal in the sediment porewater compared with the surface water samples.In contrast, the concentration of NO−3 in the sur-face and pore water samples was significantlylower compared with that detected in the NE side.The accumulation of reduced dissolved inorganicN in water pockets covered by the iron-oxide filmmay be due to poor nitrifying activity resultingfrom the suboxic conditions found along the SWshoreline (Table 1).

The accumulation of oxides on the sedimentsurface due to groundwater out-seepage resultedin the observed differences in Fe concentrationbetween both sides of the study site. The concen-tration of Fe2+ in the water underneath this iron-oxide film has been reported to reach 18.8 mg/L(Portillo et al., 2008). This concentration is highcompared with those observed in other freshwa-ter and brackish environments (Colliene, 1983;Luther et al., 1992; Shaked et al., 2004) but sim-ilar to the concentrations recorded in the TintoRiver, an acidic river (mean pH = 2.5) that drainsacross the Iberian Pyrite Belt and reaches the At-lantic coast to the west of our study sites (López-Archilla & Amils, 1999). Because the pH in thestudy pond was never as low as in the Tinto

River, where sulphur- and iron-oxidizing bacte-ria are common (López-Archilla et al., 2001),the concentration of Fe2+ in the SW margin ofthe study pond requires a different explanation.Additionally, typical iron-reducing and oxidizingspecialist bacteria were not identified as the ma-jor metabolically active bacterial components atthe study site (Table 3).

The relevance of light in the formation of iron-oxide films in the laboratory incubations of thewater and sediment samples (Table 2) suggeststhat the light-induced redox cycling of iron mayprovide an explanation for the presence of thisfloating iron-oxide film in the natural environ-ment. According to Sulzberger et al. (1989), arelatively high concentration of Fe2+ can be pre-served at oxic-anoxic boundaries when the oxy-genation rate is slow compared with the reductionof Fe3+ due to the presence of surface complexesof iron-oxyhydroxides (or FeOOH). This reduc-tive dissolution of FeOOH surfaces is a photo-catalytic process that is usually accompanied bythe reoxidation of Fe2+ by O2 and a further Fe-mediated oxidation of organic matter (Stumm &Sulzberger, 1992). As a result, a high concen-tration of insoluble FeOOH overlies a concen-tration peak of Fe2+ at the oxic-anoxic bound-ary (Sulzberger et al., 1989). In the littoral oflakes, the oxic-anoxic interface is generally setat the sediment surface or at a depth of a few cen-timetres into the sediment (Gerhardt & Schink,2005). This was the case at the NE shore of thestudy pond, where the concentration of Tot-Feincreased drastically in the sediment pore watersamples collected at depths below 5 cm probablydue to the reductive dissolution of iron-oxides.At the SW shoreline, in contrast, the oxic-anoxicboundary was positioned at the air-water inter-face likely due to the seepage of poorly oxy-genated groundwater and the low impact of thedominant wind on the shoreline protected by thedune. The iron-oxide film did not remain intact ata distance of 1 m from the SW shoreline becausethe water surface was no longer protected fromthe wind action by the dune steep, which causedthe film to start to break and dissolve. The directmixing of groundwater seepage with pond open-water was not likely the cause of the disappear-



ance of the iron-oxide film at a distance of 1 mbecause the concentration of major conservativeions in the water (Cl− and Na+) was not signifi-cantly different between the samples collected atdistances of 0 and 1 m. Therefore, the concentra-tions of Tot-Fe and Fe2+ in the water were stillhigh at a distance of 1 m, where particles of col-loidal iron were present, especially when the wa-ter pH was below 7.

Because the concentration of Tot-S andsulphide production were relatively low at theSW side, the concentration of sulphide was notsufficiently high to induce the abiotic removal ofall the iron via the formation of FeS. The con-centration of Tot-Fe in the surface and pore watersamples was higher on the SW side, and Fe2+

was maintained in solution due to the existing re-ducing conditions. On the NE side, the elevatedproduction of sulphide by sulphate-reducingbacteria at the upper sediment layers likelytrapped iron, resulting in the formation of FeSprecipitates. In fact, sulphate-reducing bacteriahave been reported exhibit decreased activityat high iron concentrations (Lovley & Phillips,1986; Chapelle & Lovley, 1992). Thus, the lowersulphate-reduction activity observed at high ironconcentration suggests a larger availability oflow-molecular-weight organic compounds (oth-erwise consumed by sulphate-reducing bacteria),which can become a nutrient resource for otherheterotrophic bacteria, as was the case in theseorganic-rich environments. These results supportthe findings obtained in previous studies (Lovley& Phillips, 1986; Chapelle & Lovley, 1992),indicating the existence of competitive bacterialphenomena regulated by the iron concentrationin groundwater and shallow marshes.

The inorganic matrix of this natural film wasmainly composed of Fe (87 %), followed by Si,Ca2+, Na+, S, Mg2+ and traces (< 1 %) of P, K,Mn, Al, Zn, B, As, Sr and Cu. Trace metalscan be expected to be found due to the metal-binding capacity of iron-oxyhydroxides (Nelsonet al., 1999). The composition of the iron-richfilms obtained in this study was very similar tothat found in the air-water interface of Fe2+-richgroundwater seepage discharged at the base ofPleistocene sand dunes in Oregon (pH ∼ 6.0).

These iron-rich films were composed of two-lineferrihydrite upon further abiotic oxidation, as iscommonly the case in organic-rich environmentswith pH levels of ∼ 5.0 (Grathoff et al., 2007).Organic matter, which reached 14 % based onthe measurement of organic carbon, was a sig-nificant component of floating surface films intwo Swedish groundwater discharge areas withpH values similar to those found in our studysites (Kleja et al., 2012). Ferrihydrite was likelypresent as small particles with humic materialsorbed onto surfaces or included in the particles,making the particles sufficiently hydrophobic toavoid settling in the absence of significant physi-cal disturbance (Kleja et al., 2012).

Using as an analogy the iron-oxide filmproduced by laboratory enrichment throughanaerobic incubations and its thickening in thepresence of light, we suggest that the formationof this floating film in nature is due to a com-bination of both biotic and abiotic processes.Common by-products of the metabolism offermenting bacteria are effective anionic ligandsfor Fe3+ reduction, and bacteria (includingnon-metabolizing iron bacteria such as Bacillussubtilis) can adsorb Fe2+ onto their cell surfacesand facilitate iron-oxide formation (Châtellier& Fortin, 2004). Furthermore, both bacterialorganic acids produced during anaerobic degra-dation and fermentation processes of complexorganic matter and bacterial extracellular poly-mers can initiate the photocatalytic binding andoxidation of iron because they adsorb positivelycharged Fe-oxides (Sulzberger et al., 1989). Analternative cycling pathway may be the rapidoxidation of Fe2+ in the presence of moderateconcentrations of both O2 and H2O2 in sunlitsurface waters and the subsequent reduction ofFe3+ by reductants, such as superoxide, formedupon solar irradiation at circumneutral pH(Emmenegger et al., 2001). The growth of facul-tative anaerobic bacteria, such as Enterobacter,may be an essential part of this redox cyclingbecause these organisms produce H2O2 duringheterotrophic metabolism under suboxic condi-tions and can use Fe2+ as an electron acceptor toneutralize this H2O2 production (Ghiorse, 1984;Sorokin, 1999). This iron-oxide film could bring


378 Reina et al.

other advantages to the growth of facultativeanaerobic bacteria as this floating film mayprovide bacteria a photoprotection mechanismbecause intense direct sunlight represents oneof the major factors affecting the viability ofGammaproteobacteria and Firmicutes, amongother bacteria (Barcina et al., 1990).

In conclusion, a combination of environ-mental factors, such as protection against wind,groundwater seepage, and the existence of asandy aquifer, and the activity of ubiquitousheterotrophic bacteria were likely responsiblefor the formation and prevalence of iron-oxidefilms floating on the water surface in the studysite. Slightly acidic and suboxic conditions weremaintained in areas where this iron-oxide filmwas intact, creating significant differences in thenutrient content, trace metal concentrations, andbacterial communities between different areasof the same water body. The accumulation ofiron-oxides increased the pool of P bound toiron oxyhydroxides in the sediment, whereas thereducing conditions promoted the solubilizationof Fe(II) and Zn and the accumulation of NH+4in both surface and pore water. In contrast toother natural environments with iron-oxidedepositions, our results describe the formationof floating iron-oxide films by microorganismsnot belonging to the typical iron-related bacterialgenera and their interaction with geohydrologicaland chemical phenomena in the iron redox cycleof a freshwater wetland. The finding of theformation of a thin iron-oxide layer in anaerobiclaboratory incubations from the natural sedi-ment samples indicates a greater contributionof the heterotrophic bacterial community toiron transformations than previously thoughtdue to the ubiquity of these bacteria in naturalenvironments rich in organic matter.

ACKNOWLEDGEMENTS

We are grateful to Han Golterman for his supportand suggestions. We thank Lotte Fleskens andChristien van der Zwart for collaborating withthe sediment collection and P-fractionation at thelaboratory. This study was partially supported by

the Spanish Ministry of Education and Science(CGL2004-03927-C02-01/BOS).

REFERENCES

ALTSCHUL, S. F., W. GISH, W. MILLER, E. W.MYERS & D. J. LIPMAN. 1990. Basic localalignment search tool. Journal of MolecularBiology, 215: 403–410.

BALCH, W. E., G. E. FOX, L. J. MAGRUM, C. R.WOESE & R. S. WOLFE. 1979. Methanogens:Reevaluation of a unique biological group. Micro-biological Reviews, 43: 260–296.

BARCINA, I., J. M. GONZÁLEZ, J. IRIBERRI & L.EGEA. 1990. Survival strategy of Escherichia coliand Enterococcus faecalis in illuminated fresh andmarine systems. Journal of Applied Bacteriology,68: 189–198.

BERTHELIN, J., G. ONA-NGUEMA, S. STEMM-LER, C. QUANTIN, M. ABDELMOULA & F.JORAND. 2006. Bioreduction of ferric species andbiogenesis of green rusts in soils. Comptes Ren-dues Geoscience, 338: 447–455.

CHAPELLE, F. H. & D. R. LOVLEY. 1992. Com-petitive exclusion of sulphate reduction by Fe(III)-reducing bacteria: a mechanism for producing dis-crete zones of high-iron ground water. Ground Wa-ter, 30: 29–36.

CHÂTELLIER, X. & D. FORTIN. 2004. Adsorptionof ferrous ions onto Bacillus subtilis cells. Chemi-cal Geology, 212: 209–228.

CHÂTELLIER, X., M. M. WEST, J. ROSE, D. FOR-TIN, G. G. LEPPARD & F. G. FERRIS. 2004.Characterization of iron-oxides formed by oxida-tion of ferrous ions in the presence of variousbacterial and inorganic ligands. GeomicrobiologyJournal, 21: 99–112.

COLLIENE, R. H. 1983. Photoreduction of iron in theepilimnion of acidic lakes. Limnology & Oceanog-raphy, 28: 83–100.

CORNELL, R. M. & U. SCHWERTMANN. 1996.The iron oxides. Weinheim VCH, New York. USA.

DAVIS, R. A. Jr., A. T. WELTY, J. BORREGO, J. A.MORALES, J. G. PENDON & J. G. RYAN. 2000.Rio Tinto estuary (Spain): 5000 years of pollution.Environmental Geology, 39: 1107–1116.

EMERSON, D. & J. V. WEISS. 2004. Bacterial ironoxidation in circumneutral freshwater ct: findingsfrom the field and laboratory. GeomicrobiologyJournal, 21: 405–444.



EMMENEGGER, L., R. SCHÖNENBERGER, L.SIGG & B. SULZBERGER. 2001. Light-inducedredox cycling of iron in circumneutral lakes.Limnology & Oceanography, 46: 49–61.

GERHARDT, S. & B. SCHINK. 2005. Redoxchanges of iron caused by erosion, resuspensionand sedimentation in littoral sediment of a fresh-water lake. Biogeochemistry, 74: 341–356.

GHIORSE, W. C. 1984. Biology of iron- and man-ganese-depositing bacteria. Annual Review of Mi-crobiology, 38: 515–550.

GOLTERMAN, H. L. 2004. The chemistry of phos-phate and nitrogen compounds in sediments. Klu-wer Academic Publisher, Dordrecht/Boston/Lon-don.

GONZÁLEZ, J. M., A. ORTIZ-MARTÍNEZ, M. A.GONZÁLEZ-DEL VALLE, L. LAIZ & C. SAIZ-JIMÉNEZ. 2003. An efficient strategy for screen-ing large cloned libraries of amplified 16S rDNAsequences from complex environmental communi-ties. Journal of Microbiological Methods, 55: 459–463.

GONZÁLEZ, J. M., J. ZIMMERMANN & C. SAIZ-JIMÉNEZ. 2005. Evaluating putative chimeric se-quences from PCR amplified products and othercross-over events. Bioinformatics, 21: 333–337.

GRASSHOFF, K. & H. JOHANSEN. 1972. A new,sensitive and direct method for the automatic de-termination of ammonia in seawater. Ice Journalof Marine Science, 34: 516–521.

GRATHOFF, G. H., J. E. BAHAM, H. R. EAST-ERLY, P. GASSMAN & R. C. HUGO. 2007.Mixed-valent Fe films (schwimmeisen) on the sur-face of reduced ephemeral pools. Clays & ClayMinerals, 55: 638–646.

HERBILLON, A. J. 2006. Ferrosic hydroxides, greenrusts and fougerite in the biogeochemical cycle ofiron. Comptes Rendues Geoscience, 338: 393–401.

HUBER, J. A., D. B. M. WELCH, H. G. MORRI-SON, S. M. HUSE, P.R. NEAL, D. A. BUTTER-FIELD & M. L. SOGIN. 2007. Microbial popula-tion structures in the deep marine biosphere. Sci-ence, 318: 97–100.

HUGHES, J. B. & J. J. HELLMANN. 2005. The ap-plication of rarefaction techniques to molecular in-ventories of microbial diversity. Methods in Enzy-mology, 397: 292–308.

IONESCU, D., HEIM, C., POLERECKY, L., VOL-KER, T. & D. DE BEER. 2015. Biotic and abioticoxidation and reduction of iron at circumneutral

pH are inseparable processes under natural condi-tions. Geomicrobiology Journal, 32: 221–230.

JAMES, R. E. & F. G. FERRIS. 2004. Evidence formicrobial-mediated iron oxidation at a neutro-philic groundwater spring. Chemical Geology,212: 301–311.

KAMPHAKE, L. J., S. A. HANNAH & J. M. CO-HEN. 1967. Automated analysis for nitrate by hy-drazine reduction. Water Research, 1: 205–206.

KLEJA, D. B., J. W. J. VAN SCHAIK, I. PERSSON& J. P. GUSTAFSSON. 2012. Characterization ofiron in floating surface films of some natural wa-ters using EXAFS. Chemical Geology, 326–327:19–26.

KNAPP, E. P., J. S. HERMAN, G. M. HORNBER-GER & A. L. MILLS. 1998. The effect of distri-bution of iron-oxyhydroxide grain coating on thetransport of bacterial cells in porous media. Envi-ronmental Geology, 33: 243–248.

LANE, D. J. 1991. 16S/23S rRNA sequencing. In:Nucleic Acid Techniques in Bacterial Systematics.E. Stackebrandt & M. Goodfellow (eds): 205–248.Wiley, Chichester. UK.

LLAMAS, R. 1990. Geomorphology of the eoliansands of the Doñana National Park (Spain).Catena Supplement, 18: 145–154.

LÓPEZ-ARCHILLA, A. I. & R. AMILS. 1999. Acomparative ecological study of two acidic riversin southwestern Spain. Microbial Ecology, 38:146–156.

LÓPEZ-ARCHILLA, A. I., I. MARÍN & R. AMILS.2001. Microbial community composition and ecol-ogy of an acidic aquatic environment: the TintoRiver, Spain. Microbial Ecology, 41: 20–35.

LOVLEY, D. R. & E. J. P. PHILLIPS. 1986. Organicmatter mineralization with reduction of ferric ironin anaerobic sediments. Applied & EnvironmentalMicrobiology, 51: 683–689.

LOVLEY, D. R., S. J. GIOVANNONI, D. C. WHITE,J. E. CHAMPINE, E. J. PHILLIPS, Y. A. GORBY& S. GOODWIN. 1993. Geobacter metallire-ducens gen. nov. sp. nov., a microorganism capableof coupling the complete oxidation of organiccompounds to the reduction of iron and othermetals. Archives of Microbiology, 159: 336–344.

LUTHER III, G. W., J. E. KOSTKA, T. M. CHURCH,B. SULZBERGER & W. STUMM. 1992. Seasonaliron cycling in the salt-marsh sedimentary envi-ronment: the importance of ligand complexes with


380 Reina et al.

Fe(II) and Fe(III) in the dissolution of Fe(III) min-erals and pyrite, respectively. Marine Chemistry,40: 81–103.

MILLS, H. J., R. J. MARTINEZ, S. STORY & P. A.SOBECKY. 2004. Identification of members of themetabolically active microbial populations associ-ated with Beggiatoa species. Applied & Environ-mental Microbiology, 70: 5447–5458.

MUÑOZ-REINOSO, J. C. 2001. Vegetation changesand groundwater abstraction in SW Doñana,Spain. Journal of Hydrology, 242: 197–209.

MURPHY, J. & J. P. RILEY. 1962. A modified singlesolution method for the determination of solublephosphate in natural waters. Analytica ChimicaActa, 27: 31–36.

MUYZER, G., E. C. DE WAAL & A. G. UITTER-LINDEN. 1993. Profiling of complex microbialpopulations by denaturing gradient gel electro-phoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Applied &Environmental Microbiology, 59: 695–700.

NELSON, Y. M., L. W. LION, M. L. SHULER & W.C. GHIORSE. 1999. Lead binding to metal ox-ide and organic phases of natural aquatic biofilms.Limnology & Oceanography, 44: 1715–1729.

NEUBAUER, S. C.,D. EMERSON & J. P. MEGONI-GAL. 2002. Life at the energetic edge: Kinetics ofcircumneutral iron oxidation by lithotrophic iron-oxidizing bacteria isolated from the wetland-plantrhizosphere. Applied & Environmental Microbiol-ogy, 68: 3988–3995.

PORTILLO, M. C. & J. M. GONZÁLEZ. 2008.Statistical differences between molecular finger-prints from microbial communities. Antonie vanLeeuwenhoek, 94: 157–163.

PORTILLO, M. C., M. REINA, L. SERRANO, C.SAIZ-JIMENEZ & J. M. GONZÁLEZ. 2008. Roleof specific microbial communities on the bioavail-ability of iron in the Doñana National Park. Envi-ronmental Geochemistry & Health, 30: 165–170.

RENTZ, J. A., C. KRAIYA, G. W. LUTHER III &D. EMERSON. 2007. Control of ferrous iron oxi-dation within circumneutral microbial iron mats by

cellular activity and autocatalysis. EnvironmentalScience & Technology, 41: 6084–6089.

REVSBECH, N. P. 2005. Analysis of microbial com-munities with electrochemical microsensors andmicroscale biosensors. Methods in Enzymology,397: 147–66.

SACKS, L. A., J. S. HERMAN, L. F. KONIKOW &A. L. VELA. 1992. Seasonal dynamics of ground-water-lake interactions at Doñana National Na-tional Park, Spain. Journal of Hydrology, 136:123–154.

SERRANO, L., M. REINA, G. MARTÍN, I. REYES, A.ARECHEDERRA, D. LEÓN & J. TOJA. 2006. Theaquatic systems of Doñana. Limnetica, 25: 11–32.

SHAKED, Y., Y. EREL & A. SUKENIK. 2004. Thebiogeochemcial cycling of iron and associated ele-ments in Lake Kinneret. Geochimica Cosmochim-ica Acta, 68: 1439–1451.

SILJESTRÖM, P. & L. E. CLEMENTE. 1990. Geo-morphology and soil evolution of a movingdune system in south-west Spain (Doñana Na-tional Park). Journal of Arid Environmental, 18:139–159.

SOBOLEV, D. & E. E. RODEN. 2004. Characteriza-tion of a neutrophilic, chemolithotrophic Fe (II)-oxidizing β-Proteobacterium from freshwater wet-land sediments. Geomicrobiology Journal, 21: 1–10.

SOROKIN, Y. I. 1999. Aquatic Microbial Ecology.Backhuys Publishers, Leiden. The Netherlands.

STUMM, W. & B. SULZBERGER. 1992. The cy-cling of iron in natural environments: considera-tions based on laboratory studies of heterogeneousredox processes. Geochimica Cosmochimica Acta,56: 3233–3257.

SULZBERGER, B., D. SUTER, C. SIFFERT, S.BANWART & W. STUMM. 1989. Dissolution ofFe(III) (hydr)oxides in natural waters; laboratoryassessment of the kinetics controlled by surfacecoordination. Marine Chemistry, 28: 127–144.

TROLARD, F. 2006. Fougerite: From field experi-ment to the homologation of the mineral. ComptesRendues Geoscience, 338: 1158–1166.


The interplay of hydrological, chemical and microbial processes in … · 2015-11-10 · Limnetica,...

Documents

Transcript of The interplay of hydrological, chemical and microbial processes in … · 2015-11-10 · Limnetica,...