<<

Portland State University PDXScholar

Geology Faculty Publications and Presentations Geology

1-1-2011 Potential Fossil Endoliths in Vesicular Pillow , Patch , Eastern North

Barbara Cavalazzi University of Johannesburg Frances Westall Centre de biophysique moléculaire Sherry L. Cady Portland State University Roberto Barbieri Università di Bologna Frédéric Foucher Centre de biophysique moléculaire

Let us know how access to this document benefits ouy . Follow this and additional works at: https://pdxscholar.library.pdx.edu/geology_fac Part of the Environmental Microbiology and Microbial Ecology Commons, and the Geology Commons

Citation Details Cavalazzi, B., Westall, F., Cady, S. L., Barbieri, R., & Foucher, F. (2011). Potential Fossil Endoliths in Vesicular Pillow Basalt, Coral Patch Seamount, Eastern North Atlantic Ocean. [Article]. Astrobiology, 11(7), 619-632.

This Article is brought to you for free and open access. It has been accepted for inclusion in Geology Faculty Publications and Presentations by an authorized administrator of PDXScholar. For more information, please contact [email protected]. ASTROBIOLOGY Volume 11, Number 7, 2011 © Mary Ann Liebert, Inc. DOl: 1O.1089/ast.2011.0657

Potential Fossil Endoliths in Vesicular Pillow Basalt, Coral Patch Seamount, Eastern North Atlantic Ocean

Barbara Cavalazze· 2 Frances Westall ? Sherry L. Cady? Roberto Barbieri,4 and Frederic Foucher 2

Abstract

The chilled rinds of pillow basalt from the Ampere-Coral Patch in the eastern North Atlantic were studied as a potential habitat of microbial life. A variety of putative biogenic structures, which include fila­ mentous and spherical microfossil-like structures, were detected in K-phillipsi te-filled amygdules within the chilled rinds. The filamentous struchtres (""' 2.5 pm in diameter) occur as K-phillipsite tubules surrounded by an Fe-oxyhydroxide (Iepidocrocite) rich membranous structure, whereas the spherical structures (from 4 to 2pm in diameter) are associated w ith Ti oxide (anatase) and carbonaceous matter. Several lines of evidence indicate that the microfossil-like structures in the pillow basalt are the fossilized remains of . Possible biosignahtres include the carbonaceous nature of the spherical structures, their size distributions and morphology, the presence and distribution of native fluorescence, mineralogical and chemical composition, and environmental context. When taken together, the suite of possible biosignatures supports the hypothesis that the fossi l-like structures are of biological origin. The vesicular microhabitat of the rock matrix is likely to have hosted a cryptoendolithic microbial community. This study documents a variety of evidence for past microbial life in a hitherto poorly investigated and underestimated microenvironment, as represented by the amygdules in the chilled pillow basalt rinds. This kind of endolithic volcanic habitat would have been common on the early rocky planets in our Solar System, such as Earth and Mars. This study provides a framework for evaluating traces of past life in vesicular pillow , regardless of whether they occur on early Earth or Mars. Key Words: Fossil microbes-Microhabitat-Vesicular pillow basalt. Astrobiology 11 , 619-

1. Introduction lisms, is commonly associated with bioalteration of modern and ancient oceanic volcanic glass (for a comprehensive re­ HE OCEANIC LITHOSPHERE contributes significantly to view sec Staudigel el al., 2008). Chemolithoautotrophic mi­ T biogeochemical cycling and the amount of biomass and croorganisms have an affinity for many of the biocsscntial 4 biodiversity on Earth. It is aisa an important habitat for mi­ elemL'J1ts (e.g., Fez ' , Mn . , ~ ,C02) present in basaltic crobial life (Fumes and Staudigel, 1999; Fu mes et a/., 2007; silicates, which include , pyroxene, feldspar, and Santelli et a/., 2008; Staudigel et a/., 2008; Mcloughlin et a/., hornblende (Fisk el al., 2006, and references therein). Altera­ 2009; Nielsen and Fisk, 2010). When the temperature of newly tion of the silicate minerals as a result of weathering and mi­ formed oceanic rocks drops to values that become tolerable crobial extraction releases the metabolically significant cations for life, which is estimated to be less than 121- 122°C (Takai (Kostka el al., 2002; Kim el al., 2004). Bioalteration of the glassy et 111.,2001; Kashefi and Lovley, 2003), they can be colonized basalts, which develops progressively along fracture surfacL"S by microbial communities that utilize the inorganic elements and in and around vesicles produced by degassing, is ac­ and compounds of the rocky substratum as energy sources companied by the formation of authigenic phases (e.g., (Thorseth eJ al., 1995). A wide variety of evidence, which in­ smectite and phillipsite) in the zone of alteration. Staudige! cludes the presence of fossilized microbes, spherical and tu­ and coworkers (2008) presented a schematic model that il­ bular al teration cavitiL"S, and geochemical and molecular lustrates the developmental stages and variety of mechanisms signatures consistent with microbial remains and metabo- involwd in microbial alteration of volcanic glasses.

l[)cpartmcnt of Geology, University of Johannesburg, Johannesburg, South Africil. "Centre de Biophysique Mol&ulaire-CNRS, Qrlcilns, Frall(e. 3Department of Geology, Portland State University, Portland, Oregon.. USA. ~Dipartimento di Scienze della Terril e Geologico-Ambientilli, Univcrsita di Bologna, I3ol0gm, Itilly.

619 620 CAVALAZZI ET AL.

Portugal ATLANTIC OCEAN

FIG. 1. Location map of sample site (star). The sample studied was dredged from the flank of Coral Patch Seamount (34°58.3lON to 34°56.773, 1 1°57.323 W to 11 °57.140; wa­ ter depth 1018m) during the SWIM2004 scientific cruise in Gulf of Cadiz, Atlantic Ocean. The bathymemc map of seamounts Ampere..coral Patch ~ and islands in the eastern North Atlantic Smt.s Ocean is modified from GddmachC'r and !1J> Hoernle (2000). Seirle Smt. ,.,. Morocco

TIle primary glassy rinds of basalts provide a microbial such as carbonates, clays, minerals, and (e.g., IvaTS­ habitat for endoliths, organisms thai reside in the interior son et al., 2008; PeckmalUl et al., 2(08). In this study, we of rocks, and euL'ndoliths, microbes that actively bore into present the results of a detailed investigation of objects with the rock substratums and create microtubular cavities -like morphologies that occur within phillip­ (McLoughlin et Ill., 2007). Chemical alteration of the glassy site-filled amygdules located at and near the outer surfaces of rinds of seafloor basalts and of extrusive volcanic rock in­ pillow bas.llts recovered from the Coral Patch Seamount in creases the range of microenvironments that could support the eastern North Atlantic Ocean (Fig. 1); the preliminary such microbial communities. The diversity of glassy habitats findings related to this research were included in Cavalazzi supports communities of epiliths (L>ukaryotic and prokary­ et al. (200S). The combination of different methods of sample otic microorganisms that live by attaching themselves to the preparation and analytical techniques, which include optical external surfaces of rocks), chasmoendoliths (endoliths that and electron microscopy, confocal laser scanning micros­ inhabit fissures and cracks in rocks), and cryptoendolilhs copy, and Raman microscopy, pennitted the investigation of (endoliths thai are hypothesized to favor the colonization of two types of putative microbial stmctures and their associ­ rocks that have preexisting interstices and porous stmctures, ated chemical compositions. Our study provides additional sellSI! Santelli et al., 200S). evidence that vesicular basalt can support microbial ecosys­ Recent studies have shown that prokaryotes and eukary­ tems and that such communities may leave traces in the otes also inhabit the primary and alteration rinds of crys­ geological record. Of relevance to paleobiology and astrobi­ talline volcanic rocks. Schumann et aL (2004) described ology search strategies is the likelihood that vesicular bas.llts putative fossilized fungal life in vesicles of deep oceanic with similar types of microenvironments would have been bas.1.it collected from a site in the North Pacific (OOP common on any wet (or formally wet) rocky planet that Si te 1224). Connell et al. (20Cl9) described fungi they discov­ experienced extrusive volcanism throughout its gt.'Ological ered on the surfaces of basalts that fanned relatively recently history. at a still-active deep-sea volcano (Vailulu'u Seamount, Sa­ moa). Ivars.'>On el al. (2008) reported the discovery of fossil 2. Materials and Geological Setting chasmoendoliths in -filled veins associated with sub­ seafloor bas.1.ltic rocks s.1.mplcd at the Emperor Seamounts in The s.1.mple, a fragment (-90x60x40 cm) of dark-reddish the Pacific Ocean. PL'Ckmann et III. (2008) and Eickmann e/ al. pillow basalt partially encrusted by fossiliferous pelagic (20Cl9) found fossi l evidence of marine cryptoendolithic , was collected (August/September) during the prokaryotes within the carbonate amygdules (mineral-filled scientific cruise SWIM 2004 (principal investigator: N. 2i­ vesicles) of Devonian pillow bas.llts from Rheinisches teltini, ISMAR-CNR, Bologna) of the R/V URANIA in the Schiefergebirge (Germany), Frankenwald, and Thtiringer Gulf of Cadiz, eastern North Atlantic Ocean (Fig. 1). It was Wald within the Saxothuringian zone in Gennany. These dredged from stalion SWIM04-29 (34°58.3lON to 34°56.773, latter workers, in particular, emphasized the need for more 11 °57.323 W to 11 0 57.140) on the flank of Coral Patch Sea­ research to reveal the ecological niche that cavities and mount in the Ampere-Coral Patch Seamounts region from amygdules in seafloor basalts cou ld provide for life in vol­ moderate depths (1OJSm) (star in Fig. I shows approximate canic rock..<;. locality). Cores (2 cm in diameter) were drilled from the outer Chasma- and cryptoendolithic microorganisms that in­ rind of the pillow bas.llt for thL<; study. habit the margins, walls, and pore spaces of fluid-filled The Amperc-Coral Patch SeamOllllts (31 Ma old) repre­ cavities, voids, veins, and fractures of basaltic rocks can be­ sent the intemK>diate phases of a >72 Ma old hot spot come fossilized via the precipitation of authigenic minerals, (Geldmacher and Hoemle, 2000), which also formed the VESiCULAR HABITAT 621

Ormonde, Unicorn, Seine, and Porto Santo SeamOlmts (Fig. spectra. Beam centering and Raman spectra calibration were 1). Located 460km 5W of the Ampere-Coral Patch Sea­ performed before spectra acquisition by using a Si standard 1 mounts region, the Archipelago is the pres<..'l1t lo­ with a characteristic 5i Raman peak at 520.4cm- • The cation of the Madeira hot spot track. The Serra de Monchique optimum power for ill situ analyses of different minerals, Complex (72- 70Ma old alkaline volcanic rocks) in southern between 1.67 and 1.70 nW at the sample surface, was deter­ Portugal represents the vestiges of the early activity of the mined experimentally. The Raman spectra and maps were hot spot (F ig. 1). Geldmacher and Hocrnle (2000) interpreted collected a few micrometers below the surface of anyone the alkaline volcanic rocks along the Madeira hot spot track, specimen to eliminate possible contamination that might including the Ampere-Coral Patch Seamounts region, as have resultt.>d from the preparation and handling of the thin resulting from the progressive melting and exhaustion of sections. Raman analyses and maps were recorded and recycled (hydrothermally altered) basaltic in a treated with WITecProject2.oo software. For final analysis, discrete pulse of plume material as it upwelled and spread the thin sections that were uncontaminated with oil were cut out at the base of the lithosphere. The composition of the into subsections, etched in an aqueous solution of 1% HCl basaltic rocks along the hot spot track evolves from transi­ for lOs (to rt.'ffiove any room contamination and hand oils), tional tholeiites to basanites to hawaiites to trachytes. The and then air dried. The thin section pieces were mounted crustal oceanic rocks at Ampere-Coral Patch Seamounts on aluminum stubs covered with carbon-conductive adhe­ form part of an alkali (hawaiite type) basaltic suite (Geld­ sive tape. macher and Hocrnle, 2000). Au-coated, C-coated, and non-coated thin sections were observed with an environmental scatming electron micro­ scope (ESEM), FEI Quanta 200, equipped with an Oxford 3. Methods Instruments TNCA 350 X-ray energy-dispersive spectrometer TIle pillow basalt was subsampled and characterized ini­ (E05) system, and a Hitachi S4200 field emission gun scan­ tially by optical microscopy. Six subsamples drilled from the ning electron microscope (FEG-SEM), equipped with Si (Li) 2 cm diameter cores were analyzed for this study. For each detectors from Oxford Instruments Link ISIS. A conductive subsample, one or two uncovered polished petrographic thin (Ag) paint was spread sparingly around the perimeter of the sections (30 lIm thick) were prepared. Thin sections were uncoated sample prior to scanning electron microscope investigated with the use of a transmitted optical light mi­ (SEM) analysis to reduce surface charging. Energy-dispersive croscope, an Olympus BX51 TH-2oo equipped with an X-ray (EDX) analyses were qualitative and semi-quantitative. Olympus DP12 digital microscope camera, and an Olympus ESEM observations were conducted in high and low (1 and BX51 TRF equipped with an Olympus DP72 digital micro­ O.5torr) vacumTI. TIle operating conditions of the scanning scope camera. Subsequently, the thin sections were examined electron microscopes were 5-25 keY accelerating voltage for with a confocal laser scanning microscope (CLSM) and a imaging and 5-15keV for elemental analyses. All the in­ confocal Raman microscope. struments used in this study are located at the Department of Confocal laser scanning microscope images, obtained with Geology, University of Johannesburg (South Africa); at the a LSM 510/3 META microscope, provided a means to cor­ Centre de Biophysique Molecu!aire, CNR5-Orleans (France); relate features observed in the petrographic thin sections at the Centre de Microscopic Electronique, UniversitC d'Or­ with native fluorescence found in the material. CLSM optical leans (France); and at the Centro interdipartimentale Grandi micrographs, acquired with the use of a Type FF fluores­ Strumenti, Universita di Modt.'lla (Italy). cence-free microscopy immersion oil, were recorded with Car!.Zeiss LSM Image Browser software. The image size was 4. Results set at 512x512 pixels, and the area captured in each image 2 4. 1. Rock textures measured O.28xO.28pm . The images were recorded with the software operating in line mode, and four scanned im­ The Coral Patch Seamount pillow basalt is characterized by ages were averaged to n.>duce background noise. The CLSM a vesicular (-15-20% vesicles) glass rind that contains nu­ images were acquired and stored as an electronic signal, merous cooling fractures (i. e., quenched during degassing due which allowed us to apply several electronic image en­ to rapid chilling) (Fig. 2). TIle glassy rind has experienced hancement methods (Amos and White, 20(3). Native fluo­ extensive palagonitization, a process that involves low­ rescence (488 run excitation filter, 505 run long-pass emission temperature hydrolitic alteration of mafic glass by seawater filter) prodUCl.>d by the sample was captured digitally in (Stroncik and Schmincke, 2001). As a result of this process, the images acquired with the use of a plan-Apochromat 63xoil glass devitrifies to microcrystalline secondary minerals as the objective lens (numerical aperture= 1.4). Sequential images sample ages. In thin section, the mainly consists of were recorded as 18 to 64 optical slices. Detaits about indi­ extremely fine-grained reddish-brown secondary mineral., vidual image acquisition parameters are provided (e.g., in (Fig. 2). Rare fractured olivine (forsterite) relicts and pheno­ caption for Fig. 5). In general, slices 0.30 pm thick were ac­ crysts (< 500 Iml in di.1.mctcr) of clinopyroxene (augite) and quired, and the stack size was defined as a hmction of the chromian spinel (magnesiochromite) occur throughout the resolution required for anyone image and corresponding glassy rind (Fig. 2; Table 1). thickness of the scanned volume. The vesicular texture occurs along the outer margin of Raman analyses and compositional maps were acquired the glassy rind of the pillow basalt. It consists of subspherical with an alphaSOO WiTec AFM-confocal Raman microscope. to elongate vesicles characterized by a diameter from 200 to Three objectives (Nikon 20x, SOx, and 1OOx) and a frc­ 300 11m to less than 1.5 mm (Fig. 2). The vesicles are par­ qUl.'llCy doubk.od Nd:YAG (532run) Ar-ion 20mW mono­ tially to totally filled by transparent to translucmt zeolite chromatic laser source were used to collect the Raman (K-phiUipsite) (Fig. 2; Table 2). Mineral-filled vesicles are 622 CAVALAZZI ET AL.

eGC- <,"" FIG. 2. Mosaic of transmitted-light photomicrographs of a petrographic thin section of the vesicular basalt. TIle palago­ nitizcd, reddish-brown glassy rind of the pillow bas.1.it is characterized by amygdulcs, thai is, partially or totally infilJcd by K-phillipsite, sparse grains of altered phenocrysts of olivine (01) and clinopyroxene (cpx), and cooling fractures. Some amygdules (black arrows) are filled by granular K-philli psite wi th disseminated Fe-Ti oxides and microbial-l ike structures. A Fe-Mn-Ni-Ti-oxide crust separates the outer edge of the pillow basalt from the encrusting -rich lithified pelagic ooze and K-phillipsilc. referred to as amygdules. Some amygdules aTe filled with elongate fibrous K-phillipsite crysta ls that are radially arranged and protrude away from discrete nucleation centers located along the vesicle walls. Other K-phillipsite­ filled amygduk>s display a granular texture (Figs. 2 and 3A). The surface of the pillow basalt is characterized by either TABLE 1. RAMAN D ETECTION O F RARE R ELICT C RYSTALS a 100-ISO/lm thick reddish-black crust of Mn-Fe-Ti-Ni· WITHIN THE G LASS RlND enriched oxides or a crust of lithified Ca-carbonate ooze that originally contained abtmdant microfossils such as calcare­ Raman spec/TIl ous narulOplankton and planktonic foraminiferal tests (Fig. 2). Although difficult to distinguish in thin section, fossils of calcareous nannoplankton (rare specimens of Spilenolitlllls ~ spp. and CycliCllrgolitJllls spp.) were identified in the car­ bonate. Rare fragments of radiolarians occur among the • , abundant remains of foraminifera. The planktonic forami­ niferal assemblage is dominated by epi- and mcsopdagic • genera, such as Globigerina, Globigerinoides, Glooorotalia, G/o­ ~ boturborotalia, and Orbulina. I'-v ,

4.2. Microbe-like morphologies " / i Possible biogenic filamentous and spherical stmctures , and small grains of Ti-Fe oxides (anatase) were fowld in the -;so 3Jo ~ ~o aJO-;JO"7.J70 , ~ K-phillipsite-filled amygd ules that display a granular tex­ Wev=enumber " (em-') ture and are located closest to the outer margin of the pillow Raman spectrum 1: measured spectrum (S32nm) of olivine, basalt (Fig. 3; Table 2). The orange-red microbe-like struc­ forsterite (M~Si04 )' Reference spectrum: forsterite tures appear diffuse and transparent when viewed with (532nm), RRUFF ill: R040018. plane polarized light, as shown in the petrographic thin Raman spectrum 2: measured spectrum (532nm) of section photomicrographs of Figs. 4-10. clinopyroxene, augite «Ca,Mg,Feh(Si,Alh06)' Reference spectrum: augite (532 run), RRUFF ID: R06lO86. 4.2.1. Filamentous structures. The fi lamentous struc­ Raman spectrum 3: measured spectmm (532 nm) of tures appear as isola ted objects that are characterized by a chromium spinel, magnesiochromite (MgCr20 4)' constant diameter of 2.5-3.5 pm and an observed maximum Reference spectrum: magnesiochromite (532 nm), RRUFF length of 30 pm (Fig. 4). Enclosed in a granular K-phillipsite ID: R060796. matrix, the filamentous structures have a sinuous to straight All peaks in the region considered for each spectrum represent morphology wi th a smooth surface (Fig. 4A), although the their vibrational mode. Raman spectra were calibrated with refer­ outer surface of some rare specimens, such as the ones ence spectra after RRUFF DataBase (Llctsch and Downs, 20(6). VESiCULAR HABITAT 623

TABLE 2. THE MINERAL PHASES ANA LYZED BY SEM-EDX MICROANALYSIS AND RAMAN MICROSCOPY OF THE AMYCDULES FROM THE GLASSY PrLLOw BASALT FRONT SEM-EDX microanalyses: K-pllillipsite

Samples SiOl All o.~ Fct) MgO eno Na ~ O K,O To/a} CPR 01.1 61.45 23.94 0.33 0.00 5.13 9.16 100.01 CPR 01.2 59.44 24.48 0.71 0.23 6.69 8.45 100.00 CPR 01.3 61.22 22.77 0.87 1.12 5.70 8.18 100.01 CPR 02.1 59.52 23.24 0.31 0.26 6.60 10.07 100.00 CPR 02.3 60.44 22.56 0.56 0.86 1.85 5.43 8.29 99.99

Phillipsite: (Ca,Naz,K2h(AI"Si 1O:>O:u 12H20. Analyses were made on C-coated thin section using 15keV accelerating VOltage.

Rnman spec/rllm: K-phillipsite SEM -EDX spectrum: K-phil1ipsitc

$ 1; •• .§' - :l I il - ~ i:. - • Wavenumber (em -') Raman s pectrum 1-; measured sped rum (532 nm) of SEM-EDX s pectrum: spectrum was acquired on C-coated mineral phase, phillipsite, filling vesicles. thin St.'Ction using 15 keY accelerating voltage. Raman spectrum 2: reference spectrum of phillipsite (532nm) (RRUFF ID: R060161).

Rmlum Spectntlll: anatase shown in Fig. 48, appear undulated. l3ecause the vesicle walls are often incmsted with Fe-Ti oxides (Figs. 2 and 3), only a few sinuous filamentous stmctures with a smooth surface are observed to be attached with one end to the vesicle walls (white arrow in Fig. 4C). llle filamentous structures filled with K-phillipsite--consist either of indi­ vidual tubular forms or short chain.. of sma!! spheres (~3- 2.5 InTI in diameter) (Figs. 5 and 6). Both types of filamentous stmctures are characterized by a distinctive native fluores­ CL'flce (Fig. 5). Some of the tubular structures are t:.'ncapsu­ lated in an amorphous C-Fe-O-rich L'flvelope less then 500 nm thick (black arrows in Fig. 6B, 6E). , Raman analyses were made of some of the objects thilt display a filamentous morphology (Fig. 7). Raman mapping (Fig. 7C- 7E) provides a means with which to correlate the objects with cellular-like morphologies and their mim.'l"alog­ Wavenumber (em ·') kal composition. The filaments consist of K-phillipsite (Fig. Raman spectrum 1-; measured spectnun (532 lUn) of mineral 7 A-C), whereas the amorphous envelope around some of the phase, anatase (ri02), associated to K-phillipsite filling filamentous structures consists of Fe oxyhydroxide, lepido­ vesicles. crocite that is enriched in carbonaceous matter (Fig. 7A- 8, Raman spectrum 2: reference spectrum of anatase (532 nm) 7D-E). In Fig. 7B, the Raman spectrum 2 includes the major (R.RUFF ID:R070582). bands of the first-order region of the disordered carbona­ ·All peaks in the region considered for the spectrum represent the ceous matter (O!; ~ 1345 em - I; Gl: ~ 1600 em - I) associated vibrational mode. with the filamentous morphologies. The relative intensity of 624 CAVALAZZI ET AL.

-20 ".)<

., ,., 0 , • ... . . K-phi ... K-phi amygdul:,. / ,. ",r./I I' , ..., ....'" •,

FIG. 4. Thin SI..'Ction transmitted-light photomicrographs ~ di!l&lem ."angeme<>t of spl>eric;lliormt 1...... 1f ..Tl QX C/U$\I that reveal the presence of individual filamentous structures ~ ~1' QraNItJan.taM (diameter size: 2..5--3..5llm) within K-phillipsite--filled vesi­ cles. The filamentous structures show smooth sinuous [black FIG. 3. Thin section transmitted-light photomicrograph of arrows in (A) and (C)] and corrugated [arrow in (B)] an amygdule from the altered glass of pillow basalt rinds morphologies. In (C), (white arrows) a filament structure is (Al. (8) Illustrates detail of the boxed area in (Al and shows attached to the vesicle wall and embedded in K-phillipsite. the distribution of spherical and filamentous structures K-phi, K-phillipsitc. (black arrows). (C) Simplified sketch of an amygdule that illustrates their dominant characteristics and distribution of purported microfossils. Details in (C) arc not to scale. K-phi, outer diameter between 2 and 2..5 11m (Fig. 8). The outer walls K-phillipsite. of the macrospheres display a reddish, transparent appear­ ance in petrographic thin sec ti on. The macrospheres occur the D1 and Gl bands indicates a poorly crystalline, disor­ mainly as paired bodies or as aggregates (Fig. SA--B). TIle dered carbonaceous matter. microspheres also appear in petrographic thin section as transpa.rcnt, smooth-surfaced individuals or paired bodies 4.2.2. Spherical structures. Two sizes of spherical (Fig. BC). SEM-EDX analysis revealed that the distinctive outt.'r structures were obst..>rwd: macrospheres, with an average rims of the macrospheTL"S and microsphercs have a similar outer diameter of 4tlffi, and smaller microspheres, with an composition that shows an enrichment of carbon, Ti, and Fe a~ygdule wall

filament consisting of small

FIG . 5. Magnified view of optiml tfilnsmittcd-light (A- B) illld CLSM images (C- F) of putative microfossils, which include filamentous structures and short chains of small spht.m> embedded in the K-phillipsitL'-fillL'CI vesicles. The CLSM unages provide details of the structures observed in the boxed area shown in (B). The CLSM unage; reveal a native florescent signal associated with the filamentous stmchlres, highlighting the internal fine-smle morphological feahu'eS, which include small spheres. CLSM ac­ quisition parameters: (C) image from stack scan mode (stack size/Ian: 35.7x3S.7xS.l/an3; stack size/pixel: 512x512x18; pixel depth: 8 bit; scaling: x=0.070 pm, y=O.07O Iml, z=o.30 lUll; wavelength: 488 lUll 9.0%); (D) ima.ge from plillle SCilll mode (stack size/ 3 pixel: I x I x I; pixel depth: 8 bit) obtained from (C); (E) image from stack scan mode (stack size: 27.3x27.3x5.1/Ull ; stack size/ pixel: 512x512xlR; pixel depth: 8 bit; scaling: x =0.053 lan, y=O.053/an, z=O.30/lffi; wavelength: 4RRnm 9.0%); (F) image from plane scan mode (stack size/pixel: I x l xl; pixel depth: 8 bit) obtained from (E). K-phi, K-phillipsite.

625 626 CAVALAZZI ET AL.

K-phi

sXI. X3 •

• 0 4 , , 0 , , 4 Ent'DI'(\

K-phl

4 EnergyQceV)

FIG. 6. ESEM images of filamentous morphotypes within K-phillipsite-filled vesicles. The filamentous structures consist of K-phillipsite microtubes [Xl: EOX spectrum in (0 ); white arrows in (A)] surrounded by an Fe-rich carbonaceous membrane­ like envelope [X2: EOX spectmm in (E); black arrows in (B)]. Some filaments are fonned by short chains of small spheres [X3: EOX spectrum in (D); white arrows in (C)]. The ESEM ilThlgcs were acquired on uncoated and lightly ('Idled (HCll%, less than 4s) thin sections in low vacuum (1 torr) with a 25keVelectron beam. The samples were subsequently Au-coa ted for ESEM-EDX analyses. EOX data were acquired with a 5 keY accelerating voltage. K-phi, K-phi!1ipsite.

relative to the matrix (Fig. 9). Ra.lThl.n microscopy revealed 2010a). Although the morphology alone of bacteria-like 0b­ that the Ti-enrichL"CI phase is anatase (Ti02) (Table 2) and that jects may not be a diagnostic biosignature, it serves an im­ it is associated with carbonaceous matter (Fig. 10). portant role in spatially locating wi thin a potential ecological niche candidate biosignatures that should be characterized 5. Discussion dlemically in more detail. The discovery of a suite of possible biosignaturcs associated with the morphological objects We infer the biogenicity of the filamL'J1tous and spherical strengthens our interpretation of their biogenic origin. structures in the K~phi1lipsite amygdules from the Coral In summary, the suite of possible biosignatures that are Patch Seamount pillow basalt based on the full suite of consistent wi th a biogenic origin for the microbe-like remains morphological, mineralogical, and chemical characteristics includes the following: that were observed with different analytical instmmcnts over a range of scales from the microscopic to submicro­ (1) Morphology-distinctive micrometer-sized objects scopic (cf. Buick, 1990; Knoll, 1999; Summons et Ill., 1999; that display sizes, size distributions, and filamentous Cady e/ Ill., 2003; Brasier et Ill., 2006; Schopf et Ill., 2007, and spherical morphologies consistent for bacteria; VESICULAR HABITAT 627

FIG. 7. Optical photograph and Raman analy­ ses of an area in a polished thin section that contains a filament. (A) Reflected light optical photomicrograph (focus plane: 7 11m below to the surface). The number indicates the location from which the representative Raman spectra of the host matrix and filamentous structure (dashed lines) were collected, whereas the boxed area indicates the location of the Raman maps. (8) The host matrix, spectrum 1, consists of phillip­ site, and the fi lamentous stmcture as shown in the multiphase spectra 2 is made of lepidocrocile (k.'P) associated with phillipsite (phi) and traces of carbonaceous matter (com). Raman spectra were calibrated after RRUFF DataBase (Laetsch and Downs, 2006), Demoulin ct al. (2010), and Marshall et al. (2010). Inset shows the reference spectrum of pure lepidocrocite (532nm) after Demoulin et aL (2010); see also reference spec­ trum (780mn) RRUFF ID R050454. (C- E) Raman maps that show the distribution of phillipsite (phi), lepidocrocite (lep), and carbonaceous ma tter (com) of the area that contains the filamentous structures (dashed lines). Data are presented as a black-white intensity ratio map where white regions indicate high concentrations of the component labeled at the bottom of each paneL The Raman maps correspond to the same focal plane imaged in (A).

(2) Chemical composilion--carbonaccous composition of for additional possible biosignatures from such structures microbial-like morphological features and a correlation (Cady et al., 2003). The purporlL>d microfossils in the bas.1 lt of the carbonaceous remains and authigenic mineral amygdules studied here display a number of key morpho­ precipitates associated with the purported microfos­ logical attributes that are consistent with a biological origin: sils, which are consistent for bacteria; (i) size distribution typical of filamentous and coccoidal (3) Ecological setting- habitat for microbial conummilies prokaryotes; (ii) minimal size variation within anyone par­ with enhanced preservation potentiaL ticular morphological group (filaments display diameters 2.5-31101, Fig. 4; macrospheres~ 4 /lm, and microspheTL"S 2-25 11m, Fig. 8); (iii) shape consistency (e.g., filamentous and 5. 1. Morphology of putative microfossils consistent spherical) with constant diameters (Figs. 4-10); and (iv) an with life arrangement and structural complexity of individual cells or Though morphology alone GUffiot serve as a definitive colonial aggregates (e.g., short chains of small spheres that biosignature (e.g., Garcia Ruiz et al., 2(02), morphological comprise filaments, Fib'S' 5 and 6). The lack of variation in the evidence of microfossil-like objects helps to focus the search morphological attributes is much more typical of microbial

K-phi

K-phi • •• K-phi •

FIG. 8. Transmitted light photomicrographs of a petrographic thin section that illustrates two sizes of spherical structures within the K-phillipsite-filled vesicles. (A, 8 ) The reddish-brown macrosphercs (~4.5 pm outer diameter) appear as thick­ walled semi-spheroids. (C) The transparent (-3 pm diameter) microspheres occur as isolated or paired bodies (black arrows). K-phi, K-phitlipsite; vw, vesicle wall. 628 CAVALAZZI ET AL.

FIG. 10. Optical photograph and Raman maps of the area containing spherical structun."S in a polished thin Sl.'Ction. (A) Reflected light optical photograph (focus plane: 5 pm below to the surface) of the spherical structures (-4/uTI diameter). FIG. 9. ESEM image of spherical structures and their che­ TIle boxed area indicates the location of the RanuUl maps. mical composition. They appear as circular structures with (B- D) Raman maps reveal the distribution of phillipsite

    On el at., 2008). colonize surfaces, fractures, and vesicles in volcanic rocks to L.. ,scr Raman microscopy has been successfully used to obtain energy and nutrients from reduced chemical species characterize natural caroonact-'Ous compounds (e.g., Jehlinca released from basalt during dis.-',olution /alteration, and they and Beny, 1992; Marshall et al., 2010) and for chemical im­ establish themselves in physically stable environments agery of microfossils (Schopf et af., 2005, 2006, 201Oa, 201Ob). where they can proliferate and, when nccess<1.ry, escape Here, we were able to confirm the intimate association of predation (e.g., Edwards et al., 2004, 2005; Cockell and Her­ disordered carbonaceous material with the candidate bio­ rera, 2008; Santelli ef aI., 2008; Chan et al., 2011). Older fila­ signd microorganisms are mineralized (e.g., Bhartia etll/., worth noting that, at present, the interactions between Ti and 2010). Sheathed Fe-oxidizing microorganisms (van Veen microorganisms and the role of Ti in microbial fossilization et al., 1978; Emerson and Moyer, 1997; Chan et al., 2011) are are poorly understood (Shabtai and Fleminger, 1994; Gla­ characterized by chains of ce!!s that, when encased in their moclija et al., 2009). TIlOugh some prokaryotes are known to tubular sheaths, appear as filaments (e.g., note the short adsorb TiOz during metabolic processes (Shabtai and FIe­ chains of sphere observed in Fig. SC- F).Though it is possible minger, 1994; Bedard et Ill., 2006), the precipitation of tita­ that all the native fluorescence associalL>d with the morpho­ nium (and iron) oxide minerals is commonly thought to logical remains of the purported microfossils is produced by represent abiogenic low-temperature alteration of basaltic small autofluorescent crystals, the nonbiological explanation glas..<; with seawater (e.g., Stroncik and Schmincke, 2(01). It (the "nu!! hypothesis" sensu Brasier e/ a/., 20(6) for the ob­ was impossible in this study to establish whether the selec­ served phenomenon is not favored because (i) the authigenic tive Ti mineralization of the spherical structures via anatase mineral precipitates (e.g., lepidocrocite, Fe hydroxide, and was a passive or microbially controlled fossilization process. anatase Ti oxide; see below) are not known to be auto­ fluorescent at the excitation wavelengths used in this study 5.4. Timing of fossilization and (ii) the native fluorescence is always correlated with microfossil-like morphologies. Vesicles are primary segregation structures that represent Native fluorescence of minerals is related to the pres­ the fossilized vestiges of degassing bubbles trapped during ence of impurities, organic activators, and the excitation extrusion and decompression from a volatile-rich melt source wavell'tlgths uSl>d (Gorobets and l~ogojine, 2002). Although (peck, 1978). Although it is difficult to determine when the few data are available in the literature with regard to the minerals infilled the vesicles, characteristics of the microhabi· potential to induce native fluorescence in minerals for the tat and microfossil taphonomy provide some infonnation excitation wavelength used here (J"xc=4R8nm), laser­ with regard to the involvement of microorganisms in the induced fluorescence of the mineral phases associated with process. Given their distribution within the K-phillipsite-filled the bacteria morphologies [the natural Fc(TlJ)-hydroxide and amygdules, the microbial structures represent endolithic Ti-oxide mineral phases] are not known to fluoresce when and, most likely, cryptoendolithic filamentous and sphL'rical excited with UV or visible radiation (Gaft et al., 2(05). Ty­ microorganisms. Unfortunately, because our microfossils were pica!!y, lepidocrocite, y-Fe{III}O{OH), cannot be the source of not associated with any multiple paragenetic sequences of the luminescence because of the nature of it; the type and cements (e.g., Peckmann eJ aI., 2(08), we were unable to de­ strudure of this mineral render it impracticable for emission termine precisely when the vesicles hosted life. lltC presence of luminescence by thc mineral itself even with impurities in of the purported fossil microorganisms in amygdules in close its structure. lepidocrocite can fonn, however, with external proximity to the outer margin of the pillow basalt, along with macroscopically mixed luminescent impurities. the attachment of some of the microfossils to the walls of the vesicles (Fig. 4C) and the absence of granular or tubular al­ teration textures in the walls of the vesicles and glass shards, 5.3. Chemical composition of authigenic minerals suggests that early precipitation of the phillipsite in the vesi­ consistent with microbial life cles and pore space reduced the fluid circulation and inhibited The mincralogy of the authigl'tlic precipitates, Fe and Ti the growth of euendolithic microorganisms (see Fisk et al., metal oxides, associated with the purported microfossils is 2003). The vesicular microhabitat would most likely have been consi.stent with the activities of a microbial population as­ a nutrient-rich micrrrenvimnment with the nutrients coming sociated with subseafloor environments. For example, the from alteration of the volcanic rocks. A circumneutral to iron-oxide lepidocrocite precipitates when sheathed iron­ alkaline (pH>7) fluid cnviroruncnt within the vesicles is in­ oxidizing filamentous bacteria remove iron from solution dicated by the presence of abundant TiCh in the K-phillipsitc­ (Emerson and Moyer, 2002). Though bacterial sheaths can filled amygdules. The Fe(ITJ) (oxy)hydroxides observL>d may provide a physically and chemically protective environment have been produced either chemically or biologically by mi­ in which the cells can grow and divide, once mineralized crobial and c1"lCmical iron(II) oxidation at a neutral pH they limit the diffusion of nutrients into cells. Microbial (Emerson, 2000). surfaces and extracellular polymeric substances arc known The fossilized microorganisms within vesicles of the Coral to enhance the concentrations of certain elements, such as Fe Patch pillow basalt are preserv<.-d as organic remains (car­ and Ti, relative to the concL'tltrations of thL"SC elements in the bonaceous mL'ffibrane) and mineral (K-phillipsite, Ti oxides, 630 CAVALAZZI ET AL.

    ,md Fe hydroxidcs}-rcplaccd stmcturcs. Because biogenic been found to colonize vesicles of basaltic rocks in terrestrial carbonaceous matter dL"grades progressively and presum­ envirorunents (Jorge ViUar et al., 2006). The discovery of ably continuously during mineralization and encrustation, populated vesicles of volcanic rocks from the ocean basins the presence of native fluorescence from the microfossils and and in terrestrial environments has widened the prospect of surrounding area indicates that the organic remains retain a finding new sites for these extreme environments on Earth. sufficient number of functional groups to produce a detect­ The astrobiological relevance of fossil cryptocndoliths within able signal. lh'ir morphologies and distribution arc consis­ vesicles in pillow is further emphasized by the finding of tent with a biotic origin and, though it is not possible to extrL'Tllely vesicular basalt within the Columbia Hills on the detennine exactly when they infiltratL>d the vesicles, their Gusev Crater floor (Mars Exploration Rover Spirit mission) mineral composition is consistent with early microbially (Schmidt et al., 2008). From this perspective, the cryptic mediated authigenesis in a marine environment. lneir micro­ habitat of pillow basalts and their as..<;oc iated microorganisms habitat suggests that they were chcmolithotrophic and het­ represent an, as yet, wlderOlppreciOltcd environment in whicll erotrophic cryplocndoliths Iha.! inhabited a low-temperature to seOlrch for evidence of life. circunmeutral to alkaline aqueous microhabitat. 6. Conclusions 5.5. Vesicular habitat: a suitable niche for life with high preservation potential Partly mineral-replaced (K-phillipsi te, Fe hydroxides, and Ti oxide) filaments and spherical structures occur within Life displays a wide range of survival strategies and can K-phillipsite-filled amygdules in the chilled margin of piUow adapt to even the most inhospitable habitats. Endolithic en­ from Coral Patch Seamotult, eastern North Atlantic vironments arc ubiquitous terreshial microbial habitats and, Ocean. as such, represent an important area of study in the fields Well-preserved and carbonaceous filamentous and of geomicrobiology, early life research, and astrobiology. spherical structures show low variability in size and shape. Endolithic microorganisms, including cryptoendoliths, can Their biogenicity was assessed on recognized biogenicity colonize a wide variety of substratums, especially in extreme criteria established for microfossils. In particular, the co­ envirorunents, and produce characteristic stmctures and occurrence of their cellular morphologies (biomorphic fea­ bio-a lteration textures with a high preservation potential tures, arrangement and distribution), their mineralogical and (Friedmann and Koriem, 1989; Wynn-Williams and Ed­ chemical compositions, and associated autofluOfCSCL'llCe wards, 2000). signal and envirorunental microhabitat strongly support TIle study of terrestrial analogues of potential extrater­ their biogenicity. The microbial population would have in­ restrial environments is a prerequisite for astrobiology and cluded lithoautotrophs and may have included mineralized planetary exploration. Oceanic bas.1.it rocks arc rich in re­ sheathed microorganisms, possibly Fe-oxidizing bacteria. duced elements (e.g., iron) that playa crucial role in the TIle presence of phillipsite and anatase permineralized mi­ microbiology of basalt endolithic habitats. In fact, (chemo­ crofossils, as well as anatase grains within the K-phillipsite lithoautotrophic) microbial alteration of oceanic basalts has a matrix, indicates that the environment within the vesicles significant influence on deep-sea carbon cycling and chemi­ had a circumneutral to acidic pH. cal exchange behveen basalt and seawater (Santelli ct aI., Since the vesicles are primary structures in the altered 2008). Oceanic volcanic rocks were prevalent on early Earth glass of pillow basalt, the fossil microorganisITL<; described in and probably on early Noachian Mars (Westall, 2005). Thus, this study must hOlve been cryptocndoliths, and the vesicular endolithic microorganisms would be strong candidates for habitat suggests that they were chemotrophic/lithoauto­ the colonization of early Earth and other volcanic planetary trophic organisms. This type of volcanic microenvironment surfaces (e.g., Friedmann and Koriem, 1989; Fisk e/ al., 2(06). and the cryptoendolithic microorganisms that could inhabit The microhabitats within vesicles would have provided a it need further study. Such environments, which would have UV-protected cnviromnent for microorganisms on planets, been common on early Earth and probably on early Mars, such as early Earth, that had no ozone shield (Cockell and represent analogues of astrobiological interest. Raven, 2(04). Vesicles also represent environments that would have been protected from the catastrophic effects of Acknowledgments volcanism and meteoritic impact on early Earth and Mars (Cockell et aI., 2002). However, the size, diversity, and We gratefully acknowledge N. Zitellini, A. Ceregato, and flUlCtional activity of endolithic microbes on, and within, the L. Capotondi, ISMAR-CNR Bologna for donating samples. rock substrate at the seafloor are poorly known (Edwards We acknowledge funding for B.C. from CNR-CNRS Interna­ et al., 2005). Biotic alteration morphologies, such as eu­ tional Cooperation Program, for F.W. from GDR-Exobiologie endolithic microborings in Archean pillow lavas in South grant, illld for S.L.e. from NASA lmder aWOlrd No. Africa and Western All';tralia (Fumes ct al., 2004; B.1.nerjee NNG04G]84G and NSF lmder award No. GE00808211. Thanks are duc to I. Raffi, Universita di Chieti-Pescara (Italy) c/ al., 2006) have been described and advanced as a new astrobiological biosignature (Mcloughlin ct aI., 2007). for help in fossil nannoplankton identification. N. Mclough­ Though euendolithic bioalteration of volcanic rock/glass lin, G. Glamociija, and anonymous reviewers provided useful surfaces has been well documented in the geological record comments. (review in Staudigel et al. , 2008), fossil cryptoendoliths Abbreviations within vesicles of submarine basalts are little known (Schu­ mann ct aI., 2004; Cavalazzi et al., 2008; Peckmann et al., 2008; CLSM, confocal laser scillming microscope; EOX, energy­ Eickmarul et al., 2009; this study). Cryptocndoliths have also dispersive X-ray; ESEM, environmental scalUling electron VESiCULAR HABITAT 631 microscope; FEG-SEM, field emission glm scaJming electron hydraulic binders analysed from 46- and 260-year-old build­ microscope; SEM, scanning electron microscope. ings. Corros Sci 52:3168-3179. Edwards, K.J., Bach, W., McCollom, T.M., and Rogers, O.R. (2004) Neutrophilic iron-oxidizing bacteria in the ocean: their h.1bita ts, References diversity, and roles in mineral deposition, rock alteration, and Amos, W.B. and White, J.G. (2003) How the confocal laser biomass production in the deep-sea. Ccomicrobiol j 21 :393-404. scanning microscope entered biological research. Biology of the Edwards, K.J., B.1ch, W., and McCollom, T. (2005) Ccomicro­ Cell 95:335-342. biology in oceanography: microbe mineral interactions at and Bach, W. and Edwards, K.J. (2003) Iron and sulfide oxidation below the seafloor. Trends MicrobioI13:449---456. within the basaltic ocean crust: implications for chemo­ Eickmann, 8., B.lch, W., Kiel, 5., Reitner, J., and Peckmann, ]. lithoautotrophlc microbial biomass production. Geochim Cos­ (2009) Evidence for cryptocndolithlc life in Devonian pillow moc/lim Acta 67:3871 -3887. basalts of Variscan orogens, Gennany. Palaeogeogr Palaeocli­ Banerjee, N.R and Muehlenbachs, K (2003) Tuff life: bio­ malol Palaroeco/283:120-125. alteration in volcaniclastic rocks from the Ontong Java Emerson, D. (2000) Microbial oxidation of Fe(lI) and Mn(H) at Plateau. , Geophysics, Geosystems 4, doi:10.1029/ circumneutral pH. In Etrviromlle!1lal Microbe-Minemllnteractions, 2002GC000470. edited by OX Lovlcy, ASM Press, Washington OC, pp 109-144. Banerjee, N.R, Fumes, H., Muehlenbachs, K., Staudigel, H., and Emerson, O. and Moyer, e.L. (1997) Isolation and chara cteriza­ de Wit, M. (2006) Prcscrvntion of -3.4-3.5G.1 microbi,ll bio­ tion of novel iron-oxidizing bacteria that grow at circumneu­ markers in pillow lavas and hyalodastites from the Barberton tral pH. Appl Elwiroll Microbiol 63:4781 ~ 792. GTl..>enstone Belt, South Africa. Earth Planet Sci Lett 241:707- Emerson, D. and Moyer, CL. (2002) Neutrophilic Fe-oxidizing 722. bacteria arc abundant at the Loihi seamount hydrothermal &xIard, D.L., B.1i1ey, J-j., Reiss, B.L., and Van Slyke JerLilk, G. vents and playa major role in Fe oxide deposition. Appl EII­ (2006) Development and chnractcrization of stable scdiment­ viroll Microbiol 68:3085-3093. free anaerobic bacterial enrichment cultures that dechlorinate Fisk, M.R, Storrie-Lombardi, M.e., Douglas, S., Popa, R., Arodor 1200. Appl Enviroll Microbiol 72:2460-2470. McDonald, G., and Oi Moo-Savoie, C (2003) Evidence of bi­ Beveridge, T.J. (1989) Role of cellular design in bacterial metal ological activity in Hawaiian subsurface basalts. Ceocilemistry, accumulation and in mineralization. All/III Rev Microbial Geophysics, Geosyslems 4, doi:10.1029/2002GC000387. 43:147- 171. Fisk, M.R., Popa, R, Mason, O.U., Storrie-Lombardi, M.e., and Bhartia, R, Hug, W.F., Salas, E.e., Reid, RD., Sijapati, K.K., Vincenzi, E.P. (2006) Iron-magneshun silicate bioweathering Tsapin, A., Abbey, W., Nealson, K.H., Lane, A.L., and Conrad, on Emth (and Mars?). Astrobiology 6:48-68. P.G. (2008) Classification of organic and biological materials Friedmann, E.!. and Koriem, A.M. (1989) : how it with deep ul traviolet excitation. A,'pl Spectrose 62:1070-1077. disappeared (i f it was ever there). Adv Space Res 9:167- 172. Bhartia, R, Sa la s, E.e., Hug, W.F., Reid, R.D., Lane, A.L., Ed­ Fumes, H. and Staudigel, H. (1999) Biological mediation in wards, K.J., and Nealson, K.H. (2010) Label-free bacterial ocean crust alteration: how deep is the deep biosphere? Earlll imaging with deep-UV-laser-induced native fluorescence. Planet Sci Lett 166:97- 103. Appl Environ MicrobioI76:7231-7237. Fumes, H., Muehlenbachs, K., Torsvik, T, Tumy r, 0., and L1ng, Brasier, M., Mcloughlin, N., Green, 0., and Wacey, D. (2006) A S. (2002) Bio-signaturcs in metabas.1Jtic glass of a Caledonian fresh look a t the fossil evidence for early Archaean cellular life. ophiolite, West Norway. Gool Mng 139:601---608. Phi/os Trans R Soc Lmd B Bioi Sci 361:887- 902. Fumes, H., B.merjee, N.R, Muehlenbachs, K., Staudigel, H., and Buick, R (1990) Microfossil recognition in Archean rocks: an ap­ de Wit, M. (2004) E.1rly life recorded in Archean pillow lavas_ prais.1.I of spheroids and fila.ments from a 3500 myoid chert­ Science 304:578-581. barite unit at North Pole, Western Australia _ Palaios 5:441-459. Fumes, H., Banerjee, N.R., Staudigel, H., Muehlenbachs, K., Cady, S.L., Fanner, J.D., Grotzinger, J.P., Schop f, J.W., and Mcloughlin, N ., de Wit, M., and Van Kranendonk, M. (2007) Steele, A (2003) Morphological biosignatures and the search Comparing petrographic signatures of bioalteration in recent for life on Mars. Astrobiology 3:351-368. to Mcso.lrchean pillow la vas: tracing subsurface life in oceanic Caval,lZzi, B., Westall, F., and B.lrbieri, R (2008) (Crypto-)En­ igncous rocks. Precambrian Res 158:156-176. doliths from vesicular pillow lavas, Coral Patch Seamount Gaft, M., Rei"feld, R., and Panczer, G. (2005) Modem Lllmines­ North Atlantic Ocean. Studi Trent Sci Nat Acta Geol83 :1 77-182. ceHce Spectroscopy of Minerals and Mnterials, Springer, Berlin. Chan, C.S., Fakra, S.C, Emerson, D., Fleming, E.j., and Edward~, Garcia Ruiz, ].M., Camerup, A.M., Christy, A.G., Welham, N.j., K.]. (2011) Li thotrophlc iron-oxidizing bacteria produce or­ and Hyde, S.T. (2002) Morphology: an ambiguous indicator of ganic stalks to control mineral growth: implications for bio­ biogenicity. Astrobiology 2:353-369. signature formation_ lSME / 5:717- 727. Geldmacher, J. and Hoemle, K. (2000) The 72Ma gecl(:hemical Cockell, e.S. and Herrera, A. (2008) Why arc some microor­ evolution of the Madeira hotspot (eastern North Atlantic): gan i ~m~ boring? Trends MicrobioI16:!01- 106. recycling of Paleozoic ( :S: 5OJMa) oceanic lithosphere. Earlll Cockel!, e.S. and Raven, J.A. (2004) Zones of photosynthetic Planet Sci Lett 183:73-92. potential on Mars and early Earth. learns 169:300-310. Glamoclija, M., Steele, A., Fries, M., Schieber, J. , Voytek, M.A., and Cockell, C.S., Lee, P., Osinsky, G., Homeck, H., and Bro.ldy, P. Cocket!, e.S. (2009) Association of anatase (Ti0.!) and microbes: (2002) Impact-induced microbial endolithic habitats. Meteorit unusual fossilization effect or a potential biosignature? Geolo­ Planel Sci 37:1287- 1298. gical Society of America Special Papers 4.58:%5-975. Connell, L., Barrett, A, Templeton, A, and Staudigel, H. (2009) Gorobets, B.S. and Rogojine, A.A., editors. (2002) Llllllillescellt FWlgal diversity aS50Ciatcd with an active deep sea volcano: Spectra of Minerals Reference·Book, Coronet Books Inc., l'hila­ Vailulu'u Seamount, Samoa. Gevmicrobiol j 26:8597-8605. delphia, PA, USA. Demoulin, A., Trigance, e., Neff, D., Foy, E., Dillmann, P., and Ivarsson, M., L1usmaa, L Lindblom,S., Broman, C, and Holm, L'Ho~tis, V. (2010) The evolution of the corrosion of iron in N.G. (2008) Fossilized microorganisms from the Emperor 632 CAVALAZZI ET AL.

    SeamOlUlts: implications for the search for a subsurface fossil Schopf, 1.W., Kudryavtsev, AB., Agresti, D.G., CZJja, AD., and record on Earth and Mars, Astrobiology 8:1139- 1157. WdowiJk, T.J. (2005) Raman ilThlgery: a new appro.1Ch to as­ Jehlinca, J. and Beny, C. (1992) Application of Raman micro­ sess the geochemical maturity and biogenicity of penniner­ spectromehy in the study of structural changes in Precambrian alized Prrolmbrian fossils. Astrobiology 5:333--371. kerogens during regional metamorphism. Drs Geochcm 18:21 1-213. Schopf, 1.W., Tripathi, AB., and Kudryavtse\', AB. (2006) Three­ Jorge Villar, 5,£', Edwards, H.GM., and Benning. LG. (2006) dimensional confocal opticJI imagery of Precambrian micro­ Raman spectroscopic and scanning electron microscopic scopic organisms. Astrobiology 6:1 - 16. analysis of a novel biological colonisation of volcanic rocks. Schopf, J.W., Kudryavtsev, A.B., Czaja, AD., and Tripathi, A.B. Icarus 184:15R-169. (2007) Evidence of Archean li fe: stromatolites and microfos­ Kappler, A. and Straub, K.L (2005) GeomicrobiologicaJ cycling sils. Precambrian Res 158:14 1-155. of iron. Reviews in Mineralogy and Geochemistry 59:85-108. Schopf, J.W., Kudryavtscv, AB., and Scrgeev, Y.N. (2010.1) Kushcfi, K. and Lovlcy, ox (2003) Extending the upper tem­ Confocal laser scanning microscopy and Raman imagery of perature limit for li fe. Science 301 :934. the late Neoproterozoic Chichkan microbiota of South Ka­ Kim, J., Dong, H.L, Seabaugh, J., Newell, S.W., and Eberl, D.O. zakhstan. J Palmntol 84:402-416. (2004) Role of microbes in the smectite-to-illite reaction. Sci­ Schopf, J.W., Kudryavtsev, AB., Sugitani, K, and WJlter, M.R ellce 303:830-832. (2010b) Precambrian microbe-like pscudofossils: a promising Knoll, A. (1999) Recognition of a biologicill signature in rocks. solution to the problem. PrecambrinII Res 179:191 - 205. Discussion summary. In Siu Limits of Very Small Microorgan ­ Schumann, G., Manz, W., Reimer, j., and Lustrino, M. (2004) isms: Proceedillgs of a Worksllop, Steering Group for the Work­ Ancient fungal life in North Pacific Eocene oceanic crust. shop on Size Umits of Very Small Microorganisms, National Gmmicrobiol J 21:241- 246. RcscJrch Council, The NJtionJI Academies Press, Washington Shabtai, Y. and Fleminger, G. (1994) Adsorption of RhodococClls DC, pp 85-87. strain GIN-1 (NCJMB 40340) on ti tanium dioxide and coal ny Konhauser, K.O. (1998) Diversity of bacterial iron mineraliza­ ash particles. Appl Environ Microbiol 60:3079- 3088. tion. Eartll-Sciellce Reviews 43:91-121. Staudigel, H., fumes, H., McLoughlin, N., and Banerjee, N.R. Kostka., J.E., Dalton, D.O., Skelton, H., Dollhopf, S., and Stucki, (2008) 3.5 billion years of gL1S5 bioalteration: volcanic rocks as J.W. (2002) Growth of iron(III)-rcducing bacteria on clay b.lsis for microbial life? Earth-SciCllc/? Reviews 89:1 56-176. minerals as the sole electron aa:eptor and comparison of Stroncik, N.A. and Schminckc, H.-U. (2001) Evolution of growth yields on a variety of oxidized iron form.;. Appl En­ palagonite: crystallization, chemical changes, and elemental viron Microbiol 68:6256-6262. budget. Geochemistry, Geophysics, Geosystems 2, doi:1O.1029/ Lletsch, T.A and Downs, R.T. (2006) Software for identification 200xx:000102. and refinement of cell pJrameters from powder diffraction Summons, R.E, Jahnke, L.L, Hope, M., and Logan, G.A. (1999) 2- data of minerals using the RRUFF Project and American Methylhopanoids as biomarkers for cyanobacterial oxygenic Mineralogist Crystal Structure Databases [abstract PQ8.25[. In photosynthesis. Natllre 400:554- 557. Program and Abstracts of tile 19J'r Gellaal Meetillg of tire IlIta­ Takai, K., Moser, D.P., DeFlaWl, M.F., Onstott, T.C., and lIational Mineralogical Association, Intemational MineralogiCJI Fredrickson, J.K (2001) Archaeal diversity in waters from Association. deep South African gold mines. Appl Ellviron Microbiol 67: Marshall, c.P., Edwards, H. G.M., and Jehlinca, J. (2010) Un­ 5750-5760. derstanding the application of Raman spectroscopy to the Thorseth, I.H., Torsvik, T., Fumes, H., and Muehlenbach.;, K detection of traces of life. Astrobiology 10:229-243. (1995) Microbes play an important role in the alteration of Mcloughlin, N., Brasier, M.D., WJcey, D., Green, O.R., and oceanic crust. Chrm CeoI126:137- 146. Randall, S.P. (2007) On biogenicity criteria for endolithic mi­ Thorseth, I. H., Torsvik, T., Torsvik, Y., Daae, F.L., and Pedersen, croborings on early £.lrth and beyond. Astrobiology 7:10-26. R.B. (2001) Keldysh-98 scientific pilrty, diversity of life in McLoughlin, N., Fumes, H., Banerjee, N.R., Muehlenbachs, K, ocean floor basalt. Earth Plmlet Sci L£tt 194:31- 37. and Staudigel, H. (2009) Ichnotaxonomy of microbial tmce van Yeen, W.L., Mulder, E.G., Jnd Deinema, M.H (1978) The fossils in volcanic gL1SS. J Ceol Soc London 166:159-169. Spliaerotillis-Leplotilrix group of b.1cteria. Microbiol Rev 42: Nielsen, M.E. and Fisk, M.R. (2010) Surface area measurements 329- 356. of marine basalts: implications for the subsea floor microbial Westall, F. (2005) Early life on Earth and analOgies to Mars. biomass. Ceopllys Res Lett 37, doi:Ul.1029/20l0GL044074. In Water 011 Mars and Life, Advances in Astrobiology and Bio­ Peck, D.L. (1978) Cooling and vesiculation of Alae LlYa Llke, geophysics 4, edited by T. Tokano, Springer, Berlin, pp 45-64. Hawaii. U.s. GeologicJI Survey Professional PJpcr 935-B, u.s. Wynn-Williams, D.O. and Edwards, H.G.M. (2000) Antarctic Geological Survey, Washington DC. ecosystems as models for extraterrestrial surface habitats. Pedmann, J., Bach, W., Behrens, K., and Reimer, J. (2()()8) Pu­ Planet Space Sci 48:1065--H175. tath'e cryptoendolithic life in Devonian pillow basalt, Rhein­ Address correspondence to: isches Schiefergebirge, Germany. Geobiology 6:125-135. 5.1ntelli, C.M., Orcutt, B.N., Banning, E., B.lch, W., Moyer, c.L., Barbara 0walazzi Sogin, M.L., Staudigcl, H., and Edwards, K.J. (2008) Abun­ Department of Geology dance and diversity of microbial life in ocean crust. Nature University of Jollannesburg 453:653-656. Johannesburg Schmidt, M.E., Ruff, S.w., McCoy, T.]., Farrand, W.H., Johnson, So,lIh Africa 1.R, Gellert, R, Ming, D.W., Morris, RY., Cabrol, N., Lewis, E-mail: [email protected] KW., and Schroeder, C. (2008) Hydrothermal origin of halo­ gens at Home Plate, Gll'leV Crater. J Grophys Res 113, Submitted 26 March 2011 doi: 10.1029 / 2007JEoo3027. Accepted 2 July 2011