TERRESTRIAL PLANET SERPENTINIZATION IN MELANGE SETTINGS

Mars and Earth have planetary crusts with exposures of mantle-derived ultramafic rocks dominated by pyroxene and olivine. These initial minerals transform to serpentine, brucite, magnetite, and other secondary phases via metamorphic hydration and other reactions. In this work, altered peridotites of the regionally extensive Coast Range Ophiolite (CRO), from localities in the UC-Davis McLaughlin Natural Reserve, Lower Lake, CA are compared and contrasted with serpentinites of the Nili Fossae, a mélange terrain located in the Syrtis Major quadrangle at approximately 22°N, 75°E, Mars. The habitability of serpentinizing systems is conveyed as a function of changing system parameters (such as temperature, Eh, pH, and activities of aqueous geochemical species), and provides insight into the biological prospects of serpentinization in mélange terrains in a general sense. Petrography of the serpentinized Jurassic age ultramafic unit in the Coast Range Ophiolite confirms the dominance of secondary phases (serpentine, other clays, carbonates, magnetite) and presence of relict primary minerals (olivine and pyroxene). Major element concentrations for crystals of olivine (from McL-239A and McL-329), show concentrations of MgO (avg: 48.9 wt%), SiO2 (avg: 40.8 wt%) and high FeO (avg: 10.2 wt%) that are high relative to San Carlos, Olivine (standard). Backscatter electron images (BSE) of such crystals in McL-239A and McL-329 show relict olivine surrounded by a serpentine-rich matrix. The crystal chemistry of these representative samples of olivine from the CRO serve as an analog for the olivine-rich protolith that underwent serpentinization observed in the Nili Fossae. Using CRO data to construct a model protolith reasonable for ultramafics of the Nili Fossae mélange (constrained by CRISM observations), I deduce evolving habitability as the model protolith reacts with feasible, co-occurring fluid chemistries. Major aqueous geochemical compositions are based on based on postulated planetary analog waters that are largely Na-Cl, Mg-Cl, or Ca-Cl solutions. The modeled water-rock reactions were performed at conditions associated with both CRO and the Nili Fossae mélange settings. Habitability was assessed using a Gibbs Free Energy minimization strategy, for serpentinization-driven methanogenesis (MG) and methanotrophy (MT). I show that the bioenergetic yields of fundamental methanogenetic and methanotrophic reactions progress favorably between -120 kJ/mol to a maximum of -400 kJ/mol as serpentinization progresses under different groundwater/hydrothermal conditions in mélange terrains.

Initial water compositions for Na-Cl ocean were based on Glein et al. (2015) for pH and activity of carbon dioxide (aCO2=10 -9 ). 4 The Na-Cl ocean is based on modeling computations with carbon dioxide partial pressure of 0.0003, averaged from Earth ocean chemistry (Hem, 1985).      Specifically, with the oxidation of Fe 2+ in olivine coupled to the reduction of hydrogen in H2O to free H2, biologically useful hydrogen is produced, and multiple known microbial metabolisms can be powered. Methanogenesis, methanotrophy, and metal reduction reactions coupled to hydrogen oxidation are key among them. These are likely to be common reactions in (bio)geochemical systems in ultramafic rocks, given the presence of liquid water.
On Earth, varied tectonic environments produce ultramafic rock assemblages.
Mantle minerals undergo hydrolysis at temperature, pressure, and compositional conditions present in the crust and upper mantle. This location, where Earth's pressures and temperatures are in the constrained range of less than the pressure-temperature resulting in the instability of antigorite, a serpentine mineral polymorph stable to approximately 950K and 5.0 Gpa (Wunder et al., 2001), provides a lithospheric limit on habitability of serpentinization. Serpentinization is considered on Earth to be a possible geological driver of the establishment of life on Earth (Schulte et al., 2006;Sleep et al., 2011) and there is a diversity of deep life tied to serpentinites in the terrestrial seabed  and low-temperature hydrothermal vents, as at Lost City Hydrothermal Vent Field (Kelley et al., 2005). At this site, observed seawater and upper mantle peridotite react and produce serpentine polymorphs, accessory minerals, and biologically relevant aqueous species dissolved in the fluids (Kelley et al., 2005).
Biological diversity at this location is significant (Brazelton et al., 2006). In continental serpentinization systems, impacts of the same class of water-rock reactions can be life on other rocky celestial bodies (Ehlmann et al., 2010;Boye Nissen et al., 2015).
Ultramafic rocks are common rock units among the terrestrial celestial lithospheres, such as Mars, Earth, and the Moon. This makes terrestrial planet lithospheres likely candidates to support astrobiological chemistries and exhibit non-equilibrium characteristics that present an interesting location for biological processes and even organisms to exist . In addition, rocky planets are large energy and chemical reservoirs to drive astrobiology and specifically, are sites where fluid-rock interactions can occur and be used to understand the habitability conditions of these interactions through specific thermodynamic constraints (see Scambelluri et al., 2004).
There is exciting scientific discussion regarding serpentine formation in terrestrial planets and icy satellites across multiple hypotheses and current research, in both terrestrial analogs (Preston & Dartnell 2014;Foing et al., 2011) and extraterrestrial environments for favorable conditions for serpentinization. These research foci include questions such as the presence of ultramafic rocks interacting with sub-ice oceans in icy-satellite systems of Enceladus and Europa (Russell et al., 2014), the interaction of Martian ultramafics with an aqueous phase in the production of serpentine minerals (Holm et al., 2015;Ehlmann et al., 2010), and serpentinization reactions on smallbodies, such as asteroids (Nathues et al., 2014) and comets (Holm et al., 2015). In the Coast Range Ophiolite Microbial Observatory (CROMO) in the McLaughlin Natural Reserve, California, ophiolite-hosted peridotite has undergone serpentinization (Cardace et al., 2013), and is present in a mélange composed of serpentinite blocks, disordered exotic and country rock clasts and inclusions, surrounded in a fragmentary and fine grain matrix (Moody, 1976;Hopson et al., 1981;Shervais & Kimbrough, 1985) gabbro and marine sedimentary sequences are regionally common also. The CROMO serpentinites are useful extraterrestrial analogs for altered ultramafic material in planetary mélange terrains because of the appropriate mineral assemblages and complicated tectonic and geological history of the location, resulting in a varied petrology and structure.
Mars has long been a target of astrobiological research focusing on the emplacement of ultramafic rock units undergoing serpentinization (Schulte et al., 2006;Michalski et al., 2013). Ehlmann et al. (2010) (Moore et al., 2003;. In the Claritas Rise (Figure 1), mineralogy includes serpentine, chlorite, kaolinite, and illite or muscovite (Ehlmann et al., 2010). Near the Nili Fossae ( Figure 1 and 2), the dominant mineralogies are serpentine, smectite, kaolinite, low-Ca pyroxene, and olivine. Other observed serpentine deposits have been identified in a small amount of impact craters, in the morphological features of the walls, ejecta, and the central peaks, in the Chia crater (Ehlmann et al., 2010). These deposits are also temporally correlated to the Noachian period of the planet's geological history, with minerals present including chlorite and smectite (Ehlmann et al., 2010). In addition, serpentine has been observed in geological settings that are very olivine-rich. These sites are located in a location where there is Early Hesperian Syrtis Major volcanic activity, representing the unit of high levels of olivine--and the underlying stratum is Noachian, with co-occurring olivine, carbonates, and serpentine (Ehlmann et al., 2010).
Furthermore, the carbonate has a distinctive absorption feature of magnesium carbonate (Ehlmann et al., 2008). The prevalence of olivine rich units along with low-Ca pyroxene indicate that the Nili Fossae is likely an ultramafic hosted rock unit (Ehlmann et al., 2010), however specific geochemistry of this area is unavailbe at the resolution necessary associated with localized habitability. To provide a first order approximation

The Coast Range Ophiolite: terrestrial site of serpentinization.
The Coast Range Ophiolite, California, USA has been the subject of diverse geophysical, geochemical, and geological studies (e.g., Hopson et al., 1981;Shervais et al., 2004;Choi et al., 2008). Over the past four decades, investigating the underlying geodynamics of the Coast Range Ophiolite (CRO) has resulted in three theories to explain its formation. First, the units of the CRO have been proposed as sourced from a mid-ocean ridge, based largely on stratigraphy (Hopson et al., 1981).
Second, the CRO has been proposed as a product of back-arc basin development, based on regional structural geology (Ingersoll, 2000). Third, CRO units have been interpreted as formed in a supra-subduction zone setting, specifically due to active fore-arc spreading, based on whole rock and mineral geochemistries (

Thin Section Petrography
Traditional optical petrography was employed to identify major mineral phases, to assess degree of alteration and deformation (if any), and to screen samples for relict

X-Ray Powder Diffraction
Samples were analyzed for semi-quantitative, bulk mineralogy, using x-ray powder diffraction (XRD) (see figures 54-97). Sample preparation followed standard procedures outlined by Cardace et al. (2013). In brief, samples were powdered with a percussion mortar (URI Dept. of Geosciences) and/or a mortar and pestle. Powders were then passed through a 150 µm pore size sieve, until approximately 15.0 mg of sample was collected, and transferred into the XRD sample chamber. Samples were analyzed on an Olympus Terra portable x-ray diffractometer with a Cobalt (Co) X-Ray tube with a charge-coupled detector (CCD) (Blake et al., 2012). The Co anode instrument parameters during analyses (10 W, 30 kV, and between 5-25 keV) were maintained throughout 1000 exposures. The diffractograms were interpreted using XPowder software, which references databases for mineral peak d-space identification, and allows for peak characterization. XRD diffractograms of clay minerals, including serpentine and kaolinite, exhibt 7-angstrom spacing which can cause major peak overlap, when interpreting the identification of the diffractogram, thus it is necessary to utilize more than one peak in the identification process using the XRD diffractogram.

Inductively Coupled Plasma Atomic Emission Spectroscopy
Serpentinites were prepared for analysis on the Thermo Scientific iCAP 7400

Electron Microprobe Analysis
Electron microprobe analysis (EMPA) was conducted at Brown University in Providence, RI on a Cameca SX-100 electron microprobe analyzer to identify major mineralogical phases and identify specific mineral chemistries for relict and neoformed (via metamorphism) grains (N=4). In situ, non-destructive analysis of the mineral veins was also carried out. Standard probe analysis was used for the system parameters of a

Data Validation and Statistical Approaches
For the ICP-AES and electron microprobe analyses, data obtained were normalized through the goodness of fit of calibration curves using linear regression and R 2 values to determine the relation between standardized values and observed values of both the bulk and mineral chemistry values (Rollinson, 2014). In addition, the mean of drift and blank count intensity values were utilized in the ICP-AES data reduction and correction.
Regarding XRD data, conventional use of XPowder peak-matching software allowed background subtraction and peak identification. Use of the semi-quantitative modeling tool in XPowder further allowed modeling of diffractograms as mixtures of ideal mineral endmembers; for internally consistent sample comparisons, XRD profiles were considered as model mixtures of selected reference olivine, pyroxene, lizardite, antigorite, etc. Modeled semi-quantitative abundances were then plotted in Excel.

Modeling Framework
To model the water-rock reaction paths located at the geophysical sites of  (Michalski et al., 2013).
Bioenergetic modeling (as in Amend and Shock, 2001;Cardace et al., 2015) utilizes model outputs from Geochemist's Workbench REACT modeling to constrain the ion activity product (Q) needed to solve for Gibbs Free Energy (G). When the thermodynamic system is at constant temperature and pressure, the Gibbs Free Energy for a chemical reaction in the system depends on the extent of the chemical reaction, Xi or ξ, and at system chemical equilibrium the change in Gibbs Free Energy over the extent of reaction equals zero. Of a reaction system, the equation to result at a G value, used in this work, is In this way, the thermodynamic feasibility of selected reactions can be determined, with a negative value resulting in a spontaneous reaction. Here, the methanogenesis

Thin Section Petrography
Representative thin sections (N=16; Table 4)  Serpentinized olivine and orthopyroxene were distinguished based on retained crystal structure: serpentinized olivine retains a rounded grain mass, and orthopyroxene retains lamellae. Further, in terms of birefringence, olivine exhibits second order interference colors and orthopyroxene exhibits first order interference colors (figure 4a). Clinopyroxene was determined through structure and birefringence, with coarse lamellae and higher order interference colors, with respect to orthopyroxene.
Magnetite was identified as an opaque phase that retains its opacity using cross polarization and in plane polarized light. Clay minerals were identified as thin sheets that could not be individually identified at the resolution of the petrologic investigation. Of the representative core samples analyzed, between the two drilling operations, the Homestake cores showed elevated amounts of relict primary minerals.
The CROMO cores exhibited 60-70% olivine-textured serpentine and 5-10% relict minerals that have yet to become serpentinized. Orthopyroxene-derived textures and relict minerals were observed between 5-20%. Clinopyroxene was not observed in the CROMO cores at the resolution of the analysis. Magnetite and clay minerals were also present in these samples, where magnetite was present in volume percent of 5-25% and 5-15% volume was reported as clay mineralogies. The Homestake cores exhibited higher percentages of olivine and serpentinized olivine structures than most of the CROMO cores and included more relict primary olivine and orthopyroxene crystals.
Clinopyroxene was observed in only two of the representative core samples, identified with lamellae and second-order blue interference colors. The Homestake cores also contained 10% magnetite and between 0-10% clay minerals.
Mean mineralogies were calculated for major rock type determination, in addition to the analyzed slides, using the peridotite-pyroxenite ternary phase diagram.
The representative samples plot in the lherzolite stability field predominantly, with one sample (CSW1,1_15_12-5) determined as a harzburgite-lherzolite. McL-2 mini core and McL-1 mini also was on the harzburgite-lherzolite, however both were greater than 75% within the stability field of lherzolite.

X-Ray Powder Diffraction
Representative samples were (N=40) were analyzed using X-ray powder

Inductively Couple Plasma Atomic Emission Spectroscopy
Representative subsamples of altered peridotite from the Coast Range Ophiolite surface samples, CROMO drilling cores, and the M81-313 Homestake core at McLaughlin were analyzed via ICP-AES for major elements (Al, Ca, Cr, Fe, K, Mg, Mn, Na, Ni, P, Si, Ti). These values were obtained in counts per second and were converted to concentration (Table 4) and then finally, represented at major element oxides (

Olivine
Olivine was identified by the high relief crystal structure and the zonation of metamorphism to serpentine, and characterized using energy dispersive spectroscopy

Orthopyroxene
Orthopyroxene was identified by using energy dispersive spectroscopy (

Serpentine
Serpentine was identified by using energy dispersive spectroscopy (EDS) at  for the EMP analysis (10 -9 wt%). CRO olivines were modeled as mean major element values obtained in the EMP analysis.

Initial seawater inputs
Terrestrial-based seawaters reaction were constrained by two boundary conditions; temperature (273 K and 373 K) and pressure schemes (1 bar and 75 bar).
The water to rock ratios were constrained as 1:10, 1:1, and 10:1. The seawater chemistry is based on modeling computations with carbon dioxide partial pressure fugacity of 0.0057 and an oxygen partial pressure fugacity 0.000876, representing the Martian major atmospheric constituents important in this modeling calculated through the partial pressures based on pressure from De Pater and Lissauer (2015). For CRO habitability in Earth-based rocks, seawater ocean is based on modeling computations with carbon dioxide partial pressure of 0.0003, averaged from ocean chemistry (Hem, 1985).

Serpentinization of CRO olivine with Na-Cl Seawater
The CRO olivine reacting with Na-Cl seawater solution generated methane aqueous species value of 44.50 mg/kg solution at 373 K ( Figure 57). Reactions were also modeled for the seawater solution at 273.15K for the entire reaction, a temperature common on Mars' surface and a methane aqueous species value of 42.83 mg/kg sol ( Figure 56). Indicating that CRO olivine deposits that underwent serpentinization would have produced more methane if the olivine was reacting with a higher system temperature, either higher fluid or a higher mineral reactant temperature. Major mineralogies produced in both simulations was antigorite, magnetite, and brucite, with minor amounts of greenalite and tephroite, and many possible minerals in abundance of less than .1% weight percent abundance. temperatures. Manganese brucite is only produced in the lower-temperature serpentinization system, as in the Na-Cl seawater system. Energy values, with 373 K being more favorable for both MG and MT in the Mg-Cl system, also.

McLaughlin Natural Reserve Mélange Terrain: ultramafic rock provenance puzzle
Expanses of ophiolite sequences span diverse spatial and temporal settings that are associated with major orogenies and the exact nature of the specific tectonic and magmatic formation of ophiolites and, more broadly oceanic lithosphere, has been investigated structurally, stratigraphically, and geochemically (see Dilek (Hopson et al., 1981), a back-arc basin setting has been based on regional structural geology (Ingersoll, 2000), and a supra-subduction zone setting with active fore-arc spreading has been based on unit-by-unit rock geochemistry (Choi et al., 2008).  (Choi et al. 2008), and prior work on serpentinized ultramafic units has suggested a supra-subduction setting (Choi et al. 2008), new data from this study suggest that McLaughlin Natural Reserve locality ultramafics are SiO2-poor and MgO-enriched, thus suggesting depleted mantle source and/or mantle plume. Further geochemical analysis and modeling resolution are required to determine the mantle source. A broader sample set could provide richer data for the major element and REE chemistry to clarify whether these rocks have a greater affinity for island arc or supra-subduction zone tectonic settings, or whether they represent a chaotically mixed landscape that integrates rocks of similar mineralogy but sourced from discrete tectonic settings. Future work on the subsurface at the CROMO site is also required, using structural and near-surface geophysical techniques to better characterize the subsurface habitability of mélange terrain   ( Lowenstein et al., 2001), could change the thermodynamics of the ultramafic rock-seawater in the reaction simulations and thus change the habitability of the fluids flowing through the serpentine veins. Future work will also look at the serpentinization at another hypothesized location of the Coast Range Ophiolite, the Cedars ultramafic mass, located in Sonoma, California (Suzuki et al., 2013), specifically the solid-phase geochemistry and mineralogies, in the context of habitability.

Nili Fossae through time (a shergottite-based analysis)
The lherzolitic shergottites reacting with the model Na- show similar aqueous species activity, the mineralogy changed with respect to the oxide manganese brucite, where tephroite, the manganese olivine variety was thermodynamically favored, which is also is exhibited in the Na-Cl seawater reaction systems. Aqueous methane produced in the 273 K system, 41.47 mg/kg solution, and in the 373 K system 41.61 mg/kg solution, and aqueous hydrogen produced in the 273 K system, 1179 mg/kg solution, and 373 K system, 1907 mg/kg solution, suggests lherzolitic olivine rock deposits that undergo serpentinization in a Ca-Cl seawater, would have produced more methane and hydrogen if the olivine was reacting with seawater at higher temperatures, as in the Na-Cl seawater system. As in the Na-Cl seawater system, the fugacity of hydrogen increased in the changing temperature regime, with 273 K hydrogen gas having a fugacity of 611.8 and the 373 K hydrogen gas having a fugacity of 1214. Methane gas released also increased in fugacity between the 273 K and 373 K system, going from a methane fugacity of 1.018 to 3.041, respectively, which is less than the Na-Cl system. This could be due to the thermodynamic favorability of the aqueous system being a preferred phase for hydrogen and methane in the Ca-  -3.50 fugacity 0.0057 fugacity 0.0057 fugacity Table 2 : Showing the model inputs for the water-rock reactions and habitability. 1 Initial water compositions for Na-Cl ocean were based on Glein et al. (2015) for pH and activity of carbon dioxide (aCO2=10 -9 ). 4 The Na-Cl seawater ocean is based on modeling computations with carbon dioxide partial pressure of 0.0003, averaged from ocean chemistry (Hem, 1985  , and mining cores (Homestake cores) at a maximum depth of 108 meters. Phase estimates were obtained through point counting minerals in the sample using the 2x objective and then confirmed at the 10x objective. The use of a 530-nanometer retardation plate and crossed-polar light was essential to identify minerals.        . Mean mineralogies calculated for major rock type determination, using the peridotite-pyroxenite ternary phase diagram. The representative samples plot in the lherzolite stability field predominantly, with one sample (CSW1,1_15_12-5) determined as a harzburgite-lherzolite. McL-2 mini core and McL-1 mini also was on the harzburgite-lherzolite, however both were greater than 75% within the stability field of lherzolite.                                   Figure 42. MgO vs SiO2 plot showing the placement of various geochemically derived tectonic settings associated with ophiolites. Values are sparsely located but fall generally in a depleted silica and enriched magnesium zone. Few points plot in the supra-subduction zone (SSZ), backarc to forearc (Backarc to Forearc), forearc (SSZ Forearc), oceanic backarc (SSZ Oceanic Backarc), and continental backarc (SSZ Continental Backarc), which are mean values from Dilek & Furnes (2011). Mean values plot in a significantly lower silica zone than these representative types. Furthermore, these samples also plot lower in silica than depleted mantle (Depleted Mantle) and mantle plume.