Volume 118, Issue 10 p. 2083-2104
Regular Article
Free Access

Assessment of environmental controls on acid-sulfate alteration at active volcanoes in Nicaragua: Applications to relic hydrothermal systems on Mars

Brian M. Hynek

Corresponding Author

Brian M. Hynek

Department of Geological Sciences, University of Colorado, Boulder, Colorado, USA

Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, Colorado, USA

Corresponding author: B. M. Hynek, Department of Geological Sciences, UCB 399, University of Colorado, Boulder, CO 80309, USA. ([email protected])Search for more papers by this author
Thomas M. McCollom

Thomas M. McCollom

Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, Colorado, USA

Search for more papers by this author
Emma C. Marcucci

Emma C. Marcucci

Department of Geological Sciences, University of Colorado, Boulder, Colorado, USA

Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, Colorado, USA

Search for more papers by this author
Kara Brugman

Kara Brugman

Department of Geological Sciences, University of Colorado, Boulder, Colorado, USA

Laboratory for Atmospheric and Space Physics, University of Colorado, Boulder, Colorado, USA

Search for more papers by this author
Karyn L. Rogers

Karyn L. Rogers

Geophysical Laboratory, Carnegie Institution of Washington, Washington, District of Columbia, USA

Search for more papers by this author
First published: 12 September 2013
Citations: 30


[1] A variety of secondary mineralogies has been detected on Mars from both orbiters and landers, indicating widespread aqueous alteration of the crust. Many of these locales exhibit sulfates, which in some cases imply acidic fluids. At present, there are few constraints on the paleoenvironmental conditions that existed during formation of the widespread and diverse classes of secondary minerals on Mars. We investigated hydrothermal systems at three active acidic volcanic systems in Nicaragua, including Cerro Negro, Momotombo, and Telica. The recently erupted materials are similar in composition to the Martian crust and are undergoing extensive acid-sulfate alteration predominately in gas-dominated settings (fumaroles). We characterized the secondary mineralogy and local variables, including temperature, pH, rock and gas composition, and fluid-rock ratio. We find that these environmental parameters exhibit strong controls on the alteration mineralogy. The environments studied include pH that ranged from −1 to 6, temperatures from ambient to hundreds of degrees Celsius, and fumaroles to hot springs. The hottest and most acidic systems contained sulfur, silica, and minor gypsum, while moderately acidic and cooler fumaroles included abundant silica, gypsum and other hydrated sulfates, and phyllosilicates. A setting with a higher fluid-rock ratio but similar temperature and acidity was dominated by phyllosilicates and, to a lesser degree, sulfates. The characterization of aqueous alteration of basalts under a variety of environmental conditions provides a conceptual framework for interpretation of similar relic environments on Mars. Finally, while identification of phyllosilicates on Mars is generally thought to require neutral to alkaline fluids, we documented significant formation of these minerals in the acidic volcanic systems.

Key Points

  • We detailed acid-sulfate alteration in Nicaraguan volcanoes
  • Environmental controls strongly affect resultant alteration mineralogy
  • The results provide a framework for interpreting similar systems on Mars

1 Introduction

[2] Viking-era measurements determined that the global makeup of the crust of Mars is basaltic. Results from the Infrared Thermal Mapper showed little compositional diversity [Christensen, 1982], consistent with results from the two Viking landers [e.g., Baird et al., 1976] and the 1996 Pathfinder rover [McSween et al., 1999]. The 1997 Mars Global Surveyor mission's Thermal Emission Spectrometer (TES) provided a similar look at Mars—a global crust of only two compositional types in low-dust regions: a basaltic signature that covers the southern half of the planet, while in the northern plains, the spectra are best matched by andesitic basalt or weathered basalt [Bandfield, 2002; Wyatt and McSween, 2002]. TES did find three small occurrences (total of <0.5% of Mars’ surface area) consistent with a signature of gray hematite, indicating significant past water interaction at these locales [Christensen et al., 2001]. The largest of these is about 1 × 105 km2 in area [Hynek, 2004] and was chosen as the 2003 Mars Exploration Rover (MER) Opportunity landing site, in part because it was one of the only places on Mars that showed mineralogy consistent with past water. All of these results implied a volcanic surface that has had little interaction with fluids through the billions of years of history.

[3] One of the most significant paradigm shifts in planetary science in the last 8 years has been the recognition of abundant water-altered minerals globally distributed across varied geologic settings on Mars. In stark contrast to the earlier results, a number of recent and ongoing missions have revealed a mineralogically diverse surface of Mars that provides strong evidence for prolonged fluid-rock interactions. The Mars Exploration Rovers (MERs) have returned close-up images of exposed bedrock from Meridiani Planum and Gusev Crater, as well as data on their chemical and mineralogical compositions [e.g., Squyres et al., 2004, ; Yen et al., 2007; Ming et al., 2008; Morris et al., 2008; Ruff et al., 2011]. The rocks reflect extensive acid-sulfate weathering of basalt early in Mars' history. At Gusev crater, a range of chemically weathered basalts, Mg-sulfate soils, and high-silica soils have been noted, and they likely formed in volcanic hydrothermal environments [Squyres et al., ; Yen et al., 2007; Ruff et al., 2011]. Schmidt et al. [2009] hypothesized that a localized, high-temperature event that affected the east side of the Home Plate deposits and a lower-temperature alteration occurred on the west side that formed nanophase iron oxides. Ruff et al. [2011] argued that high silica deposits found near Home Plate were inconsistent with the originally proposed Hawaiian analog for fumarolic acid-sulfate leaching and instead suggested that the deposits were more consistent with sinter produced by silica precipitation from hot springs. Indeed, questions still exist as to the role and magnitude of water at Home Plate. At Meridiani Planum, layered bedrock also contains signatures of interactions of basalt with acidic water or vapor, which the MER Science Team has interpreted to be a result of acid-sulfate weathering via predominately groundwater processes [e.g., Squyres et al., 2004]. Alternatively, McCollom and Hynek [2005] proposed that the acid-sulfate alteration occurred in a high-temperature, acidic volcanic environment.

[4] More recently, the Observatoire pour la Minéralogie, l'Eau, les Glaces, et l'Activité (OMEGA) investigation, on board the Mars Express mission and Mars Reconnaissance Orbiter's Compact Reconnaissance Imaging Spectrometer for Mars (CRISM), have identified discrete bedrock outcrops around the planet bearing Fe-, Mg-, and Ca-sulfates [e.g., Gendrin et al., 2005; Bibring et al., 2006, 2007; Murchie et al., 2009; Wray et al., 2009a; Ehlmann et al., 2009, 2011a, 2011b]. A variety of sulfates and iron oxides occurs within the great rift valley Valles Marineris that likely hosted high-temperature alteration of basaltic crust on Mars [e.g., Chojnacki and Hynek, 2008]. Indeed, the OMEGA Team has suggested that “sulfur-rich fluid circulation and the alteration of volcanic ashes into sulfates” may be a major contribution to the observance of sulfates globally on Mars [Gendrin et al., 2005]. In addition to sulfates, numerous clay minerals (phyllosilicates) are now recognized throughout the ancient highlands of Mars, including representatives from the smectite, kaolin, illite, and chlorite groups [e.g., Bibring et al., 2006; Murchie et al., 2009; Wray et al., 2009a; Ehlmann et al., 2011a, 2011b]. The largest of these sites is Mawrth Vallis, which was a finalist for the 2011 Mars Science Laboratory mission. Here Mg/Fe clay layers comprise the lower strata and are overlaid by Al-rich clay layers, and jarosite has been detected as well, indicating changing paleoenvironmental conditions or sediment supply [Wray et al., 2008; Farrand et al., 2009]. Beyond Mawrth Vallis, clay minerals have been detected throughout the southern hemisphere of Mars, in strata cropping out from the walls of craters and cliffs as well as in ejecta deposits. A variety of processes was likely responsible for the production of clays from the original basaltic crust and include hydrothermal alteration, low-grade metamorphism, and diagenesis [e.g., Ehlmann et al., 2011a, 2011b].

[5] Carbonate minerals have been also detected in a few small locales on Mars [Michalski and Niles, 2010; Ehlmann and Mustard, 2012]. At Nili Fossae, a Mg-carbonate is noted in association with olivine basalt and phyllosilicates [Ehlmann and Mustard, 2012]. The workers argued that the carbonate formed as a result of hydrothermal alteration of olivine with neutral to alkaline waters. Finally, putative chloride salt deposits have been identified in over 600 locales around Mars [Osterloo et al., 2008, 2010]. These deposits are mostly in topographic basins, and a number of them are associated with fluvial channels, suggesting that they are generally formed by the evaporation of briny fluids.

[6] In sum, the recent mineralogical results from Mars have revealed abundant and diverse interactions between the volcanic crust and water. The diversity of alteration minerals implies that a wide variety of paleoenvironments was present on ancient Mars. Sulfates such as jarosite imply acidic conditions in places, while at Nili Fossae, the assemblage of carbonates and phyllosilicates suggests neutral to alkaline fluids. In places, alteration is thought to have occurred at low temperatures [e.g., McLennan et al., 2005] or even within ice deposits [e.g., Niles and Michalski, 2009]. Yet at other places, like in the walls of the Valles Marineris rift system, specific minerals such as prehnite have been detected and imply formation at T > 200°C [Ehlmann et al., 2011a]. The diverse sulfate and chloride deposits seen around Mars also imply that fluids in those paleoenvironments were significantly saline. Understanding the history of these materials, including the geochemical pathways involved in their alteration from fresh to heavily weathered materials and the paleoconditions present (e.g., fluid temperature, pH, and salinity), would provide significant constraints on the geological evolution of these widespread materials and their astrobiological potential.

[7] Mars has experienced volcanism from the earliest discernible times until the past few million years, with some individual volcanoes having a string of large eruptions occurring over a time span of 3.5 billion years [Robbins et al., 2011]. In particular, early Mars had rampant magmatism and volcanism, owing to the high heat flow on the young planet [e.g., Hauck and Phillips, 2002], and the largest volcanic provinces were mostly constructed by 3.5 Ga [Phillips et al., 2001]. These times of enhanced volcanism were coeval with abundant surface water on a global scale [e.g., Hynek et al., 2010], providing the ingredients for hydrothermal circulation. Mars undoubtedly had long-lived hydrothermal systems in basalt-hosted volcanic environments, and there is evidence for such settings in the mineralogy identified by orbiters and landers, including within Gusev crater [e.g., Murchie et al., 2009; Squyres et al., ; Yen et al., 2007; Ruff et al., 2011]. There is also compelling evidence from the geological settings that include recent lavas and water erupting from fractures on the sides of volcanoes [Burr et al., 2002], valleys on volcanoes consistent with a hydrothermal origin [e.g. Dohm and Tanaka, 1999; Gulick, 2001], and volcanic rift systems and chaotic terrains that show interactions of magma and water [Montgomery and Gillespie, 2005; Chojnacki and Hynek, 2008]. Hydrothermal environments such as fumaroles and hot springs provide sources of heat, energy, and water for life [Walter and Des Marais, 1993; Schulze-Makuch et al., 2007], even when it is too dry or cold for life to persist nearby. Hydrothermal environments have been suggested as potential sites for the origin of life on Earth [e.g., Wächtershäuser, 2006], and Mars may have followed a similar pattern [Walter and Des Marais, 1993]. In modern hydrothermal environments on Earth, the extent and diversity of the microbial ecosystems are constrained by the geochemical and mineralogic characteristics of the system, and similar parameters are likely to constrain habitability in analog Martian settings.

[8] Given the strong evidence for prolonged magmatism, crustal water, and hydrothermalism on early Mars, we characterized four field sites that have chemistries, mineralogies, and processes analogous to relic hydrothermal systems identified on Mars. These areas were targeted because the lithology and bulk chemistry of the fresh basalts are similar to that of Mars' crust and SNC meteorites (Table 1) [Hynek et al., 2011]. To understand past aqueous alteration processes on Mars, it is critical to begin with rock compositions similar to Martian samples, since host-rock lithology largely controls the geochemical pathways [e.g., Tosca et al., 2004]. These volcanic field sites have fluids and gas compositions characterized by high temperature, low pH, and high salinity, which are all inferred for a number of settings on Mars [e.g., Gusev crater: Squyres et al., ; Meridiani Planum: McCollom and Hynek, 2005; Valles Marineris: Chojnacki and Hynek, 2008; Athabasca/Cerberus: Burr et al., 2002; Thaumasia Planitia: Dohm and Tanaka, 1999]. Yet while these terrestrial field sites have a similar basaltic parent rock lithology, they differ from each other in gas chemistry and have a range of environmental parameters (e.g., temperature and pH). As noted above, different suites of alteration minerals occur across Mars and may be a result of environmental variables, including fluid and parent rock composition, temperature, pH, rock and fluid chemistries, and oxidation state. Thus, the alteration mineralogy holds clues to the ancient environments and overall habitability of Mars. Obviously, the gas composition of past volcanic vents on Mars is wholly unconstrained. Here we assess acid-sulfate alteration resulting from “typical” gases associated with terrestrial basaltic systems. We studied three active acidic volcanoes in Nicaragua and an associated hydrothermal field to assess the controls on secondary mineralogy to elucidate the likely paleoenvironments of early Martian hydrothermal systems. We find that environmental conditions can play a major role in resultant alteration mineralogies. Our assessment of basalt-hosted hydrothermal environments on Earth provides a framework for interpretation of similar systems on ancient Mars.

Table 1. Average Rock Composition by Oxide Weight Percent for the Nicaraguan Volcanic Systems Compared to Martian Crustal Basalts and SNC Meteorites (See Also Figure 3)
SiO2 Al2O3 FeOa MgO CaO Na2O K2O P2O5
Cerro Negrob 49.73 19.50 9.70 4.73 11.50 2.18 0.43 0.12
Momotombob 54.40 16.90 9.10 4.47 9.20 2.89 0.91 0.15
Telicab 51.72 19.10 9.60 4.13 10.10 2.84 0.96 0.17
Adirondackc 45.90 10.60 18.70 9.90 7.90 2.60 0.15 0.63
Backstayc 49.40 13.10 13.20 8.30 6.00 4.00 1.02 1.34
Irvinec 47.50 8.30 19.70 9.50 5.80 3.00 0.60 0.94
Bounce Rockd 50.80 10.10 15.60 6.40 12.50 1.30 0.10 0.95
Shergottye 51.30 6.88 19.40 9.30 9.60 1.39 0.17 0.67
QUE 94201e 47.90 11.00 18.50 6.25 11.40 1.58 0.045 -
EET 79001e 49.40 11.20 17.40 6.57 10.80 1.74 0.075 0.193
Zagamie 50.50 6.05 18.10 11.30 10.50 1.23 0.14 0.50
  • a Total Fe given as FeO.
  • b Average composition from Carr and Rose [1987] and this study.
  • c Average composition from Gusev basalt classes of rocks from Ming et al. [2008].
  • d Composition of Bounce Rock at Meridiani from Rieder et al. [2004].
  • e Average composition from selected SNC meteorites from Lodders [1998].

[9] This study builds on our previous work [e.g., Hynek et al., 2011] and is part of a broader effort to understand the mineralogical and chemical alteration of basalt at these sites, as well as associated microbiological activity. Other reports describe mineralogical and geochemical trends during acid-sulfate alteration in areas of diffuse vapor discharge [McCollom et al., 2013a], spectroscopic study of exposed alteration mineralogy [Marcucci et al., this issue], microbial diversity of a hydrothermal site within Cerro Negro [Rogers et al., 2011], and laboratory simulations of acid-sulfate alteration of Cerro Negro basalt [McCollom et al., 2013b; Marcucci and Hynek, this issue]. Collectively, these results provide assessment of the controls on resultant alteration mineralogy and inferences into putative Martian hydrothermal systems and their astrobiological potential.

2 Description of Field Sites

2.1 The Central American Volcanic Chain

[10] This study focuses on three currently active volcanic areas—Cerro Negro, Momotombo, Telica—and one associated hydrothermal area (Telica's Hervidores de San Jacinto) of western Nicaragua (Figure 1). These areas, and most of western Central America, are volcanically and seismically active due to the subduction of the Cocos plate under the Caribbean plate, resulting in a line of 39 distinct volcanic centers that stretch throughout Central America (Figure 1) [Carr et al., 2007]. This activity has produced a volcanic front ranging from 165 to 190 km from the subduction trench and additional back arc systems farther inland [Carr et al., 2007]. Across Central America, the lavas range from basalt to rhyolite. The lithosphere in Nicaragua is thinner than elsewhere in Central America, and the subducting slab is dipping steeper [e.g., Patino et al., 2000], leading to a dominance of basaltic and basaltic-andesitic volcanism with little crustal contribution (Figure 2) [Carr et al., 2003]. All of the unaltered lavas studied herein are calc-alkaline volcanic arc products with high Al and sometimes exhibiting tholeiitic affinities [Nystrom et al., 1988; Carr et al., 2003, 2007]. The lavas are characterized by phenocrysts of plagioclase, augite, olivine, and minor iron-titanium oxides. Geochemical analyses indicate that the mantle source of basalts under Nicaragua, El Salvador, and Guatemala are congruent with a middle ocean ridge basalt source. Contrastingly, the Costa Rican and Panamanian volcanoes show a mantle source consistent with an ocean island basalt. This simplistic picture is confounded by the complex tectonics, poorly understood paleogeography, and potential past subduction of mantle plumes or “slab windows” [e.g., Vogel et al., 2006; Carr et al., 2007; Gazel et al., 2011].

Details are in the caption following the image
Digital elevation model for northwestern Nicaragua highlighting the volcanic chain. Visible images of the volcanoes discussed herein are inset.
Details are in the caption following the image
Total alkalis versus silica and the International Union of Geological Sciences classification. CN = Cerro Negro, T = Telica, and MO = Momotombo. For comparison, we have plotted selected basaltic SNCs that are mostly Shergottites [Lodders, 1998], major classes of Gusev basalts [e.g., Ming et al., 2008], and Bounce rock (a basaltic float sample) from the Meridiani landing site [Rieder et al., 2004].

2.2 Cerro Negro Volcano

[11] Cerro Negro (12.507405°N, 86.702403°W) is perhaps the world's youngest volcano (~250 m high and 1 km wide) that initiated activity in the year 1850 [Walker and Carr, 1986]. It has been built up over 24 strombolian to subplinean eruptions since that time, ranging from explosive to effusive activity [e.g., Roggensack et al., 1997] that led to basaltic lavas, cinder, and ash deposits. This study focuses on the products from the 1992 and 1995 eruptions. In 1992, the volcano came to life after 23 years of quiescence with explosive subplinean eruptions that resulted in ash plumes up to 7 km high, forming a large crater. In 1995, the next major eruptive phase occurred and was strombolian in nature. It included central vent explosions, pyroclastic flows, phreatic explosions, and lava dome extrusion. A large tephra cone resulted from this eruptive phase that lasted most of the year. The most recent eruptions occurred in 1999–2000 and were characterized by a radial fissure eruption that lasted only 2 days. Over its lifetime, the volcano produced a comparable amount of tephra (0.0574 km3, 59%) and lavas (0.0397 km3, 41%) [Hill et al., 1998] and likely represents a new composite cone versus a parasitic cinder cone [Walker and Carr, 1986; McKnight and Williams, 1997]. The recently erupted basalts have unusually large phenocrysts of plagioclase (An96–85), olivine (Fo81–72), clinopyroxene (En45 Wo38 Fs16–En45 Wo45 Fs14), and subordinate titanian magnetites (Usp16–20) [Walker and Carr, 1986]. We identified plagioclase (anorthite and bytownite), augite, and minor amounts of forsterite in fresh basalts from the 1992 eruption (methods in section 3 below). From bulk chemistry, the pristine basalts from recent eruptions are widely similar and fall in the basalt category (Figure 2), with major element abundance similar to those of the Martian crust and SNC meteorites and crustal samples examined by rovers, albeit with higher Al and lower Mg and Fe than average Martian products measured to date (Table 1 and Figure 3) [Hynek et al., 2011].

Details are in the caption following the image
Average major oxide abundance of field samples from Cerro Negro, Momotombo, and Telica compared to selected basaltic SNCs (blue) [Lodders, 1998], Gusev basalts (red) [e.g. Ming et al., 2008], and Bounce rock (orange) (a basaltic float sample) from the Meridiani landing site [Rieder et al., 2004]. Data are in Table 1.
[12] The erupted materials have undergone extensive acid-sulfate alteration within hundreds of meters of the central vent via active fumarolic processes, producing an abundance of alteration products. The chemical weathering at all field sites herein is predominately controlled by sulfuric acid that is produced during cooling of volcanic gases by the following reactions:

[13] Alteration has occurred along much of the northwest facing inner crater wall formed in the 1992 eruption of Cerro Negro, and this was a primary field site. At least a dozen focused and diffuse fumaroles exist along a 170 m portion of the wall and have altered much of its ~50 m high wall, which is composed of basaltic tephra (Figure 4). Near the center of this area, temperatures are highest (~120°C at the surface and up to 400°C a few centimeters down) with pH of condensed gas that is 0 to −1. Away from this central area, temperature decreases and gas vents are of higher pH, likely due to increased contributions from meteoric groundwater and reaction of vapors with basalt in the subsurface as they diffusely permeate through the cinder deposits. Toward the edges of the crater's interior, the temperature is ~45°C and pH increases to ~6. At least three fumarolic areas had green layers within the mineral crusts, indicating photosynthetic microbiological communities [see Rogers et al., 2011]. Several closed topographic basins were found downslope from the active vents. Rainwater has led to the erosion of fumarolic material into these local lows, with evidence of ponding and subsequent evaporation.

Details are in the caption following the image
Inner crater rim formed during the 1992 eruption of Cerro Negro. Measured environmental variables are included along the 170 m altered wall exhibiting active fumaroles.

[14] The other primary field site at Cerro Negro is on the rim and the interior of the 1995 tephra cone, which buried most of the remaining 1992 crater walls during its construction (Figure 5). This region also has diffused and localized gas emissions with pH ranging from 1 to 4 and surface temperatures from 45°C to 110°C. We visited both of these sites in 2006, 2008 (twice), and 2012 and noted little change in the distribution and vigor of fumarolic activity. Gases at all Nicaraguan field sites studied herein are dominated by water, CO2, SO2, and HCl, as is typical of high-temperature volcanic emissions. However, Cerro Negro had over an order of magnitude less SO2 than Momotombo, several orders of magnitude less HCl, but is much higher in CO2 emissions reported in recent measurements (Table 2) [Elkins et al., 2006; Fischer et al., 2006].

Details are in the caption following the image
The 100 m diamter summit crater created during the 1999 eruption of Cerro Negro. Annotated are measured fumarole parameters and alteration mineralogy, hem = hematite, mon = montmorillonite, aln = natroalunite, jar = jarosite, gyp = gypsum, s = sulfur, si = amorphous silica.
Table 2. Gas Chemistry in Weight Percenta
Cerro Negro Momotombo Telica
STotal 4.35 6.18 0.30
HCl 0.41 13.85 0.21
HF 0.073 1.16 0.0050
CO2 74.19 67.60 83.93
NH3 0.0020 0.0018 0.0078
He 0.0026 0.0009 0.0016
H2 0.010 9.59 0.0040
Ar 0.063 0.0075 0.018
O2 0.80 0.00080 0.0040
N2 20.10 1.37 15.50
CH4 0.0010 0.00065 0.013
CO 0.0064 0.22 0.00010
  • a H2O is the largest component, but compositions here are normalized without it [Elkins et al., 2006].

2.3 Momotombo Volcano

[15] The Momotombo volcano (12.424223°N, 86.531783°W) was also visited in the 2012 field campaign (Figure 1). This cone-shaped calc-alkaline construct rises 1297 m high from the north shore of Lake Managua and is the southernmost volcano of the Marabios Range of northern Nicaragua. Momotombo began growing about 4500 years ago and consists of a somma from an older edifice that is surmounted by a symmetrical younger cone with a 150 × 250 m wide summit crater. Momotombo has had over 15 eruptions since 1600 A.D., with most being explosive in nature, but also with production of some lava flows. The last eruption of Momotombo occurred in 1905, with a central vent explosion and a lava flow that stretched 5 km to the north.

[16] Fresh Momotombo eruption products have a higher silica content than Cerro Negro or Telica, and the average composition plots as basaltic andesite (Figure 2). The bulk chemistry of the lavas and tephra is quite similar to the basalts of Cerro Negro, but with slightly lower calcium and aluminum (Table 1 and Figure 3) [Carr and Rose, 1987]. The 1905 eruption produced black basaltic to andesitic-basalt products that had similar abundance of plagioclase and pyroxene [Novák, 2006]. We measured the mineralogy of fresh erupted materials near the summit crater by X-ray diffraction and found major components to be mostly plagioclase (albite) and various pyroxenes (hedenbergite, augite, and Na-rich Fe3+ silicate aegirine).

[17] Momotombo currently has strong and persistent fumarolic activity in the summit crater [e.g., Stoiber et al., 1980] (Figure 6). The temperature of these emissions was stable at about 230°C until 1973 but increased up to 800°C in 1980, and the rapid change suggested to Menyailov et al. [1986] that an eruption was possibly imminent (though none have yet occurred). We measured steam emissions ranging from 100°C to 605°C in August 2012, although we were unable to collect samples at the fumaroles above 200°C due to hazardous access. The majority of erupted products within the summit crater have been extensively altered by the dozens of fumaroles present and diffuse outgassing (Figure 6). The pH of the steam emissions was from 0 to < −1, the lower end being the limit of our measurement methods. The gas chemistry of Momotombo contains the highest values of overall S, HCl, and HF of the field sites (Table 2). Spheroidal weathering of large basalt boulders was widespread, and we collected samples of the various stages of alteration. A breach in the crater rim extends to the northeast from a late 1800s eruption, and fluvial processes have transported summit crater material to local catchments in a leveed lava channel that reaches from the summit to the surrounding jungle. The outside of the summit crater contained few signs of gas emissions or acidic alteration of the fresh basalt cinders that comprise the slopes of Momotombo. At the base of the southern side of the volcano exists a large geothermal field that was developed in 1973, with 33 drilled wells that have an energy capacity of 100 MW [Moore et al., 1981].

Details are in the caption following the image
Context image of the Momotombo summit crater and environmental data. The pH of fumaroles was consistently 0 to <−1.

2.4 Telica Volcano

[18] The Telica volcano (12.605010°N, 86.841645°W), north of Cerro Negro, was the third field area studied (Figure 1). We visited the summit crater of Telica in August 2012 and the associated Hervidores de San Jacinto mudpots area on the flank in 2006, 2008, and 2012. Telica is a stratovolcano rising ~1000 m above the surroundings. It has had persistent eruptions since Spanish settlement, with the most recent eruption in 2011. The summit is composed of two main craters ~600 m across, with most recent eruptions coming from the southern one, which was our field site (Figure 7). The 1999–2000 eruption hosted a short-lived lava lake. When we visited in August 2012, loud gas jets were forcefully shooting out of the sides of the deep crater, which extended 250 m vertically below the rim. The summit materials consisted of basaltic ash, pyroclastics, and ejected blocks. We were only able to access rocks on the rim of the active crater and not the interior of the crater. Thus, altered blocks and boulders ejected during recent eruption activity were the basis for our study at this locale. The bulk chemistry of the pristine materials is basaltic to basaltic andesite (Figure 2), with phenocrysts of plagioclase (anorthite) and olivine [Novák and Přichystal, 2006]. We measured albite, quartz, augite, diopside, petedunnite, and amorphous/glass (methods in section 3 below). The basalt at this volcano has a similar bulk chemical composition to Cerro Negro (Table 1 and Figure 2). The gas composition at Telica is similar to that of Cerro Negro (Table 2).

Details are in the caption following the image
Telica summit crater. Foreground alteration products include natroalunite, amorphous silica, montmorillonite, and hematite.

2.5 Hervidores de San Jacinto Mudpots

[19] The Hervidores de San Jacinto geothermal field lies to the south-southeast edge of the Telica volcano. Here the fluid-rock ratio was much higher than in the other gas-dominated field sites, resulting in roughly a dozen hot springs and a dozen mudpots (Figure 8). We sampled both geothermal pools and fumaroles in an area covering ~800 m2. No pristine rocks existed in this area, but since the region is on the flank of Telica, we assume that the parent rock lithology was similar to fresh Telica volcanics. The fumaroles were peppered throughout the geothermal field and had a fairly consistent temperature at the vents of 100°C, and pH ranged from 3.5 to 4.6. Green photosynthetic endolithic communities were found in gypsum crusts at fumaroles at this site and are the subject of future studies. Geothermal pools ranged from clear to turbid (mudpots) and had minor to vigorous bubbling. Temperatures of the clear hydrothermal springs and mudpots ranged from 75°C to 100°C, and pH encompassed values from 1.5 to 4.0.

Details are in the caption following the image
Hervidores de San Jacinto geothermal field. This mudpot had a pH of 4.0 and a temperature of 96°C. The muds are composed of mostly kaolinite and montmorillonite.

3 Methods

3.1 Analysis of Solids

[20] At each field site, we collected a representative set of fresh and altered rocks and sediments (dozens to hundreds of hand samples) to characterize the alteration processes and pathways of the fresh to altered rock samples, following our existing protocols for work on similar systems [e.g., Hynek et al., 2011]. Briefly, this was done in situ by characterizing the mineralogy of pristine and a range of altered deposits by traditional field methods and also with the portable Terra X-ray diffraction/X-ray fluorescence (XRD/XRF) instrument (inXitu Incorporated, Campbell, CA) that is functionally equivalent to the CheMin instrument on the Mars Science Laboratory (MSL) rover named Curiosity [Blake et al., 2009]. The Terra instrument is briefcase-sized and weighs just over 20 kg, making it field portable. A CoKα radiation source is used, and a sample run requires a few minutes for a monomineralic sample or up to an hour for a complex or poorly crystalline one. This instrument also has the advantage of requiring only a minor amount of sample for analysis (≪ 1g), so we were able to measure thin coatings and, in some cases, individual crystals. We have found that in situ mineralogy measurements are crucial in active volcanic systems—when the same samples are measured on site and in the lab, they often have different mineralogies, the latter measurements being attributed to dehydration, oxidation, and other effects associated with removal from the natural settings. Further, making measurements with an analogous instrument to one on MSL provides a basis for direct comparisons to those incoming data. We consider a mineral positively identified when a pattern matched all strong and moderate diffraction lines on the appropriate American Society for Testing and Materials cards. We used XPowder software for pattern matching and quantitative analysis along with the “difdata” library from the American Mineralogist Crystal Structure Database for phase identification, which we customized to include additional sulfate phases. We analyzed diffractograms from 5° to 55° in 2-theta space.

[21] Figure 9 shows representative XRD diffractograms from the various field locales along with library pattern peak matches. Nearly 200 monomineralic to polymineralic samples from the 2012 field excursion were analyzed and included fresh basalts and secondary products, including sediments from fumarole aprons and outwash basins, crusts, rock coatings, and muds. Roughly one third of our samples had a strong amorphous silica hump from ~18° to 27° in 2-theta space. In some cases, additional minerals could be identified in these samples, and other times, there were no clear diffractogram peaks, and we assumed that the sample was composed almost entirely of amorphous silica (Figure 9).

Details are in the caption following the image
Representative diffractograms for the various field sites analyzed by XRD as described in the text. Library mineral identifications are labeled over the main peaks. a = augite, al = alunogen, am = amorphous silica hump, ap = apjohnite, c = cristobalite, ca = calcite, ch = chlorite, d = despujolsite, gy = gypsum, h = hematite, il = illite, j = jarosite, k = kieserite, ka = kaolinite, K-al = potassium alum, p = plagioclase, pk = pickeringite, m = melanterite, md = mendozite, mo = montmorillonite, mu = muscovite, na = natroalunite, s = sulfur, z = zeolites.
Details are in the caption following the image

[22] We also characterized the field samples with thin-section petrography and XRD on a traditional instrument to fully detail the bulk and minor mineralogy. Fresh and altered samples were sent to ActLabs (Ontario, Canada) for XRF analysis that provided major oxide compositions and sulfur and chlorine contents. Moreover, we used petrographic microscopy, scanning electron microscopy (SEM), and an electron microprobe to characterize the micron-scale chemistries and mineralogies on a subset of the samples [see, e.g., McCollom et al., 2013a]. Finally, we characterized the pristine and altered samples in the field and upon returning to our laboratory with visible and near-infrared (VNIR) spectroscopy on a TerraSpec 4 High-Resolution Mineral Spectrometer from Analytical Spectral Devices, Inc. (Boulder, CO) [see Marcucci et al., this issue]. This is a nondestructive technique that provides a means for investigating geologic materials both in the field and in remote sensing applications. The instrument measures broadband reflected light in the 0.35–2.50 µm range. The hand probe was put in direct contact with the sample to be measured, thus eliminating atmospheric contributions. Spectra were imported into the ENVI software package for analysis. Mineral identification was done based on band matching with the U.S. Geological Survey mineral library [Clark et al., 2007] and the sulfate, rock, and oxide CRISM libraries [Planetary Data System (PDS), 2013]. See Marcucci et al. [this issue] for additional details on fitting spectra.

3.2 Measurements and Analysis of Fluids and Gas

[23] High-fidelity pH paper (Micro Essential Laboratories, Brooklyn, NY) and a micro-pH meter were used to determine the pH at the field sites. Measuring pH is challenging in gas-dominated environments. The vigorous fumaroles were measured by placing the instrument or the pH paper directly in the gas emissions. The acidic steam quickly condensed on the pH paper, allowing accurate measurements. For diffuse gas emissions, we used the pH paper to swab the wet rocks and sediments within the vent areas. While this technique does not measure the acidity of the gases accurately, it does provide measurement of the pH at the surfaces of the wet materials, where the alteration is taking place. Temperature was measured with several commercial temperature probes and thermocouples. In some areas, temperatures increased by hundreds of degrees Celsius within a few centimeters of depth. Temperature and pH were coregistered with field samples to better assess their impact on resultant mineralogy.

[24] At roughly a dozen Cerro Negro fumaroles, we condensed gas emissions at active fumaroles that had high discharge, and we also directly sampled fluids at the San Jacinto geothermal field from mudpots and thermal pools. We used a steam condenser to collect ~50 mL of fluid within about 20 min (depending on vigor). This apparatus was first employed by Ellis et al. [2008] to study microorganisms transported in the gas emissions. It consists of an inverted corked funnel that is filled with ice; the temperature contrast acts as a cold trap for the vapors and steam condensation on the outside of the funnel drips down into a collection vessel (Figure 10). We used sterile 50 mL Falcon tubes for fluid collection. pH was measured of the collected fluids, providing another gauge of pH for gas emissions. Fluids were analyzed for major, minor, and trace components using inductively coupled plasma–atomic emission spectroscopy (ICP-AES) at the Laboratory for Environmental and Geological Studies, University of Colorado.

Details are in the caption following the image
Steam condensing apparatus. An inverted funnel inside the main housing leads to condensation on its outside that drips down into the underlying tube.

[25] We attempted to directly measure gas chemistry following the methods of Arnórsson et al. [2006] at Cerro Negro, but the results contained significant atmospheric contributions. Thus, the gas chemistry for this work was compiled from the literature sources of Elkins et al. [2006] and Fischer et al. [2006]. Both direct (gas sampling) and indirect (remote sensing) measurements are reported in these studies. Water is the major component of volcanic gases, and Table 2 shows residual species at Cerro Negro, Momotombo, and Telica normalized with water removed [Elkins et al., 2006]. In terms of total sulfur measured, Momotombo is highest while Telica was lowest. Our field observations suggested that SO2 was the dominant species at all volcanoes, given the blue gas plumes and lack of H2S rotten egg smell. Momotombo also has much higher HCl and HF than the other two volcanoes, while Cerro Negro had much more oxygen and carbon dioxide.

4 Environmental Controls on Geochemical Pathways and Acid-Sulfate Alteration

[26] Stoiber and Rose [1974] presented one of the first and only studies that focused on fumarolic alteration of Central American fumaroles. The workers sampled 14 volcanoes in Guatemala, El Salvador, Nicaragua, and Costa Rica over a period of up to 11 years and identified 47 minerals in incrustations being deposited at approximately 100 different high-temperature fumaroles. This study focused on incrustations right at and within the mouths of hot (>200°C) active and focused vents, while our study looked over a broader area. The most abundant and most frequently found minerals were sulfur, hematite, halite, sylvite, gypsum, ralstonite, anhydrite, thenardite, and langbeinite. They also noted that incrustation suites deposited around fumaroles produced a zonal pattern that can be explained by the reaction of volcanic gases composed of H2O, SO2, CO2, HC1, and HF, which interacts with the atmosphere and the fumarole wall rock over a range of temperatures.

[27] Our results build on this initial study by focusing on the more mafic volcanoes and examining a much broader range of local depositional settings around active areas. Of the most common minerals found by Stoiber and Rose [1974], only two were common products at the fumaroles we sampled, sulfur and gypsum, but we also noted minor occurrences of hematite and anhydrite. The differences can be likely attributed to several factors. First, our focus was on volcanoes with basaltic to basaltic andesite composition, and as noted, parent rock lithology exhibits strong controls on alteration mineralogy. Second, gas composition varies widely across the Central American volcanic chain, but our sampled regions have fairly consistent gas chemistry (Table 2). Third, we did not sample many regions with temperature as high as that of Stoiber and Rose [1974], who sampled areas well over 600 K within vents. Our focus was on surface materials at the edges of the vents and in surrounding aprons. Finally, two of our three field sites had major eruptions since the earlier study, and we focused on the ongoing alteration from these nascent events.

[28] In total, the fumarole vents that we sampled ranged in temperature from 45°C to 200°C and pH from −1 to 6. In each case, we measured these local environmental variables to assess their role in alteration of mafic materials with similar compositions to those on the surface of Mars. As shown in Figures 2 and 3 and Table 1, the pristine chemistry and mineralogy of Cerro Negro, Momotombo, and Telica are quite similar. This allowed us to constrain the major variables of aqueous alteration to differences in temperature, pH, and gas chemistry. Below, we divide our mineralogical results into five main classes of environments: (1) extremely low pH (−1 to 1) and high-temperature (>100°C) fumaroles, (2) moderately acidic (pH from 4 to 5.5) and moderate-temperature (50°C–100°C) fumaroles, (3) high pH (5.5–6.5) and low-temperature (40°C–65°C) fumaroles, (4) fluvial outwash sediments at downslope topographic catchments, and (5) the high fluid-rock ratio geothermal pools and mudpots of Hervidores de San Jacinto (pH ~ 1.5–4.6; temperatures of 55°C–100°C). We also note differences between the volcanoes and our inferences as to why.

4.1 Fumaroles With Extremely Low pH (−1 to 1) and High Temperature (>100°C)

[29] We sampled roughly 12 fumaroles in the craters of Cerro Negro and Momotombo that fell into this category. At Cerro Negro, these locations were near the middle of the volcanic crater, both at the base of the 1992 wall and in locations on the inner walls of the 1995 cone (Figures 4 and 5), while at Momotombo, nearly all of the samples from the summit crater fell in this category. Table 3 shows the common major, minor, and trace mineralogies. These highest temperature and lowest pH fumaroles consisted of relatively simple secondary mineral assemblages dominated by elemental sulfur surrounding the active vents (e.g., MO28 in Figure 9) and an apron consisting of silica (Figure 11) overlaid by occasional gypsum (CaSO4⋅2H2O) encrustations (e.g., CN4 in Figure 9). Within a few years, the primary basalts are entirely reduced to these products, with the S0 precipitating from the gases, forming large crystals most proximal to the vents. Elemental sulfur was also widespread in the surrounding amorphous silica aprons. The highest temperature sites also exhibited abundant cristobalite and tridymite, high-temperature SiO2 phases (e.g., Mo16 in Figure 9). We only found these products around the active vents, suggesting that the hot vapor phase was responsible for their formation and that they did not exist a priori in the original wall rock. The amorphous silica that was ubiquitous in meter-scale aprons spanning several hundreds of square meters around the focused, high-temperature vents is likely a result of the wholesale leaching of the original volcanics [see also McCollom et al., 2013a]. The mobilized Ca2+ combined with the condensed H2SO4 on the surface of aprons, leading to the gypsum crusts up to a few centimeters thick. Interestingly, thermodynamic models predict that anhydrite should be the stable Ca-sulfate at the temperature of these sites [e.g., Catling et al., 2006] instead of gypsum, but this phase was seldom present while gypsum was widespread. Minor amounts of hematite represented the only iron oxide, and this phase was generally present in red or orange surface deposits on volcanic rubble around and in the vents. The phyllosilicates montmorillonite (a calcium-aluminum smectite) and nontronite, the iron(III)-rich member of the smectite group of clay minerals, were found in minor abundance in the aprons and formed from the weathering of basalts or the precipitation of hydrothermal fluids. Finally, the phyllosilicate muscovite was occasionally present in minor/trace abundance and is a product of the weathering of feldspars.

Table 3. Typical Major, Minor, and Trace Minerals Identified in Five Types of Environmental Settings in Nicaraguan Volcanoes
Major Phases Minor Phases Trace Phases
Fumaroles: pH = −1 to 1; T > 100°C sulfur; gypsum (CaSO4⋅2H2O); Si (amorphous); SiO2 (cristobalite and tridymite) hematite (Fe2O3) muscovite; nontronite; montmorillonite
Fumaroles: pH = 4.0 to 5.5; T = 50°C-100°C gypsum; Si (amorphous) natroalunite; jarosite (KFe3+3 (SO4)2(OH)6); alunite (KAl3 (SO4)2(OH)6); Fe2(SO4)3; sodium alum (NaAl(SO4)2·12H2O); alunogen Al2(SO4)3⋅17H2O); goethite (FeOOH); zeolites; montmorillonite szomolnokite (Fe2+SO4⋅H2O); mereiterite (K2Fe2+(SO4)2⋅4H2O); kieserite (MgSO4⋅H2O); hexahydrite (MgSO4⋅5H2O); smectite; kaolinite
Fumaroles: pH = 6; T = 60°C calcite (Ca(CO)3) gypsum brushite (CaHPO4⋅2H2O)
Outwash Basins: T = ambient gypsum; alunite; goethite pickeringite (MgAl2(SO4)4⋅22H2O); zeolites alpersite (MgSO4⋅7H2O); mendozite (NaAl(SO4)2⋅11H2O); melanterite (FeSO4⋅7H2O); epidote
Mudpots: pH = 1.5 to 4.0; T = 75°C–100°C kaolinite; montmorillonite despujolsite (Ca3Mn4+(SO4)2(OH)6⋅3(H2O)); khademite (Al(SO4)F⋅5(H2O)); fibroferrite (Fe3+(SO4)(OH) 5(H2O)); pickeringite; alunogen (Al2(SO4)3⋅17H2O) stilbite; aerinite; gobbinsite; FeOCl; potassium alum; illite; chlorite; vermiculite
Details are in the caption following the image
Active fumaroles within the Momotombo summit crater. The pH of the steam was 0, and the temperature was 120°C. s = elemental sulfur, si = amorphous silica, gyp = gypsum.

4.2 Fumaroles With Moderately Acidic pH (4.0–5.5) and Moderate Temperature (50°C–100°C)

[30] The alteration was slower at the moderate pH (4–5.5) fumaroles, occurring on decadal scales (reconstructed from eruption histories). Amorphous silica was similarly abundant in this setting (e.g., CN19 in Figure 9), composing much of the apron, while gypsum was also common and generally occurred as surface encrustations and within apron sediments (Figure 12). Little elemental S was detected; instead, this element was sequestered into a number of Al-, Fe-, Mg-, and Ca-sulfates, including jarosite (KFe3+3 (SO4)2(OH)6), natroalunite ((Na,K)Al3(SO4)2(OH)6), Fe2(SO4)3, szomolnokite (Fe2+SO4⋅H2O), mereiterite (K2Fe2+(SO4)2⋅4H2O), kieserite (MgSO4⋅H2O), and hexahydrite (MgSO4⋅5H2O) (Table 3). Alunite/jarosite existed as centimeter-thick encrustations but only at certain fumaroles. Note that a range of natroalunite compositions was observed in all our samples, with some having significant Fe3+ substitution for Al3+, and these ramifications are discussed in [McCollom et al., 2013a, 2013b]. Sodium alum (NaAl(SO4)2⋅12H2O) and alunogen Al2(SO4)3⋅17H2O) were occasionally present in the more distal portions of the aprons (e.g., CN9 in Figure 9). Fe also partitioned into goethite (FeO(OH)) and hematite phases, generally as surface coatings on basaltic tephra. Most of the other sulfates occurred as orange/red to tan to white coatings on tephra and were also distributed just under the surfaces of aprons. Zeolites were present in some samples in minor amounts. Phyllosilicate minerals were relatively abundant and included montmorillonite ((Na,Ca)0.33(Al,Mg)2(Si4O10)(OH)2nH2O)) (e.g., Tel5 in Figure 9) and smectite-kaolinite mixtures [Marcucci et al., this issue]. Temperatures ranged from 50°C to 100°C, but this variable had little effect on the resultant alteration minerals and their abundance.

Details are in the caption following the image
Active outgassing at “Monkey Face” outcrop along the 1992 Cerro Negro crater wall. gyp = gypsum, si = amorphous silica sinter, ca = calcite. Minor amounts of szomolnokite and hexahydrite were immediately below the gypsum/silica crusts at the Monkey Cheek site. Note the person for scale.

4.3 Fumarole With the Highest pH (6) and the Lowest Temperature (60°C)

[31] One fumarole existed on the far west side of Cerro Negro's 1992 crater wall with a higher pH, likely due to contributions from meteoric groundwater. Alteration at this site included white precipitate rinds on basalt cinders within actively steaming ground. Here calcite (Ca(CO)3) was the dominant secondary mineral (47%), followed by gypsum (15%) and the calcium phosphate brushite (CaHPO4⋅2H2O) (<10%) (e.g., CN22 in Figure 9). The only clearly identifiable phosphate mineral was represented at this site.

4.4 Outwash Basins

[32] At both Cerro Negro and Momotombo, outwash catchments existed downslope from the active vents (Figure 13). At Cerro Negro, “bathtub rings” were present and likely resulted from evaporation of ponded water that dissolved, transported, and precipitated altered products during rainfall events. Gypsum was a major constituent, owing to its high solubility. In addition, numerous other hydrated sulfates were present that were not seen at the upslope vents, including pickeringite (MgAl2(SO4)4⋅22H2O), alpersite (MgSO4⋅7H2O), mendozite (NaAl(SO4)2⋅11H2O), and melanterite (FeSO4⋅7H2O) (“outwash” samples in Figure 9) (Table 3). Additional products included zeolites and epidote. Interior to the bathtub rings, the main mineral phases were natroalunite and goethite, which were likely transported intact from the vent areas above but may also have formed in situ.

Details are in the caption following the image
Outwash basin (OB) surrounded by active fumaroles. White and yellow bathtub rings are seen along the margins.

4.5 Hervidores de San Jacinto Geothermal Field

[33] This geothermal area on the flank of Telica volcano had much higher fluid-rock ratios, and a couple dozen geothermal pools and mudpots existed alongside fumaroles (Figure 8). The acidity of fumaroles ranged from pH 3.5 to 4.6 and had similar secondary minerals to fumaroles elsewhere with similar temperature and pH. No elemental sulfur was observed, and white crusts were nearly pure gypsum in the cooler regions (55°C). Hotter (80°C–90°C) orange to red samples included primarily alunogen and muscovite. The hottest gray samples at the vents had alteration products dominated by amorphous silica and cristobalite, followed by the phyllosilicate illite. Minor amounts of pyrite were detected here, which was the only place this mineral was observed. Conversely, the mudpots contained a diverse amount of phyllosilicates, sulfates, and zeolites not seen elsewhere and had temperatures ranging from 75°C to 100°C and pH values of 1.5–4.0 (e.g., “SJ” samples in Figure 9) (Table 3). The dominant phases were the clay minerals. The sulfates included alunogen, despujolsite (Ca3Mn4+(SO4)2(OH)6⋅3(H2O)), khademite (Al(SO4)F⋅5(H2O)), fibroferrite (Fe3+(SO4)(OH) ⋅ 5(H2O)), and pickeringite. Phyllosilicates included dickite (a kaolin group mineral), illite, chlorite, vermiculite, montmorillonite, and smectite. Stilbite, aerinite, and gobbinsite comprised the zeolites. Other minerals seen only in the mudpots included FeOCl and potassium alum (e.g., “SJ Red” and “SJ Orange” in Figure 9).

4.6 Comparison Between the Cerro Negro, Momotombo, and Telica Volcanoes

[34] The three geographically proximal volcanoes have nearly identical parent lithology (Table 1 and Figure 3). Major oxide compositions are mostly uniform, but with Momotombo having less Al and Ca than the other two. The gas chemistry from Cerro Negro and Telica is similar, while Momotombo has much higher S, HCl, and HF and lower CO2 than the others (Table 2). Cerro Negro gases had a higher oxygen content, but this is presumably from atmospheric contamination. The gas composition resulted in Momotombo having the lowest measured pH in the steam emissions, which was never above 1.0. The measured temperature of the gas emissions was highest at Momotombo and Telica, reaching 605°C, although we did not collect samples at these highest temperature areas due to inherent hazards.

[35] Secondary mineralogy was overall similar at each volcano, but differences were notable (Table 4). In general, more high-temperature SiO2 phases and amorphous silica were seen at Momotombo, as well as greater abundance of elemental sulfur. In fact, pure sulfur was found in ribbons that appeared to have flowed downhill from many of the Momotombo fumaroles (e.g., Figure 11). We hypothesize that recent rains had mobilized sulfur from the vents, and it precipitated as the fluid moved away from the hottest ground around the fumaroles. Momotombo also had more zeolite minerals and a variety of muscovite phases compared to the other two volcanoes. Secondary mineralogy was quite similar at Cerro Negro and Telica, and this was to be expected given their similar parent lithology and gas composition. Both showed similar abundance of goethite and hematite, generally as minor phases consisting of surface coatings around the higher pH fumaroles. Members of the alunite-jarosite group were also similarly minor components at each volcano, with greater abundance of natroalunite over jarosite. Telica did show more varieties of phyllosilicates (montmorillonite, kaolinite, and smectite), while Cerro Negro only had rare occurrences of montmorillonite. Cerro Negro exhibited a greater diversity of sulfate minerals and in larger abundance, and gypsum was the most common mineral at that volcano.

Table 4. Generalized Compilation of Alteration Phases Documented at Each Nicaraguan Volcano
Very Common Phases Somewhat Common Phases Less Common Phases
Cerro Negro sulfur; gypsum; Si (amorphous) cristobalite; natroalunite; jarosite; hematite; sodium alum; alunogen; pickeringite muscovite; calcite; fibroferrite; szomolnokite; mereiterite; kieserite; hexahydrite; montmorillonite
Momotombo sulfur; gypsum; Si (amorphous); cristobalite; tridymite hematite; zeolites; kaolinite goethite; montmorillonite; nontronite
Telica gypsum; Si (amorphous); natroalunite sulfur; jarosite; cristobalite; hematite; kaolinite; smectite muscovite; montmorillonite

5 Comparison of Instrument Measurements and Limitations for Mineralogy Determinations

[36] Mineralogy of hundreds of samples was determined by either XRD from the Terra instrument (this paper; Figure 9) or by VNIR reflectance spectroscopy [see Marcucci et al., this issue]. XRD provides the benefits of quantitative results, accurate to the few percent levels. Spectroscopy is hampered in the quantification of mineralogy due to nonlinear mixing, spectral masking, fine-grained coatings, and grain size [e.g., Poulet and Erard, 2007; Bonello et al., 2004; Poulet, 2007; Marcucci et al., this issue]. XRD with the Terra instrument has the advantage of measuring small samples (≪1 g), aiding local mineral quantification where significant diversity exists (like in fumarolic deposits). On the other hand, spectroscopy, as it has been utilized on Mars, represents an average of minerals present over a larger spot size. This can range from as small as centimeters for rovers to tens of meters (CRISM) to hundreds of meters to kilometer-scale (OMEGA) from spectrometers on satellites currently orbiting Mars. Reflectance spectroscopy can only probe the upper tens of microns of the surface. Data collection efficiency is much greater for reflectance spectroscopy compared to XRD. It takes up to an hour to attain a clean diffractogram for a single sample from XRD, and in that time period, hundreds to thousands of spectra can be collected. Also, while no sample preparation is required for reflectance spectroscopy, XRD requires a dry, finely powdered and sieved sample—formidable and time-consuming tasks for a planetary rover.

[37] Much of our understanding of the geologic, aqueous, and climate histories of Mars relies on the returned mineralogy data, and thus, a clear understanding of the measurements is crucial for proper interpretations. The orbiter-derived mineralogy is also a key factor in landing site selection. Since XRD and VNIR spectroscopy are the current methods used to determine mineralogy on Mars, we analyzed about three dozen samples from pristine to heavily altered with both techniques for direct comparison of their accuracy and utility. Figures 14 shows XRD and VNIR analyses of the same samples (see detailed spectroscopy methods in Marcucci et al. [this issue]), and Figure 15 provides library reference spectra that were good matches for the field spectra. Table 5 shows a representative comparison of detected phases for a broader range of samples. There is general concordance in the measurements from the two instruments. VNIR spectroscopy could generally identify fewer total phases than XRD. Phyllosilicates were easily seen in VNIR data but were less detectable in XRD. Clay minerals are challenging to identify in XRD without special preparation for these constituents. We did not prepare XRD samples for specific clay analysis since this cannot be done by CheMin on Mars. In general though, most of the phyllosilicates identified by VNIR spectroscopy could be also seen in XRD data. Some sulfates (e.g., natroalunite and alunogen) and muscovite were often not identifiable in VNIR data, even though they composed major phases identified with XRD. Additionally, iron oxides were sometimes seen clearly in VNIR spectra but were not identified in XRD. We hypothesize that thin surface coatings of these constituents made them less identifiable in bulk crushed samples of the rocks that are required for XRD analysis. Conversely, XRD analysis was able to detect numerous minor and trace phases that were not distinguishable in VNIR data.

Details are in the caption following the image
Representative XRD and VNIR reflectance spectroscopy of the same field samples. Library spectra are given in Figure 15 for each spectrum.
Details are in the caption following the image
Reflectance spectra from Figure 14 with best library spectra matches for comparison [Clark et al., 2007; PDS, 2013]. For detailed descriptions of spectra fitting and mineral determinations, see Marcucci et al. [this issue].
Table 5. Representative Minerals Identification by XRD Versus VNIR Reflectance Spectroscopy, Listed by Decreasing Abundancea
CN1 pyroxene; augite; hematite (m); goethite (m); cristobalite (m); muscovite (m) goethite
CN2 Si (am), pyroxene; gypsum, jarosite; cristobalite (m); muscovite (m) jarosite
CN3 Si (am); gypsum (m), S0(m) silica
CN10 Si (am); pyroxene; Na-alum; alunogen silica; basalt
CN12 Si (am); zeolite silica; montmorillonite
MO2 cristobalite; S0; Si (am) diaspore; silica
MO3 S0; cristobalite sulfur; silica
MO7 gypsum; cristobalite, Si (am) goethite; silica
MOclay natroalunite; cristobalite; anhydrite (m); Na-alum (m) kaolinite; smectite; hematite
T1 natroalunite; gypsum; pyroxene; cristobalite (m) natroalunite; hematite
T3 gypsum; pickeringite (m) gypsum; montmorillonite
T6 natroalunite; kaolinite; pyroxene; cristobalite (m); montmorillonite (m) kaolinite; smectite; natroalunite; silica
T7 natroalunite; pyroxene; cristobalite (m); hematite (m) natroalunite; hematite; montmorillonite
T9 pyroxene; muscovite; xeolite (m); jarosite (m); kaolinite (m); montmorillonite (m) kaolinite; smectite; jarosite
  • a (m) = minor phase, <10%; (am) = amorphous silica phase.

[38] Mineralogy was also assessed on some sample types with microscopic imaging by SEM [McCollom et al., 2013a]. The microscale reveals the intricate relationships between alteration mineralogy. For example, spheroidal Fe-oxides that may have formed by processes similar to the hematite blueberries seen on Mars are seen as growths on top of natroalunite crystals [McCollom et al., 2013a, 2013b]. Microscopic data are also useful for determining the mineralogy of thin surface coatings that existed on many samples, whereas these trace phases were often not identifiable in XRD data. Finally, we used Mossbauer to assess iron mineralogy of field of several samples and synthetic solid-solution mixtures of natroalunite and natrojarosite [McCollom et al., 2013a, 2013b]. We found that natroalunite samples with very minor substitutions (~10%) of Fe for Al in the crystal structure were indistinguishable from a pure jarosite end-member in Mossbauer data. Proper identification of these two sulfates in XRD or reflectance spectroscopy requires a detailed mineral library that includes the full range of intermediate compositions.

6 Geochemical Model of Fumarolic Acid-Sulfate Alteration

[39] Alteration mineralogies in gas-dominated systems can come from either direct precipitation of the vapor phase or by leaching of the wall rock, and the relative contributions from each remains an active field of study [e.g., White and Waring, 1963; Stoiber and Rose, 1974; Kodosky and Keskinen, 1990]. The boiling point of pure H2SO4 is 340°C, which is above the temperatures measured for all fumaroles sampled in this study. However, Stoiber and Rose [1974] noted that at some point between the boiling temperature and room temperature, the formation of sulfuric acid aerosols will commence, providing a liquid phase in the emissions, and such constituents have been observed for Nicaraguan volcanoes [Allen et al., 2002]. Additionally, SO2 and H2O condensates on rock surfaces or H2S and O2 will produce sulfuric acid, as shown in equation (1).

6.1 Chemistry of Fluids and Gas Condensates

[40] The chemistry of the gases and fluids can help elucidate the acid-sulfate alteration, including the contributions from gas sublimates versus wall-rock leaching. In the 2008 and 2012 field campaigns, we collected gas condensates at the Cerro Negro fumaroles and directly sampled thermal waters and mudpots at the San Jacinto geothermal field. At Cerro Negro, we sampled gas emissions from both the 1992 crater wall and on the inner rim of the 1995 summit crater. As described above, a steam condenser was used [to cold trap gas emissions for collection (Figure 10)]. Fluids were analyzed with IC-AES, and results are presented in Table 6. Major cations in the condensed gases from the 1992 crater wall Cerro Negro include Al3+, SiO2 (aq), Ca2+, Fe2+, 3+, K+, Na+, and Mg2+ (in general decreasing order), while SO42− and Cl are the main anions. The condensed gas from the 1995 summit crater had similar abundance of most cations and anions when compared to a fumarole on the 1992 crater with similar temperature and pH. However, distinct enrichments in F and Cl and less SO42− existed on the younger crater (compare CN 95 Crater to CN4 Mound in Table 6), which is likely more directly linked to the underlying magma chamber. The fluids analyzed from the San Jacinto thermal pool and mudpot had, on average, an order of magnitude higher concentration of cations compared to Cerro Negro fumarolic condensates of similar acidity and temperature. Anion abundance was also similar to the comparison fumaroles, but with higher PO43− and SO42−. Overall, the major control on ion concentration is pH, with higher acidity fluids having greater abundance. Temperature, at least over the range we measured, did not have any correlation with total ionic concentrations.

Table 6. Cation and Anion Concentrations From Fluida
Cations SiO2 (aq) Mn2+ Fe(T) Mg2+ Ca2+ Al3+ Sr2+ Ba2+ Na+ K+
CN Cave (5, 100) 5.8 0.009 DL 0.15 3.9 0.3 0.016 0.022 0.26 DL
CN Office Balcony (3.0, 80) 14.2 DL 0.023 0.019 5.4 0.8 0.005 0.007 0.09 0.13
CN4 Mound (0, 110) 11.2 DL 2.08 0.08 7.0 637.6 0.020 0.010 0.17 0.78
CN 95 Crater (1, 115) 1.9 DL 0.68 0.07 0.9 338.8 DL 0.040 DL 0.77
CN Monkey Cheek (4.5, 99) 11.1 DL 0.068 0.036 1.9 4.5 0.007 0.008 0.33 0.14
SJ thermal pool (2.0, 96) 112.2 7.03 0.43 78.65 127.1 8.3 0.580 0.110 9.10 6.32
SJ Grog Mudpot (3.2, 60) 101.1 0.91 0.13 15.92 43.5 18.1 0.220 0.120 10.08 3.79
Machine detection limit 0.026 0.003 0.014 0.008 0.021 0.006 0.001 0.001 0.010 0.060
Anions F Cl NO22− Br NO PO43− SO42−
CN Cave (5, 100) 0.97 97 0.455 DL 1.2 0.32 167
CN Office Balcony (3.0, 80) 0.74 59 DL 0.06 1.3 DL 287
CN4 Mound (0, 110) DL 12,960 DL 3.22 5.0 DL 44,736
CN 95 Crater (1, 115) 339.4 28,175 DL 1.89 2.4 DL 1901
CN Monkey Cheek (4.5, 99) 2.4 266 DL 0.23 0.7 DL 238
SJ thermal pool (2.0, 96) 11.3 31 DL 1.31 6.0 2.52 12,466
SJ Grog Mudpot (3.2, 60) 3.96 16 DL 2.45 3.6 1.51 5355
Machine detection limit 0.02 0.02 0.05 0.05 0.05 0.2 0.2
  • a CN = Cerro Negro, SJ = San Jacinto. The CN results are from condensed gas samples as described in the text. The SJ results are direct samples of hot springs and mudpots. Parentheses after each sample name denote pH and temperature: (pH, T in °C) All measured values are in ppm.

6.2 Geochemical/Mineralogical Model of Acid-Sulfate Alteration in Nicaragua

[41] A schematic of the major classes of fumarole sites at the Nicaragua volcanoes and their alteration mineralogies is shown in Figure 16. As discussed above, the lowest pH fumaroles were typically the highest temperature and consisted of a relatively simple secondary mineralogy with S0 at the vent, surrounded by an apron of amorphous and high-temperature silica phases and minor gypsum crusts. Alternatively, fumaroles with moderate temperature and acidity had abundant amorphous silica, sulfates, iron oxides, and phyllosilicates. Local outwash basins existed downslope from the active vents and were composed of goethite and natroalunite in the middle and bathtub rings of hydrated sulfate deposits at the edges, a result from ponding and evaporation.

Details are in the caption following the image
Schematic of Nicaraguan fumarolic systems with moderate and extreme acidity and their major alteration products and distributions. Also included is a local topographic basin akin to the outwash basins discussed in the text. si = amorphous silica, SiO2 represents high-temperature phases, and gyp = gypsum.

[42] Elemental sulfur had the greatest abundance immediately around the most highly focused, highest temperature vents, and small to large (up to 5 cm) acicular crystals were present, ringing the sites of vapor discharge (e.g., Figure 11). S0 mainly was found in the hottest and most acidic fumaroles, although it did exist in trace quantities in the aprons of moderately acidic fumaroles. The 1992 gas condensate from the lowest pH fumarole (CN4 Mound in Table 6) contained far more sulfate than any other measurement, and this is also the vent at Cerro Negro where we observed the greatest abundance of elemental sulfur. Given these observations, we hypothesize that the majority of the elemental sulfur documented in this study formed from direct precipitation of the vapor phase as a sublimate or by reaction of SO2 with water.

[43] Conversely, most of the other alteration products were more likely to have formed by leaching of the basaltic wall rock when droplets of sulfuric acid interacted with it. This process occurred in two settings: leaching of the wall rock by the subsurface fumarolic gas as it rose to the surface and in vapor condensation on the surface environments. The gas emissions reflect leaching en route to the surface since the gas condensates have chemistries with cation abundance that follow the host rocks. For example, in both the fresh volcanic rocks and the gas condensates, the abundance of major cations follows (Al and Ca) < Fe < Mg (Tables 1 and 6). At the lowest pH fumaroles, Al3+ ≫ Ca2+, while at more moderate acidity vents, they exhibit similar concentrations in the fluids, with Ca2+ values generally being higher. At the most acidic fumaroles, few alteration minerals containing these cations are found, except for patchy gypsum crusts, and this is likely due to the lack of mineral stability of the Ca-Al-S system under such extreme acidic and high-temperature conditions. At the moderately acidic fumaroles, gypsum is a dominant product, and these cation are available in the gases (Table 6) and very abundant in the host rock (Table 1). A generalized chemical equation from Stoiber and Rose [1974] is illustrative of the geochemical pathways of the fumaroles studied here, i.e.,

[44] Anorthite (CaAl2Si2O8) is a common feldspar that we identified at these volcanoes in unaltered samples, although the similar composition bytownite is more common. Regardless, the plagioclase feldspar reaction with sulfuric acid produces gypsum and kaolinite (equation 2) at some locations, while gypsum and amorphous silica and aluminum are dominant elsewhere. These are all abundant products at the fumaroles we studied (Tables 3 and 4), particularly for the intermediate temperature and acidity systems. Other sulfate phases we identified likely formed from leaching of the primary minerals that free up Fe2+,3+, Mg2+, and Al3+ through similar reactions as equation (2). We find more aluminum sulfates compared to iron sulfates at most field sites, and this is probably due to aluminum's greater abundance in the parent rocks and gases (Tables 1 and 6, respectively). The Fe that forms hematite and sulfates (Fe-rich natroalunite, jarosite, and others) comes from dissolution of augite, olivine, and glass. The elemental sulfur that was observed around vents likely formed from the reaction of SO2 gas with water to produce H2SO3, which then reacts to form S0, sulfuric acid, and water [McCollom et al., 2013a].

[45] Silica is only found in minor abundance in the gas chemistry, and this is likely due to its inherent immobility. This phase is the major constituent left behind from the sulfuric acid wall-rock leaching process. Silica is partitioned into two phases, including high-temperature SiO2 (cristobalite and tridymite) and an amorphous hydrated phase. The former is found in great abundance only at the highest temperature fumaroles, and the amorphous variety is widespread at the full range of pH and temperatures measured across the fumaroles (Table 3). While these phases are only stable at higher temperatures than the fumaroles we observed, we suspect that they formed during the first stages of alteration of silica-bearing glass at the end of the eruption phases. These phases are metastable at the temperatures present in the sampled fumaroles for long periods [Heaney, 1994]. Overall, the lower-temperature moderately acidic fumaroles contained lower abundance of amorphous silica in terms of total mineral diversity. At a number of these locations, silica sinter was actively being deposited as widespread centimeter-scale crusts on crater walls where rising gas emissions were interacting with the surface. Additionally, the silica partitioned into a variety of phyllosilicates (Tables 3 and 4). We conclude that temperature is a major constraint on silica alteration mineralogy in gas-dominated environments.

[46] At Cerro Negro and Momotombo, a variety of basalt float samples had well-developed alteration rinds (Figure 17). In this case, they are not sinter or precipitates, and instead, they represent residual products from leaching of the parent rocks. This relationship allows an understanding of element mobility during leaching of the fresh materials. Table 7 shows the interior bulk chemistry of one sample (CN SHAN F) compared to the surface rind (CN SHAN R) (Figure 17). Table 7 also shows the chemistries of a number of other leached rinds that all had similar pristine interiors. It is clear that nearly all major oxides have been leached out of the parent rocks with the exception of silica, which as a result makes up the large majority of the bulk chemistry. Titanium is also enriched in the alteration rinds, and this is a result of its low solubility in acid fluids, similar to silica. Sulfur has been also added to these rocks during the leaching process, but at the percent level. In the thin section, the rinds maintain the phenocryst textures of the original minerals, but the constituents have been replaced. Mineralogically, the rinds are composed of (in decreasing abundance) cristobalite, tridymite, sulfur, and amorphous silica (e.g., “CN SHAN Altered” and “Mo Rind” in Figure 9). Given the high-temperature silica phases, we postulate that these rocks experienced leaching at the end of the last eruption phase. All of the Cerro Negro samples are at most 20 years old, indicating that leaching can be a very rapid process under high temperature and acidic conditions.

Details are in the caption following the image
(a) Alteration rinds resulting from acid-sulfate leaching of basaltic rocks. (b) CN SHAN F (interior) and CN SHAN R (rind) (see also Figure 9 and Table 7).
Table 7. XRF Data for Major Oxides and Sulfur Percentages in Leached Rinds on Basalts (Figure 17) in Comparison to a Fresh Interior (CN SHAN F)
SiO2 Al2O3 Fe2O3(T) MnO MgO CaO Na2O K2O TiO2 Total S
CN SHAN F 50.00 20.49 10.45 0.169 4.54 11.66 2.27 0.45 0.76 0.02
CN SHAN R 80.23 1.90 1.41 0.032 0.89 1.52 0.28 0.17 1.05 0.34
CNR 82.76 2.38 0.53 0.011 0.16 0.33 0.09 0.42 1.10 0.090
M010R 80.49 2.27 1.86 0.017 0.52 1.32 0.16 0.16 1.25 1.31
CN4R 71.74 5.16 5.98 0.113 2.87 4.41 0.74 0.5 0.76 0.06

[47] Finally, the San Jacinto geothermal field provides insight into the role of fluid-rock ratios on acid-sulfate alteration. While the parent rock lithology, gas composition, temperature, and acidity were similar to many fumaroles studied here, the alteration mineralogy was distinct (Table 3). Fluid chemistry of the San Jacinto thermal pool and mudpot had higher concentrations of cations and anions than fumaroles with similar environmental parameters. The higher fluid-rock ratio provides enhanced leaching of the host rocks, which are dominated by Si and Al and then Ca and Fe. Consequently, the secondary alteration mineralogy is dominated by phyllosilicates (kaolin group and the smectite montmorillonite) (Table 3) (Figure 9). A variety of hydrated sulfates occurs in these settings and can be attributed to the abundance of water. Many additional mineral phases that were not found elsewhere are common products at San Jacinto.

7 Applications to Mars

[48] Given the long-lived history of volcanism and crustal water, Mars likely hosted many fumarole and hot springs environments through time. Individual volcanoes were intermittently active for billions of years [Robbins et al., 2011], and outgassing around the vents likely led to fumarolic alteration of the summit caldera materials. The MER Spirit detailed local pyroclastic deposits at Home Plate [Squyres et al., ], and both high-temperature fumaroles and lower-temperature hydrothermal springs have been invoked to explain the alteration mineralogy [e.g., Schmidt et al., 2009]. Additionally, hydrothermal systems have been hypothesized for alteration mineralogies seen within Valles Marineris [e.g., Chojnacki and Hynek, 2008] and elsewhere on Mars [e.g., Ehlmann et al., 2011a, 2011b]. Solomon et al. [2005] argued that large-scale hydrothermal circulation likely operated throughout the Noachian Period on Mars, and this may in part be responsible for the widespread sulfate and clay minerals that have been identified around Mars.

[49] An understanding of the geochemical pathways involved in the transformation of these deposits from fresh basalts to heavily weathered rocks, and the paleoenvironmental conditions present during their alteration (e.g., temperature, pH, and fluid-rock ratio), can provide significant insight into the geological evolution and astrobiological potential of early Mars. We completed field studies of acidic, gas- to fluid-dominated hydrothermal systems in basalt-hosted volcanoes in Nicaragua. A major goal was to correlate alteration mineral assemblages with environmental conditions that include parent rock lithology, gas/fluid composition, temperature, and pH. Additionally, we have conducted laboratory experiments and theoretical geochemical modeling [McCollom et al., 2013a, 2013b; Marcucci and Hynek, this issue] to further constrain alteration mineralogy based on these parameters and for application to Martian materials and inferred environmental conditions. Collectively, these efforts provide a conceptual framework for interpretation of relic hydrothermal deposits on Mars and inferences into their astrobiological potential.

[50] Figure 16 shows some general trends of gas-dominated acid-sulfate alteration of the volcanic materials studied here. In general, we found that very low pH and high-temperature fumarolic alteration rapidly led to mostly S0 proximal to the vents, amorphous silica, and high-temperature SiO2 phases in surrounding aprons with occasional gypsum crusts on the surface and trace amounts of montmorillonite. Moderately acidic gas vents and surrounding aprons exhibited amorphous Si and abundance of sulfates, with gypsum being the dominant mineral. More phyllosilicate phases are seen and include representatives from the kaolin, smectite, and illite groups, as well as muscovite. The highest pH site, with moderate temperature, consisted mostly of calcite and minor gypsum. In fluid-rich environments (thermal springs and mudpots), clay minerals dominated the composition, followed by a variety of hydrated sulfates. Outwash basins below the active vents contained numerous hydrated sulfates that resulted from evaporation of pooled water. A caveat in all these results is that they represent active environments under the Earth's oxidizing atmosphere with significant weather events (i.e., rainfall). Indeed, the long-term stability of all these mineral phases under cold-dry Martian conditions is not known, but many of the minerals we identified at the active fumaroles on Earth are also abundant on the surface of present-day Mars. Therefore, our results provide an initial conceptual model for interpreting local environmental conditions at ancient Martian fumarole sites.

[51] The Terra instrument used in this study, which is functionally equivalent to Curiosity Rover's CheMin XRD/XRF, provided robust and quantitative in situ mineralogy and was able to detect a variety of phyllosilicates with no special preparation specific to clay minerals. However, VNIR reflectance spectroscopy measurements provided additional clay phases that were not seen in XRD. Similarly, hematite and goethite often existed as thin surface coatings that could not be detected in XRD but were distinct in VNIR data. Microscopic imaging with thin-section petrography and SEM showed the intimate relationships between mineralogy and very thin coatings that were not detected with XRD or VNIR data [see McCollom et al., 2013a]. We conclude that multiple techniques are required for a thorough characterization and understanding of hydrothermal sites on Earth and Mars.

[52] The fumarolic areas of the studied Nicaraguan volcanoes showed a high diversity of alteration minerals over centimeter to a few meter scale, and a similar variety is expected from relic Martian systems. Existing orbiting spectrometers at Mars have spatial footprints significantly larger than the altered regions studied here, and thus, similar-scale Martian hydrothermal features may only be detectable by higher spatial resolution instruments or landers. Fluvial or volcanic processes on Mars may have distributed acidic alteration products over a broader region, and our results suggest that the fluvial deposits would contain abundance of hydrated sulfates. Eolian processes could certainly rework relic hydrothermal deposits and would act to homogenize the chemistry and mineralogy of the materials.

[53] The presence of phyllosilicate-bearing deposits on Mars has often been used to argue for neutral to subalkaline pH of the fluids at the time of formation [e.g., Bibring et al., 2006, 2007; Chevier et al., 2007; Wray et al., 2009a, 2009b; Murchie et al., 2009; Ehlmann et al., 2009, 2011a, 2011b]. Fernández-Remolar et al. [2011] showed the presence of phyllosilicates in the acidic Rio Tinto system in Spain. However, they note that many of the phyllosilicates were originally formed in ancient geologic settings and transported into the Rio Tinto in modern times. In this study, we have documented phyllosilicate formation directly from acid-sulfate alteration of multiple active volcanic systems in Nicaragua. The extreme acid sites (pH < 1) produced only very minor amounts of clay minerals, while the moderately acidic (pH 4–5.5) fumaroles had significant amounts and varieties of phyllosilicates (Tables 3 and 4). In the San Jacinto system with a higher fluid-rock ratio, phyllosilicates were the dominant minerals, even though the pH ranged from 1.5 to 4.0. All groups of clay minerals identified on Mars (except the rare serpentite group) were identified in the acidic fumaroles and hot springs, including kaolins, smectites, illites, and chlorites, with the first two being the most abundant. Specifically, identified minerals include (in decreasing abundance) kaolinite, montmorillonite, muscovite, dickite, nontronite, vermiculite, illite, and chlorite. Our results show that Al-bearing phyllosilicates are common acid-sulfate weathering products in Nicaraguan hydrothermal systems, while Mg-bearing clays are only a minor phase. Consequently, significant care should be given to interpretations of near-neutral to alkaline fluids on ancient Mars based solely on the presence of phyllosilicates. Similarly, cooccurrences of phyllosilicates and sulfates do not require multiple aqueous episodes with neutral and later acidic conditions, as has been reported in the literature for Meridiani [e.g., Wray et al., 2009b; Wiseman et al., 2010], Gale crater [e.g., Milliken et al., 2010], or elsewhere on Mars [e.g., Ehlmann and Mustard, 2012]. We conclude that diverse Al-bearing clays are widespread products from acid-sulfate alteration, whereas Mg-smectites form in much smaller abundance under strongly acidic conditions, although they are still present.

[54] The Home Plate area in Gusev crater on Mars has been identified as a putative hydrothermal environment based on inferred volcanic sedimentology, and chemistry and mineralogy results from the Spirit Rover [Squyres et al., ]. Two distinct strata are present, and the workers argue that the lower unit likely represents accumulation of tephra, whereas the upper unit may represent eolian reworking of the same materials. Chemically, the outcrops are generally classified as moderately altered alkali basalts [Squyres et al., ; Ming et al., 2008]. Nearby the Home Plate outcrops exist two distinct suites of alteration minerals that imply acidic hydrothermal activity. Yen et al. [2007] detailed hydrated ferric and magnesium sulfates and excess silica at the Paso Robles site, and they favored a high-temperature system of hydrothermal fluids and volcanic vapors rich in sulfur. The minerals inferred for this Martian locale are similar to alteration minerals we find in Nicaragua, supporting the fumarolic interpretation proposed by Yen et al. [2007]. Yet the alteration mineralogy is not entirely consistent. Both studies show ferric sulfates and abundant silica, but we find more Ca-sulfates, Al-sulfates, and clays, while Yen et al. [2007] noted greater abundance of Mg-sulfates. However, the discrepancies can likely be attributable to the differences in the major cations available in the parent rocks in Nicaragua and on Mars (Table 1 and Figure 2), measurement techniques, and unknown differences in gas chemistry.

[55] The second putative hydrothermal deposits near Home Plate are mainly found in soils at the East Valley site but also in nearby soils and outcrops [Squyres et al., ; Ruff et al., 2011]. These materials have a strong enrichment in silica, ranging from ~70% to 90% of total oxide abundance [Ming et al., 2008; Morris et al., 2008]. The hydrated silica phase is generally identified as opaline silica [Ruff et al., 2011]. While we did not attempt to distinguish the various forms of silica in this paper, Marcucci et al., [this issue] documented that most of the amorphous silica at the Nicaraguan volcanoes are hydrated and match VNIR library spectra of opaline silica and silica sinter. Ruff et al., [2011] argued that the silica enrichment in the soils at East Valley was likely due to either acid-sulfate leaching of the host materials or via precipitation from Si-enriched fluids in a hot spring. They favored the latter interpretation, arguing that the lack of significant sulfur enrichment was inconsistent with acid-sulfate leaching. Further, they argue that a very high fluid-rock ratio was required in either scenario to account for the high silica.

[56] Our results pertaining to acid leaching of basalts (Table 7) provide some bearing on these issues. First, while sulfur is enriched in the rinds relative to the pristine basalts, it is not a significant enrichment (never above 1.5% total bulk composition, which is below all but one measured values for East Valley reported by Ruff et al. [2011]). Second, high fluid-rock ratios are not required for the given chemistries of the East Valley materials on Mars. Ruff et al. [2011] argued that this ratio must be high to transport excess sulfur away from the silica deposits in the leaching scenario, but as noted above, the very low fluid-rock ratios in the Nicaraguan fumarolic deposits did not result in large enrichments in sulfur (Table 7). A significant number of the fumarole aprons we studied were composed almost entirely of amorphous silica with little sulfur. Finally, high fluid-rock ratios are not required for precipitation of silica sinter. We found such deposits coating the surfaces around a number of fumaroles (e.g., Figure 12). Reworking this material into a soil, as is inferred from the East Valley deposits, could provide a substrate dominated by silica. Thus, our field results support silica enrichment via acid leaching or through precipitation of siliceous sinter and the fluid-rock ratios and the depositional mechanism at East Valley remains unconstrained.

[57] In conclusion, fumarolic environments are widespread around active volcanic systems on Earth, and it is very likely that many similar settings existed on Mars. Here we documented the alteration of basaltic rocks similar to Martian compositions in gas-dominated, high-temperature, acid-sulfate systems. We characterized the alteration products and provided a geochemical model for formation of these materials. We find that local environmental parameters such as temperature, pH, and fluid-rock ratios are important factors in the resultant secondary mineral suites. Thus, the specific mineral suites identified on Mars from similar hydrothermal settings can be used to infer the paleoenvironmental conditions, which in turn provide constraints on their potential habitability. Additionally, many phyllosilicates were identified in the acidic, low fluid-rock fumarolic settings in Nicaragua that include many of the same clay minerals reported from Mars. The Martian examples have frequently been used to infer neutral to alkaline conditions with high fluid-rock ratios, and our results are in direct conflict with this inference.


[58] We thank Analytical Spectral Devices, Inc., for loan of the TerraSpec4 spectrometer used in this study. We also thank two anonymous reviewers for their constructive suggestions. Additionally, we thank the Nicaraguan government agency Instituto Nicaragüense de Estudios Territoriales for in-country support of this study. This work was supported by NASA Exobiology Award NNX08AQ11G, NASA Early Career Award NNX12AF20G to B. M. Hynek, NASA NESSF Award NNX10AU39H, and the CU Department of Geological Sciences.