Biosignature Preservation and Detection in Mars Analog Environments

3.1.2.1. Habitability pros

Hydrothermal convection systems span a range of depths where deep groundwater encounters and gains buoyancy from a heat source, ascends toward the surface, cools, and then either emerges at the surface or descends back down along the flanks of the ascending thermal plume.

Submarine hydrothermal systems are globally pervasive as key components of the mid-ocean ridge system, which is the world's most extensive mountain range with a total length of 80,000 km. Early in Earth's history, greater abundances of radionuclides in the mantle created higher geothermal heat flow (Arevalo et al., 2009) that, in turn, probably sustained greater hydrothermal activity. Such activity probably hosted extensive microbial ecosystems (e.g., Van Kranendonk, 2006; Westall et al., 2015a). Widespread hydrothermal activity has also been proposed to sustain hydrologic cycles at regional scales on early Mars (e.g., Gulick, 1998, 2001).

In continental environments, hydrothermal systems can enhance the availability of liquid water in several ways. A thermal source can melt ground ice locally. The buoyant heated water can invade otherwise drier, shallower regions in the crust. Hydrothermal systems create springs or fumaroles that become oases in arid environments (e.g., the El Tatio region, Atacama Desert, Chile; Ruff and Farmer, 2016) that otherwise might be relatively unfavorable for extant life.

Hydrothermal systems can also develop in the aftermath of impact events that inject substantial amounts of thermal energy into the crust (e.g., Schwenzer et al., 2012a, 2012b). Impacts have been proposed to enhance the formation of phyllosilicates and to create local habitats on early Mars (Schwenzer and Kring, 2009).

Hydrothermal systems provide basic elements necessary for life such as carbon, hydrogen, oxygen, nitrogen, sulfur, phosphorus, magnesium, iron, potassium, molybdenum, zinc, and a host of other essential elements in trace quantities. Hydrothermal fluids extract elemental nutrients from host rocks via dissolution and chemical alteration to create solutes and mineral assemblages that are stable within the hydrothermal environments. The solutes ascend and become available at shallower, cooler depths and at the surface. Volcanic gases provide carbon, sulfur, and molecular hydrogen. Nitrogen is provided principally by meteoric water that has entered the hydrothermal system. Thus, hydrothermal systems are also “oases” for nutrients.

Hydrothermal systems enhance the opportunities for microorganisms to acquire energy either from sunlight or from energy-yielding oxidation-reduction (redox) chemical reactions. Submarine systems sustain extensive chemotrophic microbial communities (e.g., Baross, 1998), and investigations have revealed potential fossilized examples of such communities in Archean rocks (Westall et al., 2015a, 2015b; Van Kranendonk et al., 2016). Thermal springs and fumaroles can deliver liquid water to surface environments that are otherwise dry but where sunlight can provide energy for photosynthetic microorganisms. Hydrothermal systems also sustain redox reactions that provide energy for life. Deeper crustal environments tend to reflect the relatively reducing conditions of the upper mantle and lower crust, whereas surface environments are typically more oxidizing. As a consequence, convecting hydrothermal fluids create interfaces where more reduced species from depth are brought into direct contact with more oxidized species from the surface. Slow reaction kinetics result in redox disequilibria in such environments that can be exploited as metabolic energy sources by microbial communities (e.g., Spear et al., 2005; Shock et al., 2010). Examples include the mixing of H₂ and H₂S from volcanic gases and Fe²⁺ from rocks with O₂ from the surface. Hydrothermal systems actually sustain a multitude of redox reactions involving the constituents from rocks, volcanic gases, and the surface environment (McCollom and Shock, 1997; Shock et al., 2010; Amend et al., 2011). Microorganisms utilize virtually all of these reactions as sources of energy for metabolism.

Hydrothermal systems excel at meeting a key requirement for habitability, namely, that all the key resources must be available simultaneously (Walter and Des Marais, 1993). Both hydrothermal systems and life require liquid water in order to function and persist. By its very nature, geothermal convection provides chemical nutrients and sources of redox energy to aqueous solutions. Reactions with basaltic or ultramafic host rocks typically buffer hydrothermal fluids to near neutral to mildly alkaline pH values, but boiling or input of magmatic gases can cause very acidic conditions to develop (for example, from oxidation of SO₂ or H₂S to sulfuric acid).

Hydrothermal systems sustain a broad range of temperatures, pressures, redox states, and fluid compositions. These diverse conditions sustain diverse microorganisms (Brock, 1978; Inskeep and McDermott, 2005). Such diversity enhances the capacity of microbial communities to adapt to changing environmental conditions and thus to survive over the long term. In addition to phototrophs, many microbial communities in hydrothermal environments utilize chemical energy sources. Thermophilic chemolithoautotrophic and heterotrophic bacteria and archaea are prevalent at temperatures above 70°C in terrestrial hydrothermal systems. For example, the Aquificales are a very conspicuous filamentous biomass in these systems from pH ∼3 to 9, at temperatures from 60°C to 95°C (Reysenbach et al., 2005). These communities grow chemolithoautotrophically (often microaerophilically) using a wide variety of electron donors and acceptors, and are often associated with mineral precipitation. These biota also provide insights into the early evolution of the biosphere. The higher-temperature communities often harbor deeply branching archaeal groups such as the Aigarchaeota (e.g. Beam et al., 2015) and Korarchaeota (Reysenbach et al., 2000), whose inferred metabolisms range from chemolithotrophy to heterotrophy.

3.1.2.2. Habitability cons

In continental environments, the habitability of hydrothermal systems may be negatively impacted by their relatively short duration over geologic timescales (thousands to billions of years). The Yellowstone hydrothermal system, a product of hot spot volcanism, has been operating for ∼2.1 million years (Christiansen, 2001). For impact-generated hydrothermal systems, such as the Sudbury impact structure in Canada (62 km × 32 km × 15 km deep), it is thought that the crater size dictates the longevity of the system, which has been estimated to be ∼1 million years (Ames et al., 1998; Osinski et al., 2012).

This shortcoming is not as severe for extensive submarine hydrothermal systems because the global mid-ocean ridge system has been continuously active for billions of years (e.g., Cawood et al., 2006). Westall et al. (2015a) proposed that subaqueous hydrothermal springs were more abundant on early Mars.

Another aspect of continental hydrothermal systems that may negatively impact their habitability is that they tend to be geographically restricted. While there may be over 12,000 hydrothermal features in Yellowstone, the pools, outflow channels, or fumarole areas tend to be on the scale of meters to hundreds of meters. Even the world's largest hot spring in the Waimangu hydrothermal system in New Zealand is ∼200 m in diameter (Keam, 1980).

The elevated temperature and low pH of some hydrothermal systems can function to limit microbial diversity and biomass production, while some chemical species can poison certain microbes altogether. The most thermophilic oxygenic phototroph known is a cyanobacterium, a species of Synechococcus, whose maximum growth temperature is 73°C (Peary and Castenholz, 1964). This temperature can be considered the upper limit for photosynthesis. Beyond this temperature, chemosynthetic microbes (chemotrophs) flourish. The most thermophilic anoxygenic phototroph is the green nonsulfur bacterium Chloroflexus, whose upper temperature limit is 69°C (Castenholz and Pierson, 1995).

Acidity can function to limit microbial diversity, especially among phototrophs. Cyanobacteria cannot withstand pH levels of less than 5.5–6.0 (Castenholz, 1988) and thus are not widespread community members of acidic hydrothermal systems. However, red unicellular eukaryotic algae belonging to the order Cyanidiales are able to tolerate pH levels from 0.2 to 4.0 and temperatures up to 56°C (Castenholz and McDermott, 2010). The most acid-tolerant anoxygenic phototrophs are the purple nonsulfur bacteria, which can withstand environmental pH levels of ∼3 but whose growth optima are ∼5 and above (Pfennig, 1974; Madigan et al., 2005).

Because oxygenic photosynthesis is generally not limited by the availability of water as its electron donor or CO₂ for carbon fixation, it is the most productive type of photosynthesis. In contrast, anoxygenic photosynthesis is usually limited by the flux of reductants from the environment (such as H₂ and H₂S from hydrothermal systems), and rough calculations suggest that these types of communities might be 2–3 orders of magnitude less productive than oxygenic ones (Des Marais, 2000). Chemotrophs tend not to synthesize as much biomass as phototrophic communities (Campbell et al., 2016).

Other selective pressures from the geosphere include aqueous species that can inhibit photosynthesis. Sulfide at tens of micromolar and greater concentrations has been shown to poison photosystem II in cyanobacteria, effectively shutting off oxygenic photosynthesis (Oren et al., 1977).

3.1.3.1. Biosignature preservation pros

Hydrothermal systems can entomb and preserve evidence of microorganisms that lived there (Walter and Des Marais, 1993; Cady and Farmer, 1996; Walter et al., 1996; Westall et al., 2015a, McCollom et al., 2016; Munoz-Saez et al., 2016; Fig. 5). Solute-laden fluids that approach or emerge at the surface can form mineral precipitates. Because the solubility of silica is strongly temperature-dependent, hot fluids emerging from great depths typically deposit siliceous sinters. Bicarbonate-rich waters outgas CO₂ at the surface and deposit calcareous sinters (Fouke et al., 2000). Ferrous iron can oxidize to create various iron oxides and other species (Parenteau and Cady, 2010; Sklute et al., 2016) (see the “Iron-rich environments” section below for detailed discussion). Dissolved barium emerging in near-surface environments can form precipitates with oxyanions. Accordingly, the same set of processes and physical conditions that sustain habitable environments also provides mineral deposits that can entomb and preserve biosignatures.

OPEN IN VIEWER

FIG. 5.

Examples of potential biosignatures in hydrothermal spring deposits: palisade textures in terracettes sinter fabrics (Djokic et al., 2017).

The primary precipitates encrust and permineralize microbes to preserve a suite of textures, biofabrics, microfossils, and organic compounds such as lipid biomarkers. Recent work (Campbell et al., 2016; Hinman et al., 2016; Stevens et al., 2016; Szynkiewicz and Mikuchi, 2016) reviewed the variety of mineral deposits and biosignatures that occur along environmental gradients in thermal spring waters having a range of compositions. They described in detail the variety of preserved textural-mineral macroscopic and microscopic textures. These deposits and their features have parallels in some early Archean hydrothermal deposits on Earth (e.g., Westall et al., 2015a; Van Kranendonk et al., 2016).

There is a rich history of studies on the mechanisms of silica permineralization in microfossil preservation in hydrothermal systems (e.g., Cady and Farmer, 1996; Farmer and Des Marais 1999; Phoenix et al., 2000; Konhauser et al., 2001; Mountain et al., 2003; Benning et al., 2004; Ferris and Magalhaes, 2008). All examples of hot spring deposits that have been identified, both modern and ancient, contain demonstrable microfossils as well as features that appear to be microfossils but whose biological origin is difficult to prove (e.g., Walter et al., 1996; Hickman-Lewis et al., 2016; Van Kranendonk et al., 2016). Confirmed microfossils are morphological fossils that retain carbonaceous remains of microbial cells (Schopf and Walter, 1983; Buick, 1990; Schopf, 1999; Cady et al., 2003). Van Kranendonk et al. (2016) documented fossiliferous thermal spring deposits in Archean rocks in Western Australia. The sites illustrated the geometry of the hydrothermal systems and the associated mineral deposits, biosignatures, and related features that help characterize these ancient environments. This study documented how multiple indicators of ancient processes, environmental conditions, and biosignatures can confirm the presence of fossiliferous thermal spring deposits.

There have also been studies that examined the production and preservation of microbial lipids in modern and subrecent siliceous sinters that have not undergone significant diagenesis (Jahnke et al., 2001, 2004; Pancost et al., 2005, 2006; Talbot et al., 2005; Gibson et al., 2008; Kaur et al., 2008, 2011). Jahnke et al. (2016) demonstrated that siliceous deposits associated with a mildly alkaline thermal spring can be rich repositories of lipid biomarkers, including hopanoids and branched alkanes, from cyanobacteria and other organisms. The abundances of these biomarkers are related to the nature of the depositional environments and the diversity of their biota. The extent of preservation of lipid biomarkers in ancient sinters might be quite variable but should be possible in at least some cases.

McCollom et al. (2016) addressed the habitability of a variety of surface, shallow, and subsurface environments. Regarding hydrothermal systems, he noted that, volumetrically, Earth's ocean crust is the largest potential habitat on Earth, yet we know remarkably little about the organisms that live there. Also we have relatively little knowledge about the potential for biosignature preservation in these environments. The mid-ocean ridge hydrothermal system, that is, the region that hosts thermally driven convection beneath the seafloor, occupies about one-quarter of the areal extent of Earth's crust (Heezen et al., 1959; Edwards et al., 2005). This system might have hosted the most continuously habitable environments on Earth.

3.1.3.2. Biosignature preservation cons

Thermal spring deposits in subaerial environments (e.g., inactive spring mounds) are subjected to oxidizing conditions that severely impact the potential to preserve organic matter. This situation is much less severe for spring mineral deposits emplaced in subaqueous environments (e.g., diatom marshes). Key factors include the length of time that the deposits are exposed to oxidants, the abundances of oxidants, and the time required to lithify sediments and thereby greatly reduce rates of diagenesis as microbial remains are protected from later-stage oxidizing fluids.

Because the greatest variety of biosignatures are associated with phototrophic communities in hydrothermal systems, sinter facies above the upper temperature limit for photosynthesis exhibit a severe decrease in the types of textures preserved (Campbell et al., 2016). Additionally, these biosignatures tend to be heterogeneously distributed, introducing difficulties when searching for samples to collect in the field.

3.2.2.1. Habitability pros

Delta plains on Earth are fecund, highly diverse habitats dominated by phototrophy (Naeem et al., 2000; Mitsch and Gosselink, 2000; Gopal and Chauhan, 2001). Microbial communities in these habitats can be highly productive given that waters are shallow and often in the photic zone. Marine deltas, in particular, are able to accommodate significant biological diversity since this is a mixing zone for fresh and salt water (Mitsch and Gosselink, 2000). Habitability of ancient deltas on Mars may have been supported by (1) the in situ protection from radiation afforded by the water column and its turbidity and (2) chemical gradients within the water column and at the sediment-water interface (Santschi et al., 1990). Element and nutrient exchange can occur between water and sediments in pores and at the surface interface (Lerman, 1978; Callender and Hammond, 1982). Delta plains can also support diverse subaerial environments including soils and wetlands (see the “subaerial environments” section below for a detailed discussion).

While microbial biodiversity can be limited in some lacustrine systems on Earth by low nutrient availability, high salinity, and extreme pH, microorganisms are otherwise often abundant in both the water column and the sediments (Wetzel, 2001; Lynch et al., 2016). Deep-water open lake systems can host unique features such as hydrothermal vents or brine seeps that can help provide energy to sustain diverse microbial ecosystems. Two key examples of this are the hydrothermal vents at the bottom of Lake Tanganyika of the East African Rift and the brine seeps at the bottom of Lake Huron in North America (Tiercelin et al., 1993; De Wever et al., 2005; Voorhies et al., 2012). Both open and closed-basin systems can host the formation of microbialites—characteristic sedimentary structures formed by the trapping of mineral detritus in the extracellular polymeric substances formed by microbial mats (Noffke et al., 2001). Active microbialite communities can be found in the closed-basin Great Salt Lake and the open-basin Pavilion Lake and modern stromatolite formations found in Shark Bay, Western Australia (Papineau et al., 2005; Russell et al., 2014, Lindsay et al., 2017). Microbially induced sedimentary structures are preserved and exceptionally well-studied in the modern Pilot Valley Basin, Utah, and on geologic timescales in the ancient Strelley Pool Chert stromatolites and Dresser Formation microbialites of the Archean-aged Pilbara Craton (Van Kranendonk et al. 2003; Allwood, et al., 2006; Noffke et al., 2013; Russell et al., 2014; Lynch et al., 2015; Westall et al., 2015b; Lindsay et al., 2017). Likewise, microbially induced sedimentary structures can be found in most aquatic environments, both marine and lacustrine, including tidal flats, lagoons, riverbeds, dunes, and sabkhas (see Noffke et al., 2001; Bose and Chafetz, 2012; Noffke et al., 2013, and references therein).

Because groundwater supplies much of the water to playa lakes, these environments can provide a conduit between the subsurface and the surface habitats (Rosen, 1994). These environments can be stratified due to salinity, density, or temperature with episodic mixing, which can compartmentalize habitability in the water column. While desiccation of ephemeral lakes can make long-term habitability and evolution difficult for some microorganisms in the water column, both microbial and multicellular life have evolved metabolic strategies to survive these events (Jahnke et al., 2014). Additionally, for some specialized microorganisms, the salts and sediments provide needed protection from radiation and are often ecosystems independent of the intermittent water column altogether, such as the endolithic communities found in the halite rocks of the Atacama Desert and the microbial ecosystem present in the sediments of the Pilot Valley basin in Utah (Davila et al., 2015; Lynch et al., 2016). Microbes adapted to the physical and chemical conditions of these subaqueous environments could have once lived in the vast playa regions of Mars identified by morphology and evaporite minerals.

3.2.2.2. Habitability cons

The key nutrient and sediment source for many lacustrine and deltaic systems is fluvial input; however, fluvial input can be subject to seasonal or longer-term climatic variations, including ice cover. Likewise, the sizeable temperature and chemistry fluctuations that can affect lakes and other closed basins can also negatively impact habitability and decrease productivity.

The high biomass production associated with deltaic habitats on Earth is related to energy availability dependent on fluvial inputs of carbon and nutrients from productive surface environments (Milliman, 1991; Gopal and Chauhan, 2001). A sterile continental landscape would have limited nutrient runoff; thus, deltas might not be advantageous niche space for heterotrophs on a world with limited surface habitability and surface biomass (Westall et al., 2015a, 2015b). However, very little organic carbon in deep marine sediments is of terrestrial origin (Berner, 1989; Hedges and Keil, 1995). Instead, most marine organic matter is derived from marine microorganisms, indicating that at least large and deep habitable subaqueous environments can be self-sustaining. Certainly, there was ample life in the oceans of Earth before the surfaces were colonized by life; however, it is unclear whether or not smaller lake systems in a largely sterile surface environment would provide sufficient nutrients to support an isolated ecological community.

3.2.3.1. Biosignature preservation pros

Subaqueous deposits preserve a diverse suite of possible biosignatures derived from both upstream subaerial life and in situ aquatic life. Potential biosignatures would include preserved organic molecules, organic structures (e.g., cells), mineral indications of life, and isotopic markers. Deltaic and lacustrine formations have been well studied on Earth where phototrophs dominate the productivity of the ecosystems. Both deltas and ephemeral lake systems can preserve biosignatures on geologic timescales (Potter-McIntyre et al., 2014).

Some deltaic deposits can represent very long timescales of deposition (tens of thousands to tens of millions of years), and the terrestrial record demonstrates that barring any subsequent erosion, deltaic sediments can be well preserved on geologic timescales (Lowe, 1980). However, identifying long-lived versus short-lived deltaic systems from orbit is challenging, and given the relatively short-lived time of surface water on Mars, there may be no long-term (by Earth standards) deltaic systems present (Eigenbrode et al., 2016). Independent of organic matter, delta sediments can preserve the record of climate and hydrodynamic changes within the grain size and sedimentation patterns as well (Goodbred and Kuehl, 2000; Postma, 2001; Castelltort and Van Den Driessche, 2003).

Organic matter that is swept into a delta will be entombed by the near constant and often rapid sedimentation typical of deltas (Müller and Suess, 1979). On Earth, most (>70%) of this organic matter is derived from terrestrial biota, ranging from dissolved and particulate organic carbon to refractory molecules and larger detritus (Berner, 1989; Milliman, 1991; Hedges et al., 1997; Seki et al., 2010). Preservation of terrestrial organic matter in these aquatic settings can obscure biosignatures derived from less productive communities within the water column (Graham et al., 2016; Johnson et al., 2015). Organic matter is concentrated hydrodynamically in dense anoxic bottomsets where reducing conditions prevail, promoting preservation of organic compounds by slowing oxidative diagenesis (Müller and Suess, 1979; Berner, 1989). The precipitation of diagenetic cements is often influenced by microbes, and it is not well known how diagenesis might proceed in a delta where biology is absent or limited. Cements can pristinely entomb cells; however, some cement precipitates can occlude fabrics that would indicate growth (Konhauser et al., 1994, 1998; Wacey et al., 2011).

The distribution of sediments based on grain size is key to organic preservation in lakes and deltaic systems. The finer particles settle out last and generally have greater surface area to support reactions with organics, thus facilitating their deposition, burial, and long-term preservation (Mayer, 1994). The relationship between grain size and enhanced organic preservation is evident in the greater abundance of organic carbon in deep facies when compared with near-shore deposits (Eigenbrode and Freeman, 2006). Sulfurization by sulfide mineral encapsulation or direct sulfurization of organic molecules is key to the preservation of biosignatures in many terrestrial records (e.g., hydrothermal, marine, lakes), and such conditions are favored in water bodies that have anoxic water columns or sediments (Raven et al., 2015). This may also have been an important process on Mars given the widespread occurrence of sulfate minerals.

Closed-basin lakes have excellent preservation potential due to the commonly reducing to anoxic environments present in the sediment-water interface and the littoral and pelagic sedimentation processes that occur in lakes over 4 m in depth (Currey, 1990). Additionally, in playas, the precipitation of salts can entomb biosignatures and fluid inclusions and protect these features from radiation and oxidants (Benison et al., 2008; Azúa-Bustos et al., 2012; Conner and Benison, 2013). Given groundwater input, playa environments can also record biosignatures of subsurface communities, as well as surface habitats (Vreeland et al., 2000).

3.2.3.2. Biosignature preservation cons

Organic biosignature preservation is often poor in high-energy depositional settings within deltas. Typically, deposits in high-energy settings are coarser sediments. Long-term organic preservation in high-energy settings is compromised by oxidizing fluids and gases passing through the more permeable and porous sediments. For this reason, biosignature preservation is generally confined to more quiescent settings such as lakes or within the deeper delta facies, where low permeability and porosity mudstones are deposited and reducing conditions are maintained (Mayer, 1994).

Estimating the timescale of deposition in subaqueous settings can be problematic. Relative timing of deposition can be determined stratigraphically, but the duration may be either the accumulation of many years of deposition or a single rapid depositional event. Duration may be determined by close examination of sediment features, but this would not be possible from orbit.

3.3.2.1. Habitability pros

Subaerial environments support a variety of microbial metabolic pathways both at the surface and in the near-surface environment. These environments can support both chemotrophic and phototrophic microbial communities, which may be present in the form of microbial mats, endoliths, or other microbial communities of varying complexity and density. Exposure to solar radiation at the surface can support phototrophic microbial communities, and workshop participants suggested that higher-energy radiation characteristic of the modern martian surface may support radiotrophic communities in these environments. In addition, the surface-atmosphere interface creates many different types of redox gradients that support diverse communities of chemotrophs at the surface and in the near-surface. These redox gradients are formed due to downward flow of precipitation in soils, groundwater fluctuations, subglacial fluid flow, and spring emergence.

Habitable subaerial environments exist over a range of climatic conditions. Soils form even under extremely cold climates on Earth (e.g., Mahaney et al., 2014), and springs fed by groundwater can exist in virtually any climate regime. Surface microbial communities like microbial mats and snow algae can persist below the freezing point of pure water (Hoham, 2000), and subglacial environments could provide some protection from even extreme climate change. Furthermore, subaerial environments can be extensive—soils, in particular, represent the most widespread nonmarine habitable environment on Earth. The different types of subaerial environments are also often found occurring in close proximity, as the precipitation that creates snowpacks, glaciers, and soils also feeds groundwater tables that can create wetlands and springs as they interact with topography and lithologic units. Thus, subaerial environments can contain a suite of habitats that foster diverse microbial communities.

3.3.2.2. Habitability cons

The main challenge for subaerial environments is that their habitability is often highly dependent on climate and can be adversely affected by changes in climate over time. On Mars, the habitability of near-surface environments would have been strongly affected by obliquity-driven climate change and, on longer timescales, the overall decrease in water activity at the surface (e.g., Ehlmann et al., 2011; Ehlmann, 2016). Thus, here we consider how habitability in subaerial environments would have changed on Mars over time.

Under the wetter climates of Noachian Mars, continuously or seasonally wet subaerial surface environments were probably widespread (e.g., Carter et al., 2015). These long-lived environments can be identified by the presence of clay-rich paleosols, and extensive or thick spring deposits (using methods discussed below), both of which strongly suggest long-term interactions with liquid water (Horgan, 2016).

The thinning of the martian atmosphere from the Hesperian into the Amazonian would have substantially decreased the diversity and productivity of subaerial environments over time (Westall et al., 2015a; Davila and Schulze-Makuch, 2016). As the climate cooled and the surface dried up, snowpacks and glaciers may have become more common, and subglacial environments in particular may have become refugia that offered some protection from increasing radiation and climate variability (Cousins, 2011). During the Hesperian, liquid water was largely driven underground or frozen at the surface but may have emerged as local springs or perhaps even regional playas during fluctuations in the groundwater table (e.g., Andrews-Hanna et al., 2007). Both wet subglacial environments and cold springs could have potentially existed on Mars late into the Hesperian (e.g., Fastook et al., 2012), so these environments may have represented niches for habitability after Mars lost much of its surface water.

Under modern martian climatic conditions, neither liquid water nor ice is stable at the surface over much of the planet, and even glaciers and ground ice do not show evidence for sufficient melting to produce liquid water (Fassett et al., 2010). Modern soils formed under recent climatic conditions on Mars (which are typically referred to as regoliths, as soil is strictly defined as including organic material) show weak evidence for chemical alteration (Yen et al., 2005; Amundson et al., 2008; Retallack, 2014) and are exposed to extreme radiation fluxes, low atmospheric pressures, and diurnal temperature cycles that would severely limit their habitability (e.g., Hassler et al., 2014). These soils could be habitable by some measures, as multiple observations have shown evidence for soil interactions with atmospheric water vapor (e.g., Martín-Torres et al., 2015), and energy would be available in the form of incipient chemical alteration, solar radiation, and higher-energy radiation. However, this environment would, under the best circumstances, support an ecosystem of very limited biomass and diversity that would be unlikely to be preserved as clear and concentrated biosignatures.

Spring activity is also limited on modern Mars. Recurring slope lineae identified at some equatorial sites have been proposed to indicate seasonal near-surface flows of brine, potentially sourced from near-surface aquifers or from deliquescence of salts by atmospheric water vapor (Ojha et al., 2015; Stillman et al., 2016; Wilson et al., 2016). However, the water activity for brines in these scenarios may be too low to support life (Rummel et al., 2014). Additionally, our knowledge of ancient spring deposits on Earth is somewhat limited, as spring deposits have only recently been recognized in the ancient rock record (Potter-McIntyre et al., 2016; Van Kranendonk et al., 2016).

3.3.3.1. Biosignature preservation pros

Subaerial environments can preserve a variety of organic, textural, chemical, mineralogical, and isotopic biosignatures, as reviewed by Horgan (2016). Archean and Proterozoic paleosols in terrestrial cratons preserve organic carbon in concentrations ranging from 0.01 to 0.36 wt % (Gay and Grandstaff, 1980; Watanabe et al., 2000, 2004; Rye and Holland, 2000). The organics are found both at the top of the paleosol profiles as well as in their subsurface horizons, and are attributed, respectively, to fragments of surface microbial mats and downward transport of surface organics during soil formation (Watanabe et al., 2004; Rye and Holland, 2000). Some of these occurrences are associated with anoxic surface ponds (Watanabe et al., 2004), and reducing conditions during paleosol formation appear to enhance organic preservation, as oxidized paleosols contemporaneous with these examples do not contain organics (Gay and Grandstaff, 1980).

In general, highly leached and oxidized paleosol profiles, including laterites, have poor organic preservation potential. However, these paleosols can preserve chemical biosignatures, including “bleached” Fe-poor upper horizons attributed to organic acids from surface microbial communities (Gutzmer and Beukes, 1998; Neaman et al., 2005), and phosphorus depletion in the upper portion of profiles that could indicate uptake by microorganisms (Horodyskyj et al., 2012).

A common characteristic of these paleosol biosignatures is that they are concentrated at or near the top of the soil profile; thus, a major requirement for biosignature preservation in paleosols is preservation of the paleosurface through rapid burial. Paleosol sequences, which form when weathering occurs along with recurring sedimentation, can preserve paleosurfaces, diverse geochemistries and redox states, and local surface aqueous environments, enabling multiple preservation mechanisms. While no paleosol sequences have been identified in the Archean and Proterozoic rock record (only individual paleosols as discussed above), studies of more recent paleosol sequences have shown concentrations of organics occur in subsurface sulfate nodules (Noe Dobrea et al., 2016) and as lignites in wetlands sediments (Retallack et al., 2000).

Biosignature preservation in cold springs is enhanced by the diversity of mineral deposits (e.g., carbonates, clays, sulfates/salts, amorphous minerals) that they produce, which enable multiple preservation mechanisms. Ostwald ripening (where large crystals form at the expense of smaller crystals, resulting in massive, anhedral crystal habits in Jurassic tufas) has been shown to preserve microbial body fossils and trace fossils in carbonates on geologic timescales (Potter-McIntyre et al., 2016). Additionally, energy availability produces abundant biomass in cold springs; therefore, biosignatures are highly concentrated in cold springs.

3.3.3.2. Biosignature preservation cons

A key challenge for biosignature preservation in paleosols in particular is that the local concentration of biomass can be quite variable, with relatively high concentrations only at the surface associated with microbial mats or perhaps in sediments produced in a highly productive persistent wetland environment. As such, even though many paleosols may have high chemical energy availability (e.g., subsurface redox gradients), this may not be consistently associated with significant biomass. Furthermore, it is unclear how widely distributed surface microbial communities would have been on either ancient Earth or Mars, especially prior to the evolution of photosynthesis. This dependence on surface communities also makes rapid burial key for preservation. While preservation in most other surface and aqueous environments is dependent on relatively rapid burial, the surface communities that subaerial environments depend on for habitation are especially sensitive to erosion, and rapid burial of these communities is a necessity for preservation. Lastly, the mineralogy and chemistry of many paleosols may not be sufficient for organic preservation. While high clay mineral (phyllosilicate) abundances in mature soils would help prevent later fluid flow and diagenesis, many clay-rich but oxidizing soils do not retain organics in the first place, as discussed above. Spring deposits would also require limited erosion and rapid burial in order to protect against erosion and destructive radiation. However, spring deposits preserve abundant textural and mineralogical biosignatures, so organics may not be necessary for biosignature detection.

3.4.2.1. Habitability pros

On Earth, the subsurface supports a significant amount of biomass, and the same may be true of Mars. In a paper that stimulated considerable interest and discussion, Whitman et al. (1998) even suggested that the standing biomass in the terrestrial subsurface rivaled that of surface environments. Although some more recent studies have yielded significantly smaller estimates for the subsurface biomass (Kallmeyer et al., 2012), estimates of the mass of the subsurface biosphere in continental crust remain in the 10¹⁶ to 10¹⁷ g C range (McMahon and Parnell, 2013), suggesting that subsurface conditions on Mars could similarly support a substantial biomass. It must be noted, however, that the bulk of the terrestrial biomass reflected in these estimates may rely, directly or indirectly, on chemical energy inputs from surficial photosynthetic communities. The amount of subsurface biomass on Earth that is supported by locally generated chemical energy sources alone remains poorly known but is likely to be considerably smaller than that of the surface-connected microbial community (Bach and Edwards, 2003; Edwards et al., 2012). Caves can also host extensive microbial biomass, which forms important micro- and macro-biosignatures, both structural and molecular (Frape and Fritz, 1987).

The martian subsurface likely provides physical environments capable of supporting life and protection from inhospitable surface conditions. Subsurface environments can provide liquid water at depth even when the surface is below freezing (Clifford et al., 2010; Michalski et al., 2013) and provide protection from the radiation and oxidants present at the surface. Liquid water may also be present in small amounts in otherwise frozen shallow deposits owing to the presence of salts that depress the freezing point; terrestrial microorganisms can metabolize in briny solutions to temperatures below −30°C (e.g., Bakermans and Skidmore, 2011). Caves act as insulating environments and may provide conditions (e.g., temperature and humidity) vastly different from external conditions (Azúa-Bustos et al., 2009). Caves may also host ice that could persist through changes in obliquity or be seasonally present due to periodic recharge from the near-surface/subsurface (Williams et al, 2010; Boston and Alexander, 2016). The ice-rock interface may have a persistent liquid water film that is sufficient to sustain habitable conditions (Popa et al., 2012)—cell concentrations as high as 10² to 10⁹ have been observed on fracture surfaces in ice (Onstott et al., 2016).

The subsurface provides long-term stable conditions that could allow life to adapt to the environment. Conditions in subsurface environments can remain stable for very long periods (millions of years or more), allowing microorganisms to adapt to the environment and for energy sources to accumulate. For example, recent studies of the terrestrial subsurface have identified deep fluids with residence times exceeding a billion years that have accumulated high H₂ concentrations (Holland et al., 2013). These fluids can support biological activity where they intersect with more shallow fluids (Lin et al., 2006; Sherwood Lollar et al., 2007). Caves are also long-lasting terrestrial environments (millions of years) and are likely to persist for even longer periods of time in tectonically quiescent environments such as on Mars.

Subsurface environments provide a variety of energy sources that can support biological activity. Water-rock interactions and radiolysis can provide abundant, localized energy sources to support subsurface communities. Much of the crust of Mars is composed of basalt and ultramafic rocks. On Earth, groundwater circulating through basalts in both subaerial and submarine settings is thought to support indigenous chemolithotrophic microbial communities relying on energy sources generated by reaction of water with basalt, such as iron oxidation, sulfide oxidation, sulfate reduction, hydrogen oxidation, and methanogenesis (e.g., Stevens and McKinley, 2000; Bach and Edwards, 2003; Lin et al., 2006; Lever et al., 2013; Osburn et al., 2014; Simkus et al., 2016). Reaction of ultramafic rocks with water, a process known as serpentinization, produces molecular hydrogen (H₂) and methane (CH₄) that support microbial activity (McCollom and Seewald, 2013; Schrenk et al., 2013). Utilization of many energy sources in these environments may require input of electron acceptors (O₂, NO^3-, SO₄ ^2-) from the surface, but plausible scenarios exist to transport such compounds into subsurface environments on Mars if they are present at the surface (Boston et al., 1992; Fisk and Giovannoni, 1999; Michalski et al., 2013). An additional source of metabolic energy in subsurface environments is radiolysis of water, whereby radiation from radioactive elements like K and U form molecules such as H₂ and hydrogen peroxide (Lin et al., 2005). Caves may also include gases from deep within the crust, allowing an additional energy source (Boston and Alexander, 2016).

Hydrology and enhanced fluid flow along fractures can focus energy and nutrient sources with the capacity for localized sites of heightened biological activity, facilitating targeted searches for habitable environments. Cell concentrations can reach as high as 10⁹ cells/g in the subsurface, and cells on fracture surfaces can reach 10⁵ cells/cm² (Wanger et al., 2006). Although concentrations on Earth tend to be highest for sedimentary rocks, cell concentrations as high as 10⁵ to 10⁸ have been observed for ash tuffs and metamorphic rocks on Earth (Onstott et al., 2016). Caves can provide preferential groundwater conduits and reservoirs for nutrients and energy captured from surface or subsurface weathering (Léveillé and Datta, 2010).

Caves, in addition to providing many of the benefits listed above such as long-term environmental stability, protection from inclement surface conditions, and a variety of localized chemical gradients, also provide the benefit of an intermediate environment between the surface and subsurface (Boston and Alexander, 2016). Caves are likely to provide higher levels of humidity, as well as access to atmospheric gases within the subsurface, which could provide redox gradients that could be utilized for metabolic energy (Boston and Alexander, 2016).

3.4.2.2. Habitability cons

On Earth, cell density gradually decreases with depth, suggesting that biomass in deep subsurface environments on Mars may also be decreased relative to what we observe close to the surface on Earth. The habitability of many subsurface environments may also require communication with the surface to supply critical nutrients (e.g., electron acceptors, elemental nutrients, organic matter) and remove waste products. In the absence of such communication, habitable environments may be much more limited. At present, little information is available concerning the magnitude of surface inputs to subsurface environments on Mars. Environments such as caves that are more closely connected to the surface may also be more vulnerable to threats to habitability such as cold temperatures.

Habitable environments in the subsurface will also be spatially limited owing to restrictions on distribution of pore spaces (fractures), temperatures above the upper limit for life (currently 122°C for life on Earth; Takai et al., 2008), and presence of liquid water.

3.4.3.1. Biosignature preservation pros

There are many different types of biosignatures that may potentially be produced by microorganisms in subsurface environments, including biominerals, textures, microbial body fossils, macroscopic biopatterns, organic and inorganic chemical compounds, and isotopes. Potential biosignatures include minerals precipitated as a direct result of biological activity (e.g., Fe oxides, sulfides), preservation of microbial structures such as biofilms by precipitating minerals such as carbonates filling fractures (Pedersen et al., 1997), organic compounds produced by biological activity (e.g., lipids, amino acids), textures indicative of microbial activity (e.g., microtubules; Fisk et al., 1998; Furnes et al., 2004), changes in trace element distributions (Leslie et al., 2013, 2014), and isotopic composition of organic matter, the breakdown products of organic matter, or minerals (e.g., Alt and Shanks, 1998). In caves and lava tubes, various mineral textures, microstromatolitic textures, and isotopic signatures in minerals and organic matter have been identified as biosignatures (Boston et al., 2001; Léveillé et al., 2007).

The interaction of fluid and rock in the subsurface leads to mineral alteration and widespread precipitation of secondary minerals, and precipitation of these secondary minerals creates a favorable environment for preservation of biosignatures. For example, microbes may induce precipitation of minerals that possess structural characteristics indicative of a biological role in their formation (e.g., Cady and Farmer, 1996; Farmer and Des Marais 1999; Konhauser et al., 2002). Mineral precipitation may also entrap organic compounds or retain isotopic signals of biological activity (Boston et al., 2001). Preservation of biosignatures also actively occurs in terrestrial caves by mineral precipitation (Léveillé et al., 2007; Richardson et al., 2012; Boston and Alexander, 2016); these authigenic minerals will also possess a record of environmental conditions and water-rock interactions that would provide additional context to biosignature formation and preservation.

Owing to interactions of fluid and rock, conditions in the subsurface will tend to be reducing, with low levels of oxidants. In the absence of these compounds, oxidative decomposition of organic compounds may be restricted. Access to the subsurface through caves, impact excavations, and potentially recurring slope lineae means that potential biosignatures preserved beneath the surface may also be exposed at the surface for detection. Lava tubes also often form relatively near the surface, thereby facilitating possible exposure (Boston and Alexander, 2016).

3.4.3.2. Biosignature preservation cons

Subsurface environments on Earth have only recently begun to be the focus of intensive scientific attention; consequently, the types of biosignatures that subsurface communities might create remain largely unknown. Most work to date has focused on characterizing the genetic and metabolic diversity of subsurface communities (Northup et al., 2011; Popa et al., 2012; Lever et al., 2013; Osburn et al., 2014) and on characterizing the distribution of bioorganic compounds such as lipids and amino acids (Lipp et al., 2008; Lomstein et al., 2012). However, how biosignatures of subsurface communities might be preserved over time remains unknown. Although it has been suggested that microtubules observed in some subsurface basalts might be a biosignature (Fisk et al., 1998; Staudigel et al., 2008), the biological origin of these structures remains unproven, and alternative explanations have been proposed (French and Blake, 2016). Potential biosignatures that could persist for millions to billions of years in the subsurface have not yet been identified.

3.5.2.1. Habitability pros

Iron is one of the most abundant redox-active metals in Earth's crust, and the oxidation of dissolved and solid-phase Fe(II) can power microbial metabolisms in both acidic and circumneutral settings and under aerobic and anaerobic conditions (e.g., Winogradsky, 1888; Ingledew, 1982; Widdel et al., 1993; Straub et al., 1996). Chemotrophic iron oxidizers can exploit the redox gradient of Fe(II) and O₂ at oxic/anoxic interfaces to fuel their growth (e.g., Emerson and Revsbech, 1994; Baker and Banfield, 2003). Some of these chemotrophs can grow anaerobically using nitrate as an oxidant instead of O₂ (Straub et al., 1996). If other reductants such as H₂ or organic carbon are present, chemotrophs can also use Fe(III) as an oxidant, a process termed dissimilatory iron reduction (Lovely, 1993). Growth on Fe(II) is not restricted solely to chemotrophs, as anoxygenic phototrophs can use Fe(II) as an electron donor for photosynthesis (Widdel et al., 1993; Heising and Schink, 1998; Heising et al., 1999).

Many iron-rich settings may represent long-lived habitable environments. The large volume of certain iron deposits can reflect a relatively steady supply of energy and reduced iron inputs over geologic timescales. For example, most Precambrian IFs were deposited from 3.8 to 1.8 Ga (Klein, 2005), potentially representing a continuously habitable environment on the scale of ca. 2 billion years. Other large iron-bearing deposits such as massive sulfide deposits may be long-lived, on the order of thousands to hundreds of millions of years, based on the time required for such a large volume to be weathered or altered such that all the iron is oxidized and no longer able to support microbial growth (Albers and Bain, 1985; Fernández-Remolar et al., 2003). The longevity of subsurface iron-rich habitats is unknown, but modeling of the dissolution of iron minerals and transport of Fe(II)-rich groundwater indicates that formation of the concretions may have been episodic over geologic time (Chan et al., 2007).

Other characteristics of iron can enhance the habitability of surface environments. The anoxic Archean Earth experienced a much higher UV flux due to the absence of an ozone layer (Sagan, 1973; Margulis et al., 1976). Both iron and elemental sulfur have been shown to provide some shielding against UV radiation, which may delay or prevent the breakdown of organic matter due to radiolysis (Cockell, 1998; Gomez et al., 2003). Iron can also protect microbial cells from UV-induced DNA mutations and lethal cell damage (Olson and Pierson, 1986; Pierson et al. 1993). The absorption coefficient for ferric (Fe(III)) iron in particular is an order of magnitude greater than that of ferrous (Fe(II)) iron (Cockell and Knowland, 1999). Whether the oxidized ferric iron was produced biotically or abiotically, it would confer greater microbial survival in surface environments on early Earth and potentially Mars.

3.5.2.2. Habitability cons

The primary total organic carbon (TOC) input into iron systems is much lower than in other environments (e.g., marine systems) due to the energy requirements of microorganisms living in iron-bearing habitats, which are strongly to mildly acidic (Parenteau and Cady, 2010; Graham et al., 2016; Johnson et al., 2016). Microorganisms must invest a significant amount of energy to pump protons out of their cells to maintain a neutral intracellular pH, and consequently do not produce as much biomass as microorganisms growing in more benign environments. Additionally, acidity can bottleneck microbial diversity in an environment, as microorganisms must evolve to cope with the elevated acidity. In isolated environments with limited TOC sources to fuel heterotrophic growth, microbial populations consequently synthesize limited biomass (Tyson et al., 2004). However, additional allochthonous TOC input from eukaryotes (e.g., plants) in a highly acidic environment may generate greater microbial diversity, such as in the Río Tinto system (Amaral Zettler et al., 2002).

3.5.3.1. Biosignature preservation pros

Preservation of microorganisms occurs when cells are rapidly buried in fine-grained iron sediments, or rapidly encrusted or permineralized in fine-grained primary iron precipitates (Farmer and Des Marais, 1999). In these environments, anoxic conditions may be maintained that restrict the oxidative destruction of organic compounds and permit early diagenetic mineralization (Cady et al., 2003).

When microbial cells are rapidly encrusted in iron-rich minerals, they may form textural or morphologic biosignatures, which may be long-lived in select environments even after diagenetic alteration of the iron oxyhydroxides. This has been documented in gossans thousands to several millions of years old (Fernández-Remolar and Knoll, 2008; Williams et al., 2015, 2016) and in iron oxyhydroxides in lacustrine units from the Jurassic (155–148 Ma) (Potter-McIntyre et al., 2014, 2016). On Earth, iron-oxidizing chemotrophs often deposit distinctive mineral structures (stalks, etc.) that can be intimately associated with organic compounds (e.g., Edwards et al., 2004; Chan et al., 2011). These structures can be preserved in the geologic record (e.g., Alt, 1988).

In addition to textural or morphological biosignatures, the cellular remains of microorganisms can be preserved as lipid biomarkers in iron-rich deposits. The carbon skeletons of branched or cyclic lipids are more resistant to degradation after cell death and can survive in the rock record for billions of years (Waldbauer et al., 2009). Such “geolipids” provide important fossil information and can help reconstruct the record of past microbial ecosystems on early Earth and potentially Mars. Iron has been shown to enhance the early preservation of lipids by rapidly encrusting the cell and protecting the lipids by making them unavailable to enzymatic degradation by the cell's own lytic enzymes, or those of heterotrophs (Parenteau et al., 2014, 2016).

Even though microfossils have not been found in the iron-rich layers of IFs, putative cells rimmed and fully replaced with hematite have been reported in the chert layers of the Gunflint Iron Formation (Barghoorn and Tyler, 1965). Allen et al. (2001) also reported iron permineralized structures resembling cells in chert layers from the Mink Mountain locality of the Gunflint. In addition to the microfossils, iron mineralized polymeric substances were detected, which prompted the authors to suggest that microbial mats may have played a role in iron oxide precipitation and IF formation (Allen et al., 2001; Schelble et al., 2004). Such reports lend support to the hypothesis that iron may play a role in the long-term preservation of cells.

Although organic carbon is thermodynamically unstable in the presence of Fe(III), organic matter may coprecipitate or chelate in iron complexes or be deposited in interbedded evaporites or clays to be preserved over longer timescales (Lalonde et al., 2012). While organic matter is often found in very limited concentrations within the iron oxide-facies of IFs (Kaufman et al., 1990), this may reflect a lower initial TOC input rather than poor preservation by iron (Parenteau and Cady, 2010; Parenteau et al., 2016).

3.5.3.2. Biosignature preservation cons

The preservation potential of microbial biosignatures in iron-rich environments is diminished by (1) the degradation of organic matter in the presence of Fe(III) over geologic timescales or at elevated temperatures, (2) biosignature formation over short timescales, which then must be resistant over long timescales to be preserved, (3) ionizing and potentially UV radiation, which may lead to organic matter degradation, and (4) the heterogeneous distribution of iron biosignatures.

As noted above, the preservation of organic carbon by iron oxides is not thermodynamically favored in the presence of Fe(III), and organic carbon can decay quickly in oxidizing environments if not captured in an environment that exhibits preferential preservation (Cady and Farmer, 1996; Walter et al., 1996, 1998; Farmer, 1999; Sumner, 2004; Hofmann et al., 2008). For example, in the same near-surface (upper 5 m) environment, a phyllosilicate-rich layer may contain 10 times more TOC than the surrounding iron-rich layers, and iron-rich hyperacidic environments demonstrate poor preservation of dissolved organic matter (Bonaccorsi and Stoker 2008; Davila et al., 2008).

Ferric iron (Fe(III)) has been shown to confer protection to microorganisms and enhance survival in high UV environments on early Earth (see Section 2.1). However, UV radiation may conversely lead to poor biosignature preservation. This is due to the ability of radiation to drive the oxidation of Fe(II) to Fe(III) in laboratory experiments (Cairns-Smith, 1978; Braterman et al., 1983). This change in oxidation state can destroy biosignatures that formed in a previously reduced or mixed valence system. Textural and morphological biosignatures can also be destroyed when the metastable primary iron oxyhydroxides recrystallize to more thermodynamic stable forms (Parenteau and Cady, 2010). An iron-mineral environment may be less able to protect organics from ionizing radiation.

Biosignatures can also be heterogeneously distributed in subsurface, subaqueous, and subaerial iron habitats. Microbial communities can be patchy in their distribution, and their concentration can be highly dependent on microenvironmental niches (e.g., Pierson et al., 1999) or small point sources of Fe(II) (Emerson and Moyer, 2002).

4.1.1.1. Orbital observations for site selection

The first challenge in identifying an appropriate site—regardless of the environment that is being targeted—is the limit in resolution of our orbital assets. For example, the High Resolution Imaging Science Experiment (HiRISE) on the Mars Reconnaissance Orbiter is capable of sub-meter-per-pixel resolution, and the Compact Reconnaissance Imaging Spectrometer for Mars (CRISM) is capable of ∼18 m per pixel resolution. Although this is certainly sufficient resolution to identify some of the larger-scale environmental features (e.g., deltas, meter-scale outcrop features), these data sets can provide only broad context for identification of most types of environments that are under consideration here. This potentially introduces a bias toward prioritizing environments that are more easily identified remotely, including environments whose geologic record is better exposed through erosion and dust removal, and environments that have larger areal footprints, regardless of their potential for biosignature preservation.

Additionally, the orbital reconnaissance of potential sites for exploration cannot provide a reliable indication of how long any given environment may have been habitable. Whether a particular environment was persistently habitable and its record covers an extended period of time (hundreds to thousands of years or more) or whether a particular deposit accumulated relatively quickly (years to hundreds of years) would have a significant effect on the potential for biosignatures to have been created and preserved.

Although some aspects of accessibility can be remotely assessed for a particular site before selection (e.g., average slopes or rough terrain), the specific challenges that are unique to any given site may not be obvious until the surface mission is in place. As the Mars Exploration and Mars Science Laboratory rovers have traversed the surface, there have been significant improvements in correlating ground and remote observations, but how the accessibility of a particular site may affect the ability of a mission to identify and collect the appropriate samples for detection of potential biosignatures cannot be fully known in advance. This challenge would have a greater effect on environments where full characterization would likely require sampling across significant vertical or horizontal extent and thus where accessibility of the entire appropriate facies would be important.

4.1.1.2. Surface observations

Even after a particular site has been selected for the next mission, surface operations face the following challenges to proper sample selection; these apply to all the environments being considered in this report.

The first challenge would come from recognizing the particular features of a habitable environment in the absence of abundant life (or potentially life at all). Many of the features that we use to identify different facies are dependent on both biotic and abiotic processes. Additionally, the abundance of photosynthetic microorganisms on this planet certainly has an effect on the particular features that are used to identify a preserved inhabited environment, and it must be considered which features on the outcrop scale should be targeted for sampling given the possibility that photosynthesis may have never evolved on Mars. For example, phototrophs in hydrothermal spring systems synthesize more biomass than chemotrophs and produce a wider variety and abundance of preserved organic molecules and meso- and microscale textures. These textures may be detectable on the outcrop scale, aiding in the identification of past habitable environments for further sampling. In contrast, there are fewer types and less abundant preserved textures associated with chemotrophic communities, possibly making their detection at the outcrop scale above the background geologic “noise” more difficult. Studies of biosignature production and preservation in analog environments are key for characterizing signatures associated with chemotrophic communities should photosynthesis have never evolved on Mars. This challenge has a particularly strong effect on the identification of mineralogical, chemical, and isotopic biosignatures, where relatively few studies have explored the nonbiological processes that could closely mimic biotic features if given longer timescales to do so.

In all the environments under consideration here, a major challenge for selecting the best sampling location is the spatial heterogeneity of the most promising deposits, which includes local, chemical, or sedimentological variability. The initial concentration of microorganisms can be highly dependent on microenvironmental niches (e.g., Pierson et al., 1999), or small point sources of nutrients may dramatically affect the density of biosignatures in a given environment (Emerson and Moyer, 2002). Within lacustrine environments, a particular section of mudstone may have significantly more organic material than another just a few meters away, or within a hydrothermal spring environment there may be a particular section where the microfossils are more abundant than in adjacent sections (e.g., Campbell et al., 2015, 2016; Van Kranendonk et al., 2016). While this challenge may not be as much of a problem for macroscopic biosignatures, it is the reason why multiple field seasons interspersed with lab work are typically required in order to achieve a complete environmental characterization and sampling campaign in terrestrial environments. This challenge can be mitigated somewhat due to the advanced instrumentation on the Mars 2020 rover that is typically not available during terrestrial field campaigns. However, as noted here, these instruments might not be equally effective in all environments.

4.1.1.3. Sample collection

One of the final common challenges—and one that in terrestrial environments can be a mitigation strategy for spatial heterogeneity—is the number and size of samples collected. While overcollecting is a way to obtain the maximum number of samples with detectable biosignatures on Earth, there is a limit to the number and volume of samples that will be collected (e.g., as determined by the capabilities of the Mars 2020 rover sampling system) and could be returned (determined by the size of the potential sample return missions). Currently, the Mars 2020 rover sampling system is designed to collect between 35 and 40 ∼15 g samples of rock and regolith (Farley, personal communication). This extremely limited amount of sample mass—some of which would be committed to other science objectives and planetary protection assays—will be a challenge for detecting all types of biosignatures and in all potential environments.

4.1.1.4. Mars 2020 rover instrument suite

Given that the instrumentation for the Mars 2020 rover has already been selected, a challenge in identifying the correct samples to collect will certainly come from the limits of the Mars 2020 instrument suite. Though the in situ science capabilities of the Mars 2020 rover have a high potential to identify features consistent with habitable environments and life, certain instruments may be more or less appropriate for identifying sampling locations in different environments (e.g., the fluorescence of hematite in Raman imaging can overwhelm the organic carbon signal at certain wavelengths) (Emry et al., 2015). This challenge is dependent on both the environment and the biosignatures that are being identified. Therefore, analog studies in a variety of relevant analog environments will be important in the coming years.

4.3.1.1. Observations

The former presence of hydrothermal activity can be indicated by the alteration zones along former fractures, minerals that frequently accompany hydrothermal activity, and geomorphic features related to spring mounds and other related sedimentary deposits (e.g., Ruff et al., 2011). Some examples of hydrothermal minerals (among many) include opaline silica, carbonates, phyllosilicates, and sulfides.

4.3.1.2. Sample selection

Ideally an important suite of samples from a hydrothermal system would be collected along the former thermal gradient from the high-temperature source to the distal deposit emplaced under surface ambient conditions. Geochemical and mineralogical analyses of such a sample suite could reconstruct the conditions during the system's active phase. Any biosignatures in this sample suite could potentially indicate the diversity of life-forms that had adapted to the range of conditions along this environmental gradient. In a subsurface hydrothermal system, samples of fracture fills of varying mineralogy, as well as samples of unaltered and altered parent rock, would also be important.

4.3.2.1. Observations

On the ground, deltaic deposits are best identified by lithology, texture, and bed geometry. The distinct patterns of distributary channels and discharge lobes can be distinguished from alluvial fans, as was found at Gale Crater (Mangold and Ansan, 2004; Wood, 2006). In the field the coarsening-upward sequences of the prodelta can be observed—possibly even from orbit (Pondrelli et al., 2008; Schon et al., 2012). Well-preserved deltas will include the continuous vertical transition from a full subaqueous facies to subaerial features such as soil development or dune deposition, although erosion of any of these features can occur after deposition. These features, as well as outflow channels and prodelta deposits, can be seen from orbit. However, the underlying lacustrine deposit is most promising for preserving biosignatures, yet it is difficult to identify from orbit.

One challenge of deltaic deposits, even on the ground, is accurately constraining their age and duration. In some cases, the age of the deltaic deposits can be determined from orbit by relating the stratigraphy to the cratering record, but the small exposure area of deltas makes this comparison challenging (Warner et al., 2015). Timescales for sediment emplacement could potentially be estimated by applying sediment transport models. Channel width and meander wavelength can be used to calculate flow magnitudes that, when compared with expectations of precipitation, can indicate time of deposition, although the range is often quite large. For example, some estimates indicate that deltaic deposits at Holden Crater would have taken between thousands and millions of years to accumulate (Moore et al., 2003). Physical models such as these are imperfect for application to Mars given that the effects of reduced gravity on sedimentary processes are not well understood. Exposure ages are also important for estimating the radiation damage that deltaic sediment may have experienced. Such ages can be estimated from orbit based on crater densities and evidence for recent erosion; however, a better estimate would use the exposure age of rocks based on cosmogenic nuclides, a measurement that must be made in situ (Farley et al., 2013) or with returned samples.

A rover could identify horizontal planar stratification with interbedding of fine-grained siliciclastics and salts, including carbonates and/or sulfates. Thick sulfate deposits that grade laterally into shoreline sands (identified by ripple marks and basinward dipping strata), fluvial channels (from episodic flooding into a desiccated lake) are also identifying characteristics. Other sedimentary structures that could be present are mudcracks and paleosols. The best samples for detection of biosignatures in this environment would be wavy, laminated morphologies indicative of biofilm fabrics, evaporites that could host fluid inclusions or even endoliths, and other microbial morphologies such as desiccated flexible microbial mats that have become curled.

The SHERLOC and SuperCam instruments on Mars 2020 might detect organic biosignatures preserved in deltaic systems. The Mars Science Laboratory mission has characterized the sedimentology of fluvial-lacustrine systems in detail at the rover scale. The lack of a quantitative mineralogy instrument on Mars 2020 will prevent any detailed analysis like that of Mars Science Laboratory, but the SuperCam VNIR spectrometer should be able to remotely identify clay minerals like smectites that are critically important for organic preservation in these environments. The fluorescence spectrometers on SHERLOC can detect condensed carbon and aromatic organics by deep UV-induced fluorescence, and SHERLOC's Raman spectrometer will allow classification of aromatic and aliphatic organics. Raman spectrometry can also be used to detect minerals relevant to aqueous chemistry. While these measurements would allow us to identify reduced carbon compounds, there may not be sufficient structural information to distinguish between a biological signal and extraterrestrial organic input.

4.3.2.2. Sample selection

While the Mars 2020 instrument suite may be limited in its ability to confirm the biological origin of subaqueous organic matter, there are other targets of astrobiological interest from these environments. Morphological targets that are common to these environments and are well preserved regardless of exposure age can include biofilm fabrics, evaporites, and microbial morphologies.

4.3.3.1. Observations

At outcrop scales, paleosols can be identified by the presence of mineral, textural, or color horizons, pedogenic morphologies (centimeter-scale granular, platy, or prismatic structures), chemical profiles indicating top-down leaching, and the association of these features with paleosurfaces or unconformities (Retallack et al., 1999; Sheldon and Tabor, 2009). Further characterization of paleosols can be accomplished through measurements of isotopes and elemental chemistry to determine paleoclimate during formation, as well as chemical and mineralogical analyses to determine the redox state and chemistry of leaching fluids. Reducing soils and wetlands that are good targets for enhanced biosignature preservation would be indicated by concentrations of ferrous iron in clays, sulfides, or amorphous phases. The instrument suite on Mars 2020 would be well suited to identify and investigate paleosols on Mars: the imaging systems could be used to identify paleosurfaces and surface aqueous features, and pedogenic textures; Mastcam-Z multispectral imaging could be used to identify the distribution of ferric and ferrous minerals and thus redox gradients; SuperCam could measure elemental profiles; SHERLOC and SuperCam Raman and SWIR spectrometers could be used to identify clay and sulfate minerals; and SHERLOC could be used to search for organics at or near paleosurfaces.

Glacial environments are most easily identified when glacial morphologies are exposed at the surface; however, this requires long-term exposure that would adversely affect preservation. How glacial deposits like subglacial alteration rinds, glacial flour, and so on would manifest or be identified in the geologic record is unclear; thus, the biosignature preservation potential of glacial sediments or environments is not well understood. However, glacial sediments could also be preserved in downstream lakes or alluvial fans.

Cold spring deposits could be characterized from a rover with relative ease due to the abundant macroscopic textural and mineralogical diagnostic characteristics, including terraces and terracettes (centimeter-scale terraces), wavy laminae, and alternating, layered mineralogies [e.g., calcium carbonate and iron (oxyhydr)oxides] and a variety of facies such as vent, pond, channel, and proximal and distal facies (described in Fouke et al., 2000).

4.3.3.2. Sample selection

The main samples of interest from many of these subaerial environments would be mineral deposits precipitated in surface or subsurface redox gradients produced by groundwater fluctuation at the surface, water flow along subsurface barriers, or at springs—in particular, sulfates, and iron oxides. Ideally, redox gradients should be sampled. Depending on the mineralogy, these samples could contain organics, mineralogical, chemical, and isotopic biosignatures. In addition, other mineral deposits produced at springs, such as carbonates and silica, would be key astrobiological samples, as they could contain organics, microfossils, textural biosignatures, mineralogical, chemical, and isotopic biosignatures. Samples of sediments accumulated at paleosurfaces, as indicated by unconformities and tops of chemical profiles, would also be desirable, as these are the most likely place to find biosignatures produced by microbial mats and could contain organics, chemical, mineralogical, or isotopic biosignatures. Within this category, reduced clay-rich soils—ideally wetlands-type saturated soils—would be of the highest priority. Finally, samples representative of chemical profiles with depth in soils would also be useful to better understand the climate, environment, and chemistry of the subaerial habitat, as well as to search for chemical/isotopic biosignatures. In particular, sampling parent material at depth in the soil profile would be key for establishing a baseline for mineralogical/chemical/isotopic biosignatures.

4.3.4.1. Observations

Far-range cameras such as Mastcam-Z will be able to identify areas that are promising to investigate, such as where subsurface material has been exposed in crater walls or the ejecta blanket and can be further examined. Such cameras would also identify the mouths of caves, although those would be challenging to access and explore for a Curiosity-style rover.

Remote compositional instruments such as Mastcam-Z, ChemCam, and SuperCam would also be able to help identify signatures of deep habitable environments such as minerals potentially indicative of redox disequilibrium (e.g., Fe-containing minerals) or subsurface alteration. These instruments would also be able to identify evidence of the type of focused fluid flow that might provide a more heavily inhabited environment, such as a fracture or vein. They might also be able to detect evidence of discharge of groundwater that might provide biosignatures from the subsurface. Laser-based instruments (LIDAR, LIBS, LIF, Raman spectroscopy), in particular, would be effective in subsurface cave environments without the need for constant illumination. Contact chemical and mineralogy measurements would be able to indicate potentially habitable conditions such as mineralogy indicative of subsurface fluid-rock alteration, potential biosignatures such as evidence of redox disequilibrium, and elemental chemistry that indicate chemical changes such as leaching consistent with interaction with biota and biotic organic compounds (e.g., Hausrath et al., 2008, 2009, 2011; Liermann et al., 2007; Baumeister et al., 2015). Measurements of organic compounds such as by the SAM instrument or Raman measurements would be able to provide direct evidence of potential organic biosignatures such as lipids, amino acids, DNA, and so on.

4.3.4.2. Sample selection

The top samples to collect from subsurface environments would include samples likely to capture potentially habitable environments, such as aqueously altered samples from caves, impacts, megabreccias, exhumed materials, deep aquifers, and chemical and redox gradients.

4.3.5.1. Observations

Subsurface iron-rich habitats can be recognized by identifying massive red-orange iron oxides or extensive iron-oxide cementation within permeable rock (e.g., sandstone). These environments may contain macroscale iron concretions within the outcrop or weathered out and resting on the surface of the ground. Zoning may be visible on the outcrop face, indicative of diffusion and alteration within the outcrop. Subaqueous IF-like deposits would appear as thickly to thinly laminated bands of iron and silica, or a granular texture with no apparent banding, with interbedded carbonates and shales. These deposits can be thick and laterally extensive. Subaqueous hydrothermal vent deposits may be visible as “thin skins” of iron (oxyhdr)oxides on top of reduced iron-bearing country rock. However, this cap of Fe(III) may readily weather away. Subaerial iron seeps and springs would be recognizable as mounds of flocculent, ochreous iron oxyhydroxides (e.g., ferrihydrite), or denser iron oxides (e.g., hematite or goethite) on top of compositionally different host rock (likely igneous rocks underlying hydrothermal springs). Lastly, gossans or massive sulfide deposits would be visible as extensive red outcrops of iron oxides and may exhibit a spotty presence of residual pyrite/sulfide minerals or efflorescent iron sulfate salts.

Instrument packages for a rover would be well-suited to analyze iron-rich environments. In terms of imaging, a reconnaissance (long distance) camera would see the characteristic red color of iron oxides in all of these iron-rich environments, even on the generally red surface of Mars. The iron oxides may appear more massive, laminated, or concentrated at the formation than in surrounding substrate. Topographically, these deposits may form capping or resistant units that retain higher elevations than the surrounding landscape. A high-resolution (“hand lens”) camera may see carbonates, sulfides, phyllosilicate layers, efflorescent sulfate salts, and thinly laminated iron oxides and silica and/or a variety of iron oxide textures and biofabrics (as described above).

In terms of composition, a long-distance mineralogy instrument may detect iron oxides, oxyhydroxides, and oxyhydroxysulfates (hematite, goethite, jarosite), iron sulfides (pyrite), sulfates (gypsum and a variety of ferric and ferrous sulfate salts), carbonates, and possibly quartz. In addition to the major phases described above, a high-resolution mineralogy instrument may detect minor phases: iron oxides, oxyhydroxides, and oxyhydroxysulfates (magnetite, ferrihydrite, schwertmannite), iron sulfides (pyrrhotite, chalcopyrite, sphalerite, etc.), carbonates (siderite, ankerite, etc), sericite, illite, other phyllosilicates, barite, and amorphous silica, among other minor phases. It would be critical to map these phases in relation to textures and fabrics possibly detectable with the “hand lens” camera. An instrument capable of measuring elemental abundances would see Fe, S, and Si. Lower-abundance elements would include metals such as Cu, Zn, Ag, and Au, cations associated with phyllosilicates (e.g., Al, K, Mg), and elements from carbonates such as Ca and Mg. Lastly, an organic molecule detection instrument may detect aromatic and short- and long-chain aliphatic molecules if they are (1) relatively young or (2) preferentially preserved in subenvironments within the iron oxides.

These general rover strategies translate well to the Mars 2020 rover payload. Both Mastcam-Z and SuperCam can determine mineralogical composition, while SuperCam can also determine elemental composition, and PIXL can map fine-scale elemental compositions. SHERLOC can characterize aliphatics and aromatics in association with millimeter- to centimeter-scale structures/biofabrics, and the associated instrument WATSON can image macroscale biogenic textures. Potential limitations include the ability of SHERLOC to resolve organics on mixed mineralogy iron or silica matrices (Emry et al., 2015), and the potential transformation of metastable iron phases by higher power lasers (Hanesch, 2009).

4.3.5.2. Sample selection

To best capture the diversity within an iron-rich environment that may contain preserved biosignatures, we propose sampling the following zones for potential sample return: (1) perform a survey and sample along any redox gradient (e.g., within an aquifer, along a gradient of oxidative products), (2) after identifying a putative cycle of deposition, sample both the base and top of the cycle, and (3) sample diversity of mineralogical types and macroscale textures along the lateral extent of the outcrop exposure.