Two of the best studied habitable environments on Mars are within Gale and Jezero craters (Figure 2). Gale crater is a Late Noachian- to Early Hesperian-aged crater that is ∼150 km in diameter (Thomson et al., 2011). It is located along the crustal dichotomy, a topographic boundary that separates the heavily cratered southern highlands of Mars and the sparsely cratered northern lowlands. The crater hosts a 5-km-high sedimentary mound, Aeolis Mons (i.e., Mt. Sharp), which is the Curiosity mission’s primary science target (Grotzinger et al., 2012). This is due to the fact that lower strata of Mt. Sharp contains hematitic and hydrated clay- and sulfate-bearing rock, which were detected from orbit (Milliken and Bish, 2010; Fraeman et al., 2013), but are accessible on the ground by the Curiosity rover. Investigations by the rover, as well as from satellite imagery, have determined that Gale crater hosted a series of large lakes, both before (Grotzinger et al., 2015) and after (e.g., Palucis et al., 2016) the emplacement of Mt. Sharp. These lakes may have been as large as ∼5,800 km² in area and up to ∼0.7 km deep (Palucis et al., 2016). The sources of water to the lakes were likely from a combination of groundwater and surface runoff, the latter potentially resulting from snowmelt (Grotzinger et al., 2015; Palucis et al., 2016; Roseborough et al., 2021). The climate during the time of lakes is debated, with modeling and geomorphologic arguments suggesting a relatively arid- to semi-arid climate with temperatures above freezing (Horvath and Andrews-Hanna, 2021), whereas the mineralogy and geochemistry support more dynamic climates with periods of prolonged warm and semi-arid conditions on a mostly cold/snowball Mars (Fukushi et al., 2019). Fukushi et al. (2019) estimated that early lakes in Gale developed in ∼10⁴ to ∼10⁶ year-long semiarid climates based on pore water chemistry from the last wetting event near the Curiosity landing site (i.e., Yellowknife Bay), which is based on finding water with neutral pH, mild salinity, low levels of dissolved CO₂, and highly oxidizing conditions (E _h > 300 mV, where E _h is redox potential).

Jezero crater is a ∼45 km diameter crater located in the Nili Fossae region of Mars. It was selected as the landing site of the Perseverance rover due to the presence of two sedimentary fan-like structures inferred to be deltas from orbit (e.g., Fassett and Head, 2008b; Goudge et al., 2015) which was later confirmed with sedimentological ground observations (Mangold et al., 2021). These deposits and their mineralogy indicate an ancient habitable lake (with similar water conditions to Gale) that formed during the Late Noachian to the Early Amazonian (Ehlmann et al., 2008; Goudge et al., 2015; Mangold et al., 2020). Jezero is an ideal location to search for biosignatures on Mars due to its geologic context (an ancient lake), its mineral diversity from orbit (including hydrated magnesium carbonates and olivine, clay minerals, and silicas), and its potential to preserve biosignatures based on these specific mineral deposits (Grant et al., 2018). There are also potential hydrothermal deposits within Jezero, as well as aqueously altered mafic and ultramafic rock, allowing for the search for traces of prebiotic chemical processes (Bosak et al., 2021).

Based on these two landing sites, terrestrial analogs would ideally meet several specific Mars-like conditions. Both the Gale and Jezero lakes are thought to have been active during the Late Noachian through the end of the Hesperian. There is debate about the range of temperatures on the Martian surface at this time, but there is no clear geomorphic or sedimentologic evidence that the lakes were ice-covered (e.g., Palucis et al., 2016), such that average temperatures may have been above freezing when lakes were present. Work by Fukushi et al. (2019) suggests redox conditions at Gale supported distinct periods of warming on an otherwise cold (∼0°C) Mars. Given these constraints, terrestrial lake systems that span mean annual temperatures at and above 0°C are desirable targets. Both the Gale and Jezero lakes appear to have been fed from a combination of groundwater and surface water, which is the case for many terrestrial lake systems, though some studies suggest that the runoff was sourced from snowmelt (versus direct precipitation) (Grotzinger et al., 2015; Palucis et al., 2016; Roseborough et al., 2021). Mars is a mostly basaltic planet (Tanaka et al., 2014), so ideal Earth analogs would be hosted in basaltic terrains, with mafic to ultramafic rocks that lack extensive metamorphism or tectonics. Ideal analogs would also have similar mineralogies to those found at Gale and Jezero, namely the presence of magnesium carbonates, Fe/Mg clay minerals, amorphous silica, hematite, and sulfates (Vaniman et al., 2014; Goudge et al., 2015; Rampe et al., 2017). In order to document how polyextremophile microorganisms respond or vary as a function of water activity and chemistry, analog systems should also have well-documented water salinities, pH, and redox potential, in addition to well-characterized microbial populations, and process and/or metabolism specific biosignatures. Lastly, as it is thought that Martian lake systems went through periods of wetting and drying (Stein et al., 2018), knowing the hydrologic history of each analog site is important, specifically periods of lake desiccation due to climate change or glaciation and changes in water sources or chemistries over time.

Located on Devon Island in the Canadian Arctic, the Haughton Impact Crater (HIC) is a large crater (23 km in diameter) formed by an impact approximately 39 Ma (Osinski and Lee, 2005). While a crater lake may have been present immediately after the impact, subsequent erosion removed any evidence along with impactites (Osinski and Lee, 2005). A lake formed within ∼8–10 million years of the impact, resulting in the Haughton Formation of dolomite-rich Miocene-age lacustrine sediments (Hickey et al., 1988). The paleohydrology of the lake responsible for the Haughton Formation is poorly constrained, and Quaternary glacial and fluvioglacial sediments overlay the formation itself (Lacelle et al., 2008). The modern climate at HIC is broadly consistent with a polar desert (Lacelle et al., 2008), and water is present mainly as ice, including surface ice, ground ice, and semi-permanent snowbanks (Clarke et al., 2019), with seasonal streams that transport cobble to gravel-sized sediment constituting the only liquid water (Osinski and Lee, 2005; Lacelle et al., 2008). Geochemical characterization of these water sources is consistent with dolomite dissolution (Lacelle et al., 2008), with several lacustrine sediment samples from the Haughton Formation yielding high sulfate concentrations (Léveillé, 2007). Solute concentrations and pH values (7.5–8.3) were lowest in residual snowbanks and increased in snow meltwater and streams, indicating progressive carbonate dissolution (Lacelle et al., 2008).

As a desiccated impact structure that once held a paleolake and currently has a cold desert climate, HIC is a compelling analog to Gale and Jezero craters. Despite the harsh environment, there is abundant evidence of biological activity and biomarker preservation in HIC. Biofilms form on meltwater streaks (several m wide and 100 s of m long) on the crater wall that are comparable to recurring slope lineae on Mars (Clarke et al., 2019) and may be possible to characterize with sub-meter resolution imagery (i.e. aerial, drone, or field-based data). Large sulfur isotope fractionations in the sulfate-rich bedrock (Parnell et al., 2010) provides evidence for persistent microbial sulfate reduction. Endolithic microbial communities, including cyanobacterial species, can thrive in the impact-shocked rocks at HIC (Cockell et al., 2002, 2003). Finally, various lipid biomarkers of biogenic origin, including tricyclic terpenes, hopanes, steranes, and pregnanes (Parnell et al., 2006; Lacelle et al., 2009; Lindgren et al., 2009), occur in HIC microenvironments such as melt breccias, dolomitic limestone bedrock, and young surface ice (Parnell et al., 2004, 2005).

The Atacama Desert in Northern Chile is the oldest continually arid region on Earth, with desert conditions having developed in the Late Triassic (Clarke, 2006; Jordan et al., 2014). Precipitation of up to 25 mm yr⁻¹ falls mainly in the austral summer and potential evaporation is ∼2000 mm yr⁻¹, resulting in a negative hydric water balance (Risacher et al., 1999). Persistent desert conditions have formed distinctly Mars-like soils, which are up to 2 million years old, and are characterized by extremely low moisture, low concentrations of organic matter, and Mars-like salt compositions, including perchlorates (Ericksen, 1983; Navarro-González et al., 2003; Ewing et al., 2008). Atacama surface soils have the highest concentrations of perchlorate found naturally on Earth (up to 0.6 wt%), which approach the concentrations reported for Martian soils (0.5–1 wt%, Davila et al., 2013). This mountainous region contains many closed (up to hundreds of meters) evaporitic basins, known as salt flats or salars, some of which contain saline lakes or liquid brines.

Salar de Atacama is the largest evaporitic basin (Arriagada et al., 2006; Flahaut et al., 2017). Smaller basins, such as Salar de Laco and Salar de Aguas Calientes, are located at higher elevation (Flahaut et al., 2017; Aerts et al., 2020). Considered together, these salars represent a natural gradient in elevation, temperature, evaporation, and soil moisture (Risacher et al., 2003; Aerts et al., 2020). Solar irradiance is uniformly high. Salar lakes (lagoons) and pools are fed by ephemeral streams, snowmelt, springs, and/or groundwater; evaporation exceeds inflows in all cases and salinities range from hypo to hypersaline (Boschetti et al., 2007). Salar brines are hypersaline (350–550 gL^-1 of salt), near neutral pH, with low redox potential (<-30 mV) (Risacher et al., 2003; Boschetti et al., 2007; Arias et al., 2017; Cubillos et al., 2018). No brines in the Atacama are alkaline (Lowenstein and Risacher, 2009). Mean annual air temperatures range from 1 to 14°C, depending on elevation, and mean daily water temperatures range from 10 to 25°C, depending on season and elevation (Boschetti et al., 2007; Cubillos et al., 2018).

Despite the multiple extreme conditions, patchy microbial communities survive in both surface and subsurface soils and in the hypersaline brines, though the population of bacterial cells in soils is low (or was not detected) in some areas (Navarro-González et al., 2003; Connon et al., 2007; Crits-Christoph et al., 2013). Fueled by the hygroscopic abilities of halite, endoevaporitic microbial communities (dominated by phototrophs), persist 3–7 mm beneath salt crusts (Wierzchos et al., 2006; Davila et al., 2008). Halophilic archaea from the family Halobacteriaceae, some of which are capable of surviving in high lithium environments, thrive in salt crusts, brines, and pools (Cubillos et al., 2018).

Salar soils contain biomarkers in Neogene evaporites that record past aqueous conditions (Pueyo et al., 2002). Lipid biomarkers, which are diagnostic of prokaryotic life (Wilhelm et al., 2017, 2019; Sánchez-García et al., (2018)), are well-preserved in salars due to the hyperaridity (Wilhelm et al., 2017) and presence of halite and gypsum salts (Fernández-Remolar et al., 2013; dos Santos et al., 2016).

The McMurdo Dry Valleys (MDV) contain numerous lakes and are the largest ice-free portion of the Antarctic continent (McKnight et al., 1999). Conditions in this polar desert ecosystem are cold (mean annual temperature is −17 to −25°C) and arid (mean annual precipitation is <10 mm), and experience strong katabatic winds from the polar plateau (Fountain et al., 1999; Doran et al., 2002; Doran and Fountain, 2021).

The MDV region contains several ice-covered lakes in hydraulically closed basins that receive inputs mainly from local glaciers (Fountain et al., 1999; Gooseff et al., 2011). Lake Bonney is the most-well studied, beginning in the 1960s (e.g., Armitage and House, 1962), and receives water inputs from four ephemeral streams fed by Taylor Glacier, which it abuts. Water remains liquid (−5°C to +6°C) due to heat from solar radiation, stream inflow, and the latent heat of ice formation (McKay et al., 1985). Bonney consists of two lobes (East and West) that are connected by a shallow sill; both lobes have highly stratified water columns that contain a fresh, well-oxygenated epilimnion that overlies a hypersaline (>70 ppt), suboxic or anoxic, nutrient rich hypolimnion (Spigel and Priscu, 1998; Gooseff et al., 2011). Of relevance to Mars, perchlorate has been measured in the water column (0.05–0.5 μgL⁻¹) (Jackson et al., 2012).

Antarctic perennially ice-covered aquatic environments are considered oases for life in a harsh, polyextreme environment. Lake Bonney’s water column hosts a complex microbial ecosystem with evidence for phylogenetically diverse bacteria, microalgae and fungi and many microbial metabolisms, including photoautotrophy, mixotrophy, and chemolithoautotrophy (Paerl and Priscu, 1998; Vick-Majors et al., 2014; Vick-Majors and Priscu, 2019; Patriarche et al., 2021). Despite the ice cover limiting light to <3% of incident, photosynthesis by cryptophytes and chrysophytes occurs in the epilimnion and phytoplankton and bacterioplankton production are maximized directly beneath the ice cover and just above the chemocline, where nutrients are supplied via diffusion (Priscu et al., 1988). Microbial sulfate reduction occurs below the chemocline in both lobes where sulfate is present (Lyons and Welch, 2016). Cyanobacterial mats are common in the lake margins (moats) that form during the short summer and in glacial meltwater streams (Wood et al., 2008; Kohler et al., 2016); dominant taxa include Phormidium, Nostoc, and Oscillatoria species, though diatom algae are also present (Gooseff et al., 2011). Recent research demonstrates the potential for these microbiological mat communities to be characterized via orbital data (Power et al., 2020; Salvatore et al., 2020).

Lake sediments contain many biomarkers, including cyanobacterial filament sheaths, diatom frustules, protozoan cysts, photosynthetic pigments, and minerals (e.g., carbonates) associated with microbial activity (Doran et al., 1994). This site is the only analog that is completely devoid of plants, metazoans, and human influence and therefore provides a unique opportunity to study microbial interactions in a Mars-like system with few confounding factors.

Lonar Lake (LON) is shallow (max depth 6 m) and occupies a young impact crater formed between 15 and 570 ka (Sengupta and Bhandari, 1988; Sengupta et al., 1997; Storzer and Koeberl, 2004; Jourdan et al., 2011). The crater is located on a ∼65 Ma Deccan basalt that overlies the southern Indian shield in the Buldhana district, Maharashtra, India (Kumar, 2005; Wright, 2014). Primary minerals include plagioclase (labradorite) and pyroxenes (augite and pigeonite) (Maloof et al., 2010). Hydrothermal alteration materials including saponite, celadonite, and calcite (Hagerty and Newsom, 2003), are proposed to be the result of groundwater interactions with remnant impact energy (Nayak, 1996). The impact excavated into the multi-layer basaltic aquifer, forming a closed basin lake approximately 25–65 m below local springs (Menzel et al., 2013, 2014; Komatsu et al., 2014). Lonar lies in the core monsoon zone, and surface runoff is the dominant water source during the monsoon season from July to September, with groundwater springs on the crater walls predominant in the dry season (Komatsu et al., 2014; Menzel et al., 2014). Streams have incised Dhara Canyon and formed an alluvial fan inside the crater (Menzel et al., 2013, 2014). The lake is saline (8–40 ppt) and hyperalkaline (3,000–5,000 mg/L) due to evaporative concentration, with pH from 9.5 to 10.5 (Borul, 2012; Menzel et al., 2013; Pedge and Ahirrao, 2013). The water column is highly eutrophic, and the water appears yellow-green to dark green due to the high concentration of algae and dissolved solids (Surve et al., 2021). Laminated sediments form in the anoxic zone (below 4 m), constituting one of the few climate records in Central India (Menzel et al., 2013).

Microbial life is diverse in the waters of Lonar lake, including cyanobacteria, planctomycetes, spirochaetes, and an alkalophilic methanogen (Thakker and Ranade, 2002). Cytophagales and microaerophilic Bacteroidaceae are abundant in soils and contribute to organic matter remineralization (Wani et al., 2006). Lake sediments preserve a variety of biomarkers (i.e., n-alkanes from leaf waxes, n-heptadecane, carotenoids, diplotene, phytane, and moretene) ranging from 1.4 to 10.1 Ka (Sarkar et al., 2015).

The Río Tinto system, located in the Iberian Pyritic Belt in Southwest Spain, is the world’s largest acidic river (∼95 km long) (Boulter, 1993). The estuary system, formed via hydrothermal activity approximately 2 Ma (Fernández-Remolar et al., 2005), flows through ore deposits (iron and copper sulfides) to the Atlantic Ocean. These subsurface deposits include copiapite, jarosite, schwertmannite, halotrichite, and gypsum and are generally depleted in organic compounds (Edwards et al., 2007). Despite freshwater inputs and seasonal wetting-drying cycles, the Tinto River headwaters are characterized by high acidity (pH 2–2.5) and high concentrations of iron and other heavy metals (i.e., Cu, Zn, As, Mn, Cr, etc.) (Amils et al., 2007; Hubbard et al., 2009; Chacon-Baca et al., 2021). The headwaters are an extremely oxidizing environment with redox potential ranging from 220 to 700 mV (Fernández-Remolar et al., 2005; Hubbard et al., 2009). Many minerals with relevance to Mars form in the seasonally arid streambeds, including iron oxides and ferric sulfates (such as hydronium jarosite) (Fernández-Remolar et al., 2005).

Most biomass at Río Tinto is constrained to the riverbed and rock surfaces, where filamentous algae and fungi form biofilms that trap prokaryotic organisms (Amaral Zettler et al., 2002; Amils et al., 2007; Aguilera et al., 2010), but iron and sulfur oxidizing microbes do grow in subsurface sulfide deposits via chemolithotrophic metabolisms (Fernández-Remolar et al., 2005; Amils et al., 2007). High concentrations of subsurface methane suggest anaerobic methane oxidizing microbes may also be present (the MARTE Team et al., 2008). While subsurface deposits are generally organic-poor, smectite phyllosilicates contain organic carbon and preserve a diversity of biomarkers (Fernández-Remolar and Knoll, 2008). Ironstones in headwater terraces contain well-preserved macroscopic and microscopic biosignatures within iron oxide precipitates (Fernández-Remolar et al., 2005).

Field sites relevant to astrobiology experience environmental conditions that expand our understanding of the limits for life (defined as microbial growth or metabolism). On Earth, microbial habitability is limited primarily by water availability and activity (a _w ), where water activity is related to the amount of water in a liquid (Price and Sowers, 2004; Merino et al., 2019). Microbes prefer higher a _w liquids and the lower limit observed for life is a _w = 0.6, but most microbes cannot grow at a _w < 0.9 and only extremophilic microbes can survive below a _w = 0.75 (Oren, 2008). Other geochemical characteristics of water, such as temperature and salinity, influence life via the modulation of water availability. In addition to water, life requires an energy source to drive chemical reactions (Jelen et al., 2016). Physiochemical gradients generate variable redox conditions and these gradients along with proton gradients, likely fueled metabolism and growth as microbial life arose on Earth (Lane et al., 2010).

Each terrestrial analog experiences multiple extreme conditions, and yet microbial communities persist at all sites, demonstrating the ability of extremophilic life to adapt simultaneously to multiple stressors. This occurrence suggests that life may have evolved to withstand the conditions of nearly every niche on Earth, albeit at varying abundances and levels of diversity (Merino et al., 2019). To examine the limits of habitability on Earth and make inferences about the potential habitability of Mars, it is useful to compare the unique combination of boundary conditions that control habitability and influence the structure and function of microbial communities at terrestrial analogs to those predicted for ancient Mars lakes. Here, we characterize four key environmental parameters at each analog—pH, temperature, salinity, and redox potential (Figure 5). The similarity of analog boundary conditions to the conditions predicted for Late Noachian to Early Hesperian lakes in Gale and Jezero craters is indicated by overlapping polygons in this multidimensional space. The polar desert analogs (HIC and BON) are characterized by cold, neutral water with moderate salinity (<50 ppt) and a wide range of redox potentials, and in these ways are most similar to our current understandings of the Martian paleolakes in Gale and Jezero craters. While the remaining sites show less similarity in these four parameters than HIC or BON, they are important for gaining insight into the capability of life to adapt to environments with very low water activity (ATC brines), where mineral assemblages are similar to Mars (RIO), and in basaltic crater lakes (LON).

These comparisons between environmental conditions at terrestrial analogs and in Mars paleo-lakes (Figure 5) may assist researchers to select the analog that is most appropriate to address specific research questions and goals. One limitation of these comparisons is that they depict only the endpoints of published environmental parameters in the present and do not consider how boundary conditions co-vary spatially or temporally (e.g., over seasonal, interannual, or geological time scales). For example, shortly after impacts, hydrothermal systems are known to develop (Osinski et al., 2013). Hydrothermal systems have been proposed as locations where life may have originated on Earth or on Mars. Most impact events that form large craters (>2 km diameter on Earth, >5 km on Mars) have the potential to general a hydrothermal system (Osinski et al., 2013). The geochemical parameters of the hydrothermal system will evolve over time as the impact crater cools (eg., Kring et al., 2020). On Earth, salinities in hydrothermal fluids are often initially high (∼20% NaCl) at high temps (75–100°C) and decrease as temperatures cool (Lüders and Rickers, 2004). Anoxic or potentially euxinic conditions can develop depending on how isolated the crater basin is from the surrounding aqueous environment (O’Sullivan et al., 2016). Redox conditions depend on the minerals present, and hydrothermal craters on Earth can support several microbial metabolisms, including iron reduction, sulfate reduction, sulfur reduction, and methane oxidation (Kring et al., 2020). Habitable temperatures (<120°C) can persist in impact-induced hydrothermal systems for ∼10³–10⁵ years, depending on the impact size and host rock permeability (Abramov, 2004; Kring et al., 2020).

Microorganisms are present within the entirety of the phase space represented at each analog. The diversity, evenness, and abundance of microbial life, however, will vary greatly within the phase space as fewer organisms can tolerate the more “extreme” range of conditions, especially when multiple extreme conditions co-occur. Understanding how these parameters combine to limit life and shape its distribution in time and space is critical to further our understanding of whether life could have evolved on Mars and where to look for it.

Biosignature preservation potential refers to the likelihood that an organic biomarker is lithified and survives diagenesis and/or other degradation processes to be detectable in the present day. Biosignature preservation is extremely heterogeneous due to many factors that vary over small spatial scales (i.e., mm to cm), such as water activity and availability, UV radiation, oxygenation (redox potential), and the formation of organic-mineral complexes which can protect lipids from degradation (Hays et al., 2017). By compiling previous research conducted at the selected Mars analog sites we can examine how mineralogy and environmental conditions impact biosignature preservation potential. A variety of organic biomarkers have been recovered from Mars analogs (Table 1, Supplementary Figure S1, Supplementary Table S3), however, we focused our analysis on lipids, which are important targets for identifying biosignatures on Mars because these complex organic molecules have a low probability of abiotic origin (Des Marais, 2013) and are likely to survive long-term storage (i.e., ∼10⁶ years) on Mars (Georgiou and Deamer, 2014; Srivastava et al., 2020). Lipid biomarkers have been recovered at all five sites, demonstrating that every site preserves biosignatures over Quaternary to Jurassic timescales (up to 10⁸ years) regardless of differences in local mineralogy (Table 1, Supplementary Figure S1, Supplementary Table S3). Lipid biomarkers are categorized into five classes based on hydrocarbon identity and structure after Brocks and Summons (2003). While the taxonomic specificity of lipid biomarkers varies, in general, they provide some insight into what classes of organisms were present in ancient ecosystems (Summons et al., 2021).

Because the surficial mineralogy of Mars can be characterized remotely, understanding how mineralogy affects preservation of biomolecules on Earth is a priority. For example, studies have shown that soils and sediments with high abundances of clay minerals (phyllosilicates) prevent diagenesis and tend to preserve organics well (e.g., Hays et al., 2017). Clay mineral assemblages are found at all analog sites except HIC and may promote biomarker preservation. All six classes of lipid biomarkers have been detected in clay minerals at BON, though the age of these biomarkers has not been determined (Doran et al., 1994). Lipid biomarkers are also well-preserved in halide minerals, such as the halite and gypsum salts found in ATC (Fernández-Remolar et al., 2013; dos Santos et al., 2016). At RIO, 2–8 Ma fossils are well-preserved, likely due to the presence of iron oxide (microfossils) and goethite (macrofossils) (Fernández-Remolar and Knoll, 2008). Iron oxides increase preservation by providing radiation protection (Gómez Gómez et al., 2004). This process is especially relevant to sedimentary rocks formed at Meridiani Planum, the landing site of the Opportunity rover, and suggests these rocks are capable of preserving evidence of microbial life (Fernández-Remolar and Knoll, 2008).

Environmental factors also impact preservation. Radiation exposure decreases preservation potential because long exposures can destroy organic molecules and alter their associated isotopic ratios (Summons et al., 2011). Ultraviolet radiation exposure is especially high at high latitudes, like BON and HIC, and at high elevations, like regions of ATC. UV radiation protection might be encountered in deep lakes, lakes with high suspended solid concentrations, or areas with rapid soil/sediment accumulation (Santschi et al., 1990).

A second key environmental factor impacting biomarker preservation on Earth is the oxygen exposure time (Reiche et al., 2018). Highly productive aquatic systems tend to see high organic matter preservation owing to high respiration rates and correspondingly low oxygen levels (Wetzel, 2001). Similarly, high sediment accumulation rates in deltaic settings reduce the oxygen exposure time, thereby promoting preservation (Naeem et al., 2000; Hays et al., 2017). Deeper lakes promote preservation because anoxic conditions reduce oxidative diagenesis (Berner, 1989) and support sulfurization, which can preserve biosignatures (Raven et al., 2016). Sulfurization enhances preservation over geologic time scales by converting organic matter into more recalcitrant macromolecules. Sulfate minerals are common in Gale, for example in the Murray Formation, and sulfurization is a potential mechanism for preserving organic matter in mudstones. Exploring sulfurization preservation in analogs with deep, closed basin lakes with anoxic bottom waters, like Lake Bonney, would have strong relevance to Gale Crater. In particular, future studies should assess how organic matter structure, sedimentation rate and redox conditions impact the process of sulfurization. Most organic matter preserved in lacustrine and deltaic environments, however, is of terrestrial origin, and may obscure the signal of less productive in situ aquatic communities. Lake Bonney and Haughton Impact Crater, whose watersheds are either completely or partially devoid of terrestrial vegetation, may be especially useful for future biomarker research.

Subaerial environments on Earth’s surface that do not have standing lakes or pools, like Haughton Impact Crater and Atacama soils, preserve a variety of organic, chemical, and isotopic biosignatures (Horgan, 2016). Extremely dry soils, like those of the Atacama, appear to preserve biomarkers well, especially when shielded from UV radiation (Fagliarone et al., 2020). The concentration of organics in these environments is variable, where high concentrations of organic matter are rare and often associated with microbial mats (Hays et al., 2017). Given the uncertainty in surface coverage of microbial communities in desert soils and potentially Mars, determining biomarker preservation potential in these environments is difficult.

The biomarker record at analog sites is complex and there are numerous caveats for interpretation even at locations whose geologic history and modern environment are well-characterized. More research is necessary to further elucidate the process of biomarker preservation within mineral matrices at terrestrial analogs, especially over longer timescales. Our analysis identifies which minerals are currently underexplored with respect to lipid biomarkers at each analog site (Supplementary Figure S1). Such work is vital to predict what diagenetic processes could occur in similar environments on Mars and can inform site selection for future astrobiological missions.

Many of the sedimentary rocks on the Martian surface have experienced minimal alteration since their deposition due to the lack of robust plate tectonics on Mars (Breuer and Spohn, 2003). For example, the ancient smectite-bearing mudstones found in Gale crater show little mineralogical evidence for burial diagenesis, despite burial depths of up to 5 km (Borlina et al., 2015). This phenomenon likely increases the potential for biomarkers to be preserved. However, there are also many challenges for biomarker preservation on Mars, which were reviewed extensively by Hays and others (2017). Briefly, these include (1) water activity and availability, which impact both habitability and biosignature preservation potential; (2) chemical oxidants, such as perchlorates, are common in Martian regolith and can degrade organic biosignatures; and (3) high ultraviolet radiation on Mars’ surface. Exposure ages on Mars vary and some areas of Gale Crater are as young as 80 ± 30 million years (Farley et al., 2014), but continued exposure to ionizing radiation will eventually destroy all biomarker signals in the top few cm’s of exposed regolith and can impact biomarkers in the top 1 m of the surface (Hassler et al., 2014; Hays et al., 2017).

The rovers currently exploring Gale and Jezero craters, Curiosity and Perseverance, respectively, both possess analytical instrumentation capable of detecting lipid biosignatures (i.e., alkanes, isoprenoids, fatty acids, or alcohols) and other organic biomarkers. Curiosity, in Gale Crater, is outfitted with a suite of instruments known as the Sample Analysis at Mars (SAM), which can detect a variety of biosignatures ranging from polar and non-polar organics, large organic molecules (>20 C atoms per molecule), and refractory organics (Mahaffy et al., 2012). Perseverance, in Jezero Crater, carries the Scanning Habitable Environments with Raman and Luminescence for Organics and Chemicals (SHERLOC), which uses a UV Raman spectrometer and fine-scale imaging capabilities to detect biosignatures (Beegle et al., 2015).

Detecting certain lipid classes via mass spectroscopy methods often requires extraction in organic solvents and separation of the total lipid extract into component parts based on polarity (Azua-Bustos et al., 2020). SAM has the capabilities to perform some extractions, including volatilizing lipids for GC-MS analysis with tetramethylammonium hydroxide (Williams et al., 2019). The spectroscopic method used by SHERLOC does not require pre-derivatization but is more susceptible to matrix effects, which can occur when organics are mixed with regolith or sediment and have the potential to confound biomarker detection (Abbey et al., 2017). The detection limits for SAM (on the order of ppb, Williams et al., 2019) are approximately 1000 times more sensitive than for SHERLOC (on the order of ppm, Beegle et al., 2015). The Rosalind Franklin is scheduled to launch in September 2022 and expected to land in Oxia Planum, a clay-rich plain (Mandon et al., 2021), in mid-2023. This rover will detect biomarkers via the Mars Organic Molecular Analyzer (MOMA), which contains a gas chromatographic-mass spectrometer and a Raman spectrometer (Siljeström et al., 2014; Goesmann et al., 2017). MOMA cannot perform complex pre-treatments and has a detection limit similar to SHERLOC (Siljeström et al., 2014; Goesmann et al., 2017). Unlike the other rovers, the Rosalind Franklin can drill to 2 m depths (Kereszturi et al., 2016), which may allow the collection of regolith with higher preservation potential. The concentration and distribution of biomarkers will have a large impact on the ability of rovers to detect biomarkers in situ due to these inherent instrument sensitivities. Thus, identification of targets that are most likely to harbor preserved biomarkers is critical. Field studies with rover-scale technologies at terrestrial analogs provides an opportunity to identify favorable working conditions for specific instruments or techniques and can increase the chance of detecting features of interest (e.g., Bontognali et al., 2021).

Remote sensing techniques are a cornerstone of Mars research because they enable frequent coverage of large areas and are a cost-effective precursor to in situ analysis (Malin et al., 2007; van der Meer et al., 2012). Interpretation of satellite data informs landing site selection (Golombek et al., 2017; Grant et al., 2011, 2018) and rover-based spectral and image data are used to determine targets for follow up in situ analysis, and in some cases, sample collection (i.e., Farley et al., 2020). In addition to providing information on topography and mineralogy, remote sensing may also be useful for detecting evidence of life and identifying regions on Mars amenable to subsurface sampling (Kereszturi, 2012a).

Applying remote sensing to terrestrial analogs with well-characterized biological features provides an opportunity to develop techniques to remotely detect signs of extant (and possibly extinct) microbial life. These techniques could be used to inform rover landing site selection and sample collection on Mars, as well as other terrestrial bodies with potential astrobiological importance. In Section 3.4.1, we consider the spatial scale and morphology of macrobiogenic features (> cm scale) at each analog in comparison with the resolution of Mars remote imagers to determine what types of observations would be necessary to detect these microbially-generated features. In Section 3.4.2, we discuss visible and near infrared (VNIR) spectral features of minerals present at each analog in relation to instrumentation coverage of Mars imagers. Further, we consider which of the minerals that can be spectrally detected have high biomarker preservation potential, which can inform the search for putative biomarkers on Mars (Bosak et al., 2021).