INTRODUCTION
Because life on Earth appeared early in our planet's history and therefore must have originated relatively rapidly, it is possible that liquid water persisted on Mars long enough for life to begin there as well. Thus, the fundamental question of whether life exists or existed elsewhere in the universe may be answered in our own galactic back- yard. This mandates that we search for a martian fossil record. The current surface environment of Mars is hostile to life as we know it, and an ancient biosphere might have become extinct. The discovery of an extinct martian biosphere would be of immense scientific importance: the demonstration that life originated independently in two places in our solar system would have far-reaching implications for the distribution of life elsewhere in the Universe.
Below is a concept to search for a biosphere which became extinct perhaps a few Gyr ago. The search is naturally allied with investigations of prebiotic chemistry and extant life because all three efforts will be largely chemical studies and because the locations to be targeted on Mars will overlap extensively. Such an investigation inevitably draws heavily on our experience with studies of earth's early biosphere.
EXPERIENCE WITH EARTH'S EARLY FOSSIL RECORD: RELEVANCE FOR MARS
Tectonic activity has altered, concealed or destroyed most of Earth's early crust. Careful regional geologic mapping has therefore been necessary to locate and characterize older Precambrian sequences. Many early Archean aqueous sediments occur typically as relatively thin-bedded units within thick sequences of volcanic rocks. These sequences have been explored for those sediment types most likely to contain fossils. Fossils have typically been found in cherts, which are impermeable siliceous rocks that resist weathering. The best-preserved specimens are found in rocks having fine-grained, stable mineral textures with well-preserved (and/or abundant) organic matter or other reduced chemical species. These samples have been searched for morphological, chemical or isotopic evidence of ancient life, and they have been interpreted in the context of their preservation and paleoenvironment. It is significant that the antiquity of the fossil record (3.5 Gyr) corresponds with the age of the oldest sediments which have been sufficiently well-preserved to have retained conclusive evidence of life. Thus, the antiquity of the fossil record is probably limited by the preservation of rocks favorable for fossil preservation, and life is likely to have appeared significantly earlier than the oldest sedimentary sequences.
Certain concepts have emerged from Precambrian studies which seem relevant to a search for life on ancient Mars. For more than five- sixths of our own biosphere's history, life existed predominantly as single-celled organisms. Thus the most ancient and ubiquitous life form in a putative martian biosphere would presumably have been microbial. Also, mineralization and/or rapid sedimentation which occurs in close proximity to microbial communities will enhance their fossilization and preservation. This is because decomposition proceeds to completion unless the cells have been isolated from decay. Short-term isolation can sometimes occur in high-salinity environments where decomposition by microbial activity has been suppressed; however, long-term fossilization additionally requires entombment in an impermeable mineral matrix. Some of the best preserved examples of Precambrian fossils were probably rapidly encased in primary silica precipitates before decay could occur.
Because multiple lines of evidence (morphologic, sedimentological and chemical -- including isotopic) have been crucial for interpreting Earth's own Archean fossil record, they will probably be required to prove that a prospective martian fossil deposit is indeed biogenic. Our ability to identify the remains of hypothetical martian organisms depends ultimately upon comparisons with extant terrestrial analogs, and such comparisons may prove to be inaccurate. Furthermore, biological information preserved in sediments is often altered or destroyed by biological degradation, elevated temperatures or pressures, or oxidation.
Morphologic evidence includes forms which are visible at various size scales, ranging from microscopic cells to macroscopic microbial constructs, such as stromatolites. "Chemical fossils" include those biologically produced substances which can be conclusively distinguished from nonbiological ones. Most notable among these are distinctive organic compounds (e.g., certain lipids or amino acids), although inorganic substances, such as certain phosphate minerals, can also be diagnostic. Differences in isotopic composition, such as the contrast observed on Earth between sedimentary organic carbon and carbonate, can retain the signature of biological isotopic fractionation.
Fossil evidence can be quite abundant in sedimentary rock sequences which have escaped extensive degradation. One illustrative example is the 600 to 900 Myr-old sequence from Svalbard, an island located midway between Scandinavia and the North Pole. About 25 percent of all carbonates in these rocks contain fossil stromatolites formed by microbial communities. (Note that macroscopic evidence of stratiform or domal laminated rocks is not sufficient to prove microbial genesis. Abiotic structures as various as calcretes, tufas, stalagmites, and even malachite bodies share similar physical characteristics. A role for microbial mat communities must be demonstrated on the basis of detailed petrologic study.) About 25 percent of all fine-grained siliciclastic samples contain organic-walled microfossils. About 5 percent of all siliciclastics contain diverse, well-preserved fossils that have been invaluable for taxonomic studies. About 10 percent of all carbonates contain microfossils; however, more than 50 percent of silicified carbonates contain fossils. Essentially all carbonates and fine- grained siliciclastic sediments that contain organic matter provide carbon isotopic evidence indicating biological discrimination.
The Svalbard rocks offer perspectives which are relevant to the Mars exploration effort. The likelihood for retaining evidence for life in well-preserved rocks of the right mineralogy is high. The best preservation is found in cherts and unoxidized fine-grained siliciclastic rocks, although carbonates also provide important information. The best strategy for finding fossils should combine visual with chemical and isotopic observations. Based on experience with the Precambrian on Earth, the detection of morphological microfossils has an inherently lower probability of success than the detection of chemical evidence, but it offers great rewards if successful. Chemical and isotopic observations greatly improve the overall probability for locating ancient evidence of life. In either case, success rests largely on targeting appropriate lithologies.
Compared to the Earth, the ancient martian crust has been less affected by destructive tectonic forces, thus the discoveries of remarkably abundant microbial remains in Earth's well-preserved Precambrian carbonates and shales are encouraging for Mars exploration. However the fraction of ancient sediments deposited in aqueous environments is likely to be smaller on Mars than on Earth, particularly during the past 3 Gyr. Thus, it should be easier to find well-preserved ancient martian crust, but it will be perhaps more challenging to locate and sample aqueous sediments within crustal sediments. The principal strategy, then, is to locate and analyze aqueous sediments, particularly those that are good repositories for a fossil record.
KEY SITES ON MARS
Life's fundamental requirements for liquid water, energy and nutrients form the basis of a search for sites on Mars which are most prospective for locating a fossil record. If we allow that life might have been initially autotrophic rather than heterotrophic, then it follows that all plausible energy sources which could drive metabolism should be considered. Solar radiation is likely to have been the major, reliable source of energy, but access to it requires elaborate photochemical systems for conversion of physical (light) energy into chemical energy. It is presently unknown whether such systems evolved early on Mars or the Earth. On the other hand, it is conceivable that an early metabolic system may have used sources of chemical energy. Rather than requiring the machinery with which to capture light energy, early microbes could have been chemoautotrophs and perhaps chemolithoautotrophs. Therefore all potential sources of chemical energy that could arise at or near the surface of Mars or the early Earth should be considered.
Thermal-spring deposits
Subaerial thermal-spring deposits have been identified as
important targets for locating a martian fossil record because such
springs might have been oases in the literal sense, and they also would
have provided reduced gases to serve as sources of energy and reducing
power for organic synthesis. Thermal-spring waters also can sustain
the high rates of mineral precipitation which, on Earth, typically occur
in the presence of microbial communities. Volcanic terrains are
widespread on Mars, and some include outflow channels of simple
morphology that may have formed by spring sapping.
The association of such features with potential heat sources, such as volcanic cones or thermokarst, provides evidence for near-surface hydrothermal systems. Minerals most commonly deposited by subaerial thermal springs include silica, carbonates and iron-oxides. Siliceous sediments are particularly favored because they tend to be fine-grained and are relatively stable during post-depositional changes. Organic-rich cherts (fine-grained deposits of silica) provide some of our best examples of microbial preservation in the Precambrian. Although rates of organic- matter degradation appear to be quite high within thermal environments, a great deal of biological information is retained in the macroscopic biosedimentary structures (stromatolites) and biogenic microfabrics of spring deposits. Many primary biogenic features of subaerial spring deposits survive diagenesis (e.g., recrystallization, phase changes). Spring deposits are excellent targets for fluid inclusions, which may preserve samples of original liquid and volatile phases and, potentially, micro-organisms and bio-markers. Such deposits are also prime targets for prebiotic compounds.
Lakes
Lake beds are important targets, because ice-covered lakes
might have been sites for life's "last stand" on the martian surface.
Subaqueous spring carbonates (tufas) and sedimentary cements
deposited at ambient temperatures often precipitate rapidly in the
presence of microbial communities. Such deposits form when fresh
water emerges from springs at the bottom of alkaline lakes. Mineral
precipitation is apparently driven primarily by inorganic processes,
although microbial mats may also influence precipitation during
periods of peak productivity in areas where rates of inorganic
precipitation are lower. Tufa deposits often contain abundant microbial
fossils and organic matter. Sublacustrine springs commonly occur in
volcanic settings, in association with crater and caldera lakes, and can
include thermal deposits. Volcani-lacustrine deposits are frequently
strongly mineralized and comprise some of our best examples of well-
preserved microbial communities in terrestrial settings
Evaporite deposits
When a lake shrinks or disappears by evaporation, the more-
soluble salts can precipitate and capture other constituents. When
evaporites crystallize from solution, they commonly entrap large
numbers of salt-tolerant bacteria within brine inclusions. Evaporites
have been suggested as potential targets for extant life on Mars,
although considerable debate currently exists regarding the long-term
viability of such micro-organisms within salt. Still, brine inclusions in
evaporites may provide excellent environments for preserving fossil
microbes and biomolecules. The disadvantage of evaporites is that
their high solubility limits them to comparatively short crustal residence
times. Consequently, most Precambrian evaporites are known from
crystal pseudomorphs preserved by early replacement with stable
minerals, such as silica or barite. On Mars, the most likely places for
evaporites are terminal lake basins where standing bodies of water
existed perhaps intermittently. The central portions of terminal lake
basins, including impact craters and volcanic calderas, are potential
targets for evaporites on Mars. The typical "bulls eye" distribution
pattern for evaporites within such settings indicates that carbonates are
normally found in marginal basin areas, with sulfates and halides
occurring progressively more basinward. Such basins also might
contain fine-grained, clay-rich siltstones and shales. These rock types
are significant in that, on Earth, they sequester the bulk of sedimentary
organic carbon and other reduced species such as biogenic sulfides.
Although these rocks are usually less resistant to weathering than cherts
or carbonates, their sometimes high organic contents on Earth
emphasize their potential importance on Mars.
Cemented regolith
As surface water percolates downward through soils, more-
soluble compounds tend to be dissolved from the upper horizons and
redeposited at depth as mineralized "hard-pans" (e.g., calcretes,
silcretes). Mineralized soils commonly contain microfossils of the soil
microbiota entombed in hard-pans or duricrusts. Mineralized horizons
within paleosols may be widespread on Mars and should not be
overlooked as potential targets for exobiology. Indeed, images from the
Viking landers indicate that, in places, soils are indurated and form
surface crusts which are resistant to ablation by wind. Rock varnish,
dark coatings observed in arid environments, often reflect biological
processes, and therefore should be searched for on Mars.
OBJECTIVES FOR FLIGHT EXPERIMENTS
It will be important to achieve a global perspective of the extent to which liquid water has altered the chemical composition of the martian crust. Elements such as calcium and aluminum respond quite differently during aqueous weathering of igneous rocks. This creates a wide range in the Ca/Al elemental abundance ratio among the various products of aqueous weathering. Similar, albeit somewhat less dramatic, patterns are also observed for the elements magnesium, sodium and potassium, relative to aluminum. Correlations between the ratios of these elements and other morphological indicators of the activity of liquid water could be informative. Gamma-ray spectroscopy can perform elemental analysis of the martian crust from orbit and thus is important for evaluating elemental distributions on a global scale, although the prevalence of a widespread aeolian layer comparable in thickness to the gamma-ray penetration depth could be a complicating factor. The best information will likely be obtained from areas of aeolian erosion where bedrock is exposed at the surface, and such areas should be targeted for orbital spectroscopy.
On Earth, high local concentrations of certain elements are particularly diagnostic of biological processes. Phosphorus is notable among these elements; it is a major constituent of bone and its deposition as phosphate-rich rock often reflects the decomposition of sedimented organic matter. Thus, in addition to the ability of phosphate to entomb and preserve fossil materials, an elevated abundance of phosphate minerals might be a key indicator of past biological activity.
A major effort should be made to locate rock types which are favorable for the preservation of fossils. These key rock types include carbonates, phosphates, evaporites, and silica-rich precipitates such as cherts. This effort should be pursued from orbiters, landers or rovers. These key minerals typically have characteristic spectral signatures in the near- and mid-infrared (IR). In addition, a number of diagnostic siliciclastic minerals, including clays, are formed by the aqueous alteration of igneous rocks. These minerals can also be identified using IR spectroscopy. Epithermal hydrothermal deposits have also been detected using airborne magnetometers.
Aqueous minerals that are both fine-grained and diagenetically stable are favored for good preservation. These minerals can be identified in near-IR reflectance spectra or mid-IR emission spectra. The presence of reduced chemical species such as organic matter may also be detected spectroscopically. Because such rocks might be relatively rare at a given landing site, rover-based spot spectroscopy would be useful for surveying and quickly evaluating numerous rocks in order to assess the chemical, mineralogical and petrologic diversity of a site. Imaging at visible and infrared wavelengths is fundamentally important for locating favorable rock types for a chemical or morphological fossil record. Camera capabilities should allow not only panoramic views but also telephoto and close-up "hand-lens" options for evaluating rock textures. Microscopy should be done with a range of light sources, including UV to detect fluorescence. This can provide important information about mineralogy and also be used to search for organic matter. For rocks found to be promising, more detailed analyses of their elemental composition, mineralogy and volatile contents are warranted. Because weathering and other processes can alter rock surfaces, an accurate assessment of mineralogy or volatile content might require that rock interiors be sampled. Thus, a drill or other device must be developed for penetrating rocks.
SAMPLE RETURN
Virtually all investigators in the Precambrian paleontology community believe that definitive discovery of martian fossils will require observations made in Earth-based laboratories. Thus, the exploration activities described above might constitute only a critical prelude to the actual revelation itself. The need for sample return is based on the conviction that fossil evidence will be challenging to obtain, and that the analytical capabilities of Earth-based laboratories are indeed much more formidable than flight instrumentation. Secondly, the latest technologies can be applied and new methods can even be developed and tailored to the task. Perhaps most importantly, the fullness of human perception, flexibility and insight can be brought to bear on returned samples.