Mars has a well-preserved sedimentary record that date as far back as 4.3 Ga and perhaps earlier 10,11, and the early presence of habitable environments is now well-documented 1–3. Early Mars had liquid water, energy sources, CHNOPS elements, catalytic transition metal elements, as well as B, Mg, Ca, Na and K, all of which are linked to existing and evolved life as we know it. Yet, much more uncertain is whether Mars’ surface conditions were conducive to an independent origin of life 4,5, or abiogenesis, which is the natural process by which life may have arisen from initial non-living matter, such as simple organic compounds. Prebiotic chemistry leading to incipient life forms is thought to be restricted not only to certain limited conditions, but may even require a specific succession of distinct environmental changes for it to succeed 6. As such, it differs substantially from more broadly defined requirements for habitability.
Here we report on new lithological and geochemical data from Hesperian-aged (~3.6 Gyrs old) strata of Mars that indicate wet-dry cycling, a process considered supportive of, and perhaps essential, for prebiotic chemical evolution 7–9. In situ investigation of hundreds of meters of sedimentary stratigraphy within Gale crater has revealed a record of ancient aqueous surface environments from fluvio-lacustrine 12 to more intermittent lake or lake margin settings up-section. The latter is interpreted as seasonal or secular drying 13–15, or as discrete dry intervals 16. After exploring strata dominated by clay-bearing mudstones, the rover reached the sulfate-bearing unit which is evidence for a major environmental transition 17 characteristic of stratified terrains found in other regions across Mars 18. At this transition data collected by the rover uncovered a new type of sulfate-enriched evaporitic-clastic deposit.
Pervasive centimeter scale polygonal patterns in the basal sulfate-bearing stratigraphic unit manifest as straight ridges that intersect with triple junctions, with the most prominent occurrence observed on the 3154th mission sol (Fig. 1 and Extended Data Fig. 1). Several additional exhumed strata nearby show comparable, as well as incipient and altered variants of these patterns (Extended Data Fig. 3). The polygons persist to at least decimetric depth within the bedrock, as is evident from their stepped appearance on thick blocks of bedrock (Fig. 1b). On bedding planes these polygons are ~1 cm in height, and about 4 cm in diameter – varying between 1 and 7 cm (Extended Data Fig. 4). The ridges commonly consist of aligned nodules, variably juxtaposed, and irregular in shape and size (Fig. 1d). In contrast, polygons cores are smooth surfaced and represent the host bedrock. Chemical composition documented by the ChemCam instrument shows a significant increase of Ca-sulfate content and variable Mg-sulfate enrichment within the polygonal ridges and other nodular features, whereas the host bedrock is of basaltic bulk composition and sulfate-poor (Fig. 2 and Extended Data Table 2).
Although polygonal ridges commonly form in salt crusts and playas on Earth as a consequence of subsurface salinity convection 19, we do not favor this interpretation here. Terrestrial salt crusts are mostly pure and consist of ephemeral salt deposits that form larger polygons 0.5 to 2 m in size 20, and the lower gravity on Mars should have given rise to convection cells of even larger size than observed on Earth. Instead, we interpret the polygonal sulfate-bearing ridges as the fill of open desiccation cracks in muds by variably coalescent, salt-bearing and sediment-inclusive nodules (Fig. 3). Whereas desiccation cracks in fresh mud layers initially form T-junctions, maturation over repeated drying cycles results in convergence towards hexagonal shapes with junction angles near 120°, i.e. Y-junctions 21. In experiments, using multi-millimetric clay layers, joint angles tend towards 120° after 10 consecutive dryings with significantly more cycles required to mature pattern into hexagonal shapes 22.
The abundance of 30 to 50 wt.% Ca-sulfate and up to 40 wt.% Mg-sulfate measured within the ridges and nodular bedrock and their much lower abundance in the host bedrock (Fig. 2) collectively suggest that sulfate minerals precipitated with evaporation in muds and incorporated detrital sediment in the process. The present appearance of the sulfate-bearing ridges likely is not the original configuration of these features. Rather they started out as evapoconcentration focused on initially formed cracks that then evolved over a longer history of drying cycles and burial diagenesis (Fig. 3). They are now exposed as erosion resistant polygonal ridges due to their higher degree of cementation relative to the host bedrock, and an early bias of surface salt precipitation in original mud-crack polygons (Fig. 3d).
Recurrent wetting of surface muds likely reflects a combination of flooding and groundwater recharge. Flooding could have dissolved salts that formed ephemeral surface crusts, as well as adding sediment and promoting burial, although air-fall is another possible source for sediment addition. The polygons’ size range of a few centimeters (Fig. 1), suggests that desiccation and mud contraction affected only the uppermost few centimeters of the surface muds 22. The variability in shape and size of the sulfate-rich nodules that precipitated within the ridges indicates several generations of nodule growth, in consistence with significant intra-sediment salinity fluctuations due to repeated drying cycles (Fig. 3).
Given the apparent repetition and limited desiccation depth, the associated wet-dry cycles could have been seasonal but potentially also active on shorter time scales. The time span during which these wet-dry cycles left their intermittent signature in the sediments may be approximated via the thickness of the polygon-bearing stratigraphic interval (18 m). Using terrestrial analogs for sedimentation rate 23, thousands to millions of years may be indicated. Even under assumption of the shorter estimate, the duration over which wet-dry cycling in evaporative ponds may have recurred in Gale crater is significant relative to the timescale considered favorable to prebiotic chemical evolution 24.
The two major observations on polygonal features combine the hexagonal shapes that indicates repeated drying cycles, and their persistence within stratigraphic thickness implying wet-dry conditions were maintained in the long-term. Mud cracks were observed in underlying strata of the Murray formation, but predominantly formed T-junctions that suggest a single desiccation event 14,16.
The polygonal features described here are the first tangible evidence for sustained wet-dry cycling on early Mars, a process considered strongly supportive for prebiotic chemical evolution. During dryings, water activity is lowered and the concentration of soluble ingredients in the residual liquid is increased, boosting reaction rates, especially for higher-order reactions 7. For instance, the reactions that form nucleotides from their constituent nucleobases (ribose, phosphate) produce H2O and hence are favored at low water activity. Most importantly, the polymerization reactions necessary to advance from nucleotides to RNA or DNA, and from amino acids to proteins, require dehydration steps which have been demonstrated to be facilitated by wet–dry cycles 25–27. Under the right environmental conditions Darwin’s proverbial “warm little pond” could significantly promote the reactions for macromolecule polymerization, and through sustained wet-dry cycling increase the likelihood of chemical evolution towards the origination of life 24.
In a broader regional context, there is widespread documentation of preserved organics in ancient Gale crater strata, containing up to ~0.5 kg/m3 3,28, in addition to a variety of other soluble elements 3. The addition of direct evidence for a series of repeated wet-dry cycles presented here supports that conditions in ancient Gale crater were conducive to prebiotic polymerization processes. There is definite potential for wet-dry cycling to have occurred more broadly on Mars (Fig. 4) in the period when both intrabasin sulfate salts and clays were deposited 1 as Gale is a stratigraphy of global significance for this mineral assemblage near the Noachian-Hesperian transition 29. Some of these strata may thus harbor well-preserved evidence of prebiotic chemical evolution, a record that is no more available on Earth 30. On the basis of this new evidence for sub-aerial wet-dry cycling within early habitable environments, and considering the delivery of organics and accumulation of volatiles on the Martian surface for almost a billion years prior (Fig. 4), our findings suggest that the Noachian-Hesperian transition period could have been favorable for the emergence of life; possibly more than the earlier Noachian eon that has potential for perennially wet surface environments 18.