The presence of perennially wet surface environments on early Mars is well documented1,2, but little is known about short-term episodicity in the early hydroclimate3. Post-depositional processes driven by such short-term fluctuations may produce distinct structures, yet these are rarely preserved in the sedimentary record4. Incomplete geological constraints have led global models of the early Mars water cycle and climate to produce diverging results5,6. Here we report observations by the Curiosity rover at Gale Crater indicating that high-frequency wet–dry cycling occurred in early Martian surface environments. We observe exhumed centimetric polygonal ridges with sulfate enrichments, joined at Y-junctions, that record cracks formed in fresh mud owing to repeated wet–dry cycles of regular intensity. Instead of sporadic hydrological activity induced by impacts or volcanoes5, our findings point to a sustained, cyclic, possibly seasonal, climate on early Mars. Furthermore, as wet–dry cycling can promote prebiotic polymerization7,8, the Gale evaporitic basin may have been particularly conducive to these processes. The observed polygonal patterns are physically and temporally associated with the transition from smectite clays to sulfate-bearing strata, a globally distributed mineral transition1. This indicates that the Noachian–Hesperian transition (3.8–3.6 billion years ago) may have sustained an Earth-like climate regime and surface environments favourable to prebiotic evolution.

1. Nature | Vol 620 | 10 August 2023 | 299
Article
Sustainedwet–drycyclingonearlyMars
W. Rapin1✉, G. Dromart2
, B. C. Clark3
, J. Schieber4
, E. S. Kite5
, L. C. Kah6
, L. M. Thompson7
,
O. Gasnault1
, J. Lasue1
, P.-Y. Meslin1
, P. J. Gasda8
& N. L. Lanza8
ThepresenceofperenniallywetsurfaceenvironmentsonearlyMarsiswell
documented1,2
,butlittleisknownaboutshort-termepisodicityintheearly
hydroclimate3
.Post-depositionalprocessesdrivenbysuchshort-termfluctuations
mayproducedistinctstructures,yetthesearerarelypreservedinthesedimentary
record4
.IncompletegeologicalconstraintshaveledglobalmodelsoftheearlyMars
watercycleandclimatetoproducedivergingresults5,6
.Herewereportobservations
bytheCuriosityroveratGaleCraterindicatingthathigh-frequencywet–drycycling
occurredinearlyMartiansurfaceenvironments.Weobserveexhumedcentimetric
polygonalridgeswithsulfateenrichments,joinedatY-junctions,thatrecordcracks
formedinfreshmudowingtorepeatedwet–drycyclesofregularintensity.Insteadof
sporadichydrologicalactivityinducedbyimpactsorvolcanoes5
,ourfindingspointto
asustained,cyclic,possiblyseasonal,climateonearlyMars.Furthermore,aswet–dry
cyclingcanpromoteprebioticpolymerization7,8
,theGaleevaporiticbasinmayhave
beenparticularlyconducivetotheseprocesses.Theobservedpolygonalpatternsare
physicallyandtemporallyassociatedwiththetransitionfromsmectiteclaysto
sulfate-bearingstrata,agloballydistributedmineraltransition1
.Thisindicatesthat
theNoachian–Hesperiantransition(3.8–3.6 billionyearsago)mayhavesustainedan
Earth-likeclimateregimeandsurfaceenvironmentsfavourabletoprebioticevolution.
Mars has a well preserved sedimentary record that dates as far back
as 4.3 billion years ago and perhaps earlier9,10
. The early presence of
habitableenvironmentsandevenperenniallywetsurfaceenvironments
hasbeenwellestablished1,2
.Littleisknown,however,aboutshort-term
episodicity and potential periodicity in early hydroclimate regimes3
.
Post-depositional processes driven by short-term fluctuations in a
hydroclimateregimemostlyleavesurficialimprints(forexample,mud
cracks). Although these surficial imprints are prone to erosion, they
arenonethelesscriticalforunderstandingpastsurfaceenvironments4
.
Moregenerally,widelydivergingmodelsoftheseasonalityandepiso-
dicityofearlyMars’swatercycleandaridityarepoorlyconstrained11–16
.
Here we report on well preserved polygonal patterns exhumed from
Hesperian-aged(about3.6 Gyrold)stratathatindicatewet–drycycling
andprovideinsightsonthehydroclimateandastrobiologicalpotential
of early Mars.
Insituinvestigationofhundredsofmetresofsedimentarystrataby
theMarsScienceLaboratoryCuriosityroverhasdocumentedancient
aqueoussurfaceenvironmentsfromfluvio-lacustrine17
tomoreinter-
mittentlakeorlake-marginsettings18,19
.Afteryearsofexploringstrata
dominatedbysmectite-bearingmudstonesinthelowerportionofthe
stratigraphic succession, the rover arrived at a sulfate-bearing unit,
marking a major environmental transition20
that is characteristic of
stratifiedterrainsacrossMars21
.Heredatacollectedbytheroverhave
uncovered a type of sulfate-enriched evaporitic-clastic deposit.
Pervasive centimetre-scale polygonal patterns in the basal
sulfate-bearing stratigraphic unit manifest as straight ridges that
intersect with triple junctions. The most prominent occurrence was
observed on the 3,154th mission sol (Fig. 1 and Extended Data Fig. 1).
Several additional occurrences were observed nearby within an 18-m
elevation interval, and show comparable, as well as incipient and
altered, variants of these patterns (Extended Data Fig. 3). The poly-
gons persist vertically to at least decimetric depth as shown by their
stepped appearance on thick blocks of bedrock (Fig. 1b). On bedding
planes, these polygons show approximately 1-cm relief and an aver-
agediameterofabout4 cm(varyingfrom1 cmto7 cm),withjunction
anglesclusteringat120°(ExtendedDataFig.4).Theridgescommonly
consistofalignednodules,variablyjuxtaposed,andirregularinshape
andsize(Fig.1d).ChemicalcompositiondocumentedbytheChemCam
instrumentshowsasignificantincreaseofcalcium(Ca)-sulfateandvari-
ablemagnesium(Mg)-sulfateenrichmentwithinthepolygonalridges
and other nodular bedrock, whereas the smooth host bedrock in the
polygoncoreshasbasalticbulkcompositionwithsporadicCa-sulfate
detections but is dominantly sulfate poor (Fig. 2 and Supplementary
Table 2).
AlthoughpolygonalridgesinevaporiticsettingsonEarthcanform
asaconsequenceofsubsurfacesalinityconvection22
,wedonotfavour
thisinterpretationhere.Suchterrestrialsaltcrustsaremostlypureand
consist of ephemeral salt deposits that form larger polygons 0.5 m to
2 m in size23
, 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
freshmudlayersinitiallyformT-junctions,maturationoverrepeated
https://doi.org/10.1038/s41586-023-06220-3
Received: 8 November 2022
Accepted: 15 May 2023
Published online: 9 August 2023
Check for updates
1
Institut de Recherche en Astrophysique et Planétologie, Université de Toulouse 3 Paul Sabatier, CNRS, CNES, Toulouse, France. 2
LGL-TPE, ENS de Lyon, Lyon, France. 3
Space Science Institute,
Boulder, CO, USA. 4
Indiana University, Bloomington, IN, USA. 5
University of Chicago, Chicago, IL, USA. 6
University of Tennessee, Knoxville, TN, USA. 7
University of New Brunswick, Fredericton,
NB, Canada. 8
Los Alamos National Laboratory, Los Alamos, NM, USA. ✉e-mail: william.rapin@irap.omp.eu

2. 300 | Nature | Vol 620 | 10 August 2023
Article
drying cycles results in hexagonal shapes with junction angles near
120°,thatis,Y-junctions24
.Inexperiments,usingclaylayers,jointangles
progressivelytendtowards120°after10consecutivedryingswithmore
cycles required to reach a homogeneous distribution centred at 120°
and mature patterns of hexagonal shapes25
.
Abundant sulfates in the ridges and nodular bedrock (30 wt% to
50 wt% Ca-sulfate and up to 40 wt% Mg-sulfate) and their much lower
abundance in the host bedrock (Fig. 2) collectively suggest that sul-
fate minerals precipitated owing to evaporation in muds and incor-
porated detrital sediment in the process. The present appearance of
thesulfate-bearingridgesisprobablynottheoriginalconfigurationof
thesefeatures.Moreplausibly,theystartedoutasevapoconcentration
deposits focused on initially formed cracks that then evolved over a
longerhistoryofdryingcyclesandburialdiagenesis(Fig.3).Theyare
nowexposedaserosion-resistantpolygonalridgesowingtotheirhigher
degree of cementation relative to the host bedrock, and an early bias
of surface salt precipitation in original mud-crack polygons (Fig. 3f).
Recurrentwettingofsurfacemudsprobablyreflectsacombination
offloodingandgroundwaterrecharge.Floodingcouldhavedissolved
saltsthatformedephemeralsurfacecrusts,aswellasaddingsediment
and promoting burial, although air-fall is another possible sediment
source.Thelengthofsurfacecracksforahomogeneousmediumscales
toabout1.2to1.3timesitscontractiondepth25
.Thus,thepolygonsize
(afewcentimetres;Fig.1fandExtendedDataFig.4),impliesdesiccation
and recharge events that affected only the uppermost few centime-
tres of the original muds. Deeper persistence of the polygonal pat-
tern within bedrock (Fig. 1b) indicates upwards propagation over the
long term as the surface aggraded sediment (Fig. 3). The distribution
of junction angles well centred at 120° and the absence of larger size
lower-order polygons (Extended Data Fig. 4) suggests shallow-depth
repeated cycles of regular intensity. The variability in shape and size
ofthesulfate-richnodulesthatprecipitatedwithinthesecracks(now
formingridges)indicatesmultiplegenerationsofnodulegrowth,con-
gruous with substantial intrasediment salinity fluctuations owing to
repeated drying cycles.
Given the apparent repetition and limited desiccation depth, a few
centimetres,wet–drycyclescouldhavebeenseasonalbutpotentially
may also have occurred on shorter timescales. The time span over
which these wet–dry cycles imprinted their signature is uncertain
a
c
c
e
e
b
b
f
d
d
20 cm
5 cm 5 mm 10 cm
Fig. 1 | In situ observations of polygonal ridges.a,Generalviewofbedrock
surroundingtheroveronsols3,154to3,156showingwidespreadpolygonal
ridges(mcam100270;largerimageinExtendedDataFig.2).b,Close-up
showing‘stepped’exposureofpolygonswithinlargebedrockblocks.c,View
ofbedrock(mcam100287)withpolygonsandlocationsofChemCamanalysis
onridge(redrectangle)andAPXSanalysisonsmoothhostbedrock(dotted
circle). d,Remotemicro-imageofcementedridgewithspotsanalysedby
ChemCam(reticles1to5),highlightingdetailsofnodulartexture(ccam01156).
e,f,Bedrockwithpolygonalpattern(e)andinterpretativeoverlay(f)thatshows
prominentridges(solidredlines),lesscertainridges(dottedredlines)and
cross-cuttinglater-stageCa-sulfate-filledveins(whiteareas)(mcam100276).
0
0.005
0.010
0.015
S
signal
(normalized)
CaO + MgO (wt%)
Mg/(Ca + Mg)
30% Ca-sulfate
40% Mg-sulfate
Ca-sulfate
(bassanite)
Mg-sulfate
(amorphous 8 wt% H2O)
0 1
0 10 20 30 40
15 25 35
5
Host bedrock
Nodular bedrock
Polygonal ridges
Fig. 2 | Bedrock enrichments in Ca-sulfate and Mg-sulfate.ChemCamsulfur
signalasafunctionofCaOandMgOforsmoothhostbedrock(blue)and
nodularbedrockincludingpolygonalridges(red-purple).Thedottedlines
correspondtomixingofsulfateswithaveragesiliciclastichostbedrock;
end-membersareshownbyblackcrossesandthebluecrossindicatesthehost
bedrockmeancomposition.ThesulfursignallimitofdetectionatSO3
>10.6 wt%isalsoshown(horizontaldashedline).Theverticalerrorbarson
end-membersrepresentsulfatecontentcalibrationuncertainty.

3. Nature | Vol 620 | 10 August 2023 | 301
but may be estimated by considering the polygons’ vertical distribu-
tion. Polygonal ridges are identified at multiple locations within an
18-m-thick stratigraphic interval (Extended Data Fig. 3). Using flood-
plainandlake-marginterrestrialanaloguesforasedimentationrateof
0.01 mm yr−1
to10 mm yr−1
(refs.26–28),thousandstomillionsofyears
maybeindicatedbythissequence,withperhapspunctuatedorspatially
variableepisodesofpersistentwet–drycycling.Thepolygonalpattern
probably propagated continuously through a stratigraphic interval
>2 mthick(ExtendedDataFig.4b)withoutdepositionaleventsproduc-
ingvisiblehorizons.Thissuggeststhatthesedimentationrateremained
uniform and low enough for the pattern to propagate vertically with
only thin clastic inputs during wettings, and without substantial
erosion during dryings (Fig. 3).
Insummary,twokeyobservationsare(1)maturehexagonalshapes
thatindicaterepeateddryingcycles,and(2)theirexistenceacrossstrati-
graphic thickness that implies that regular wet–dry conditions were
maintainedatleastepisodicallyinthelongterm.Mudcrackswereobser­
vedinunderlyingstrataoftheMurrayFormation,butthesewerepre-
dominantly T-junctions within a single horizon that suggest a single
desiccationevent18,29
.Thepolygonalfeaturesreportedhereareinstead
tangible evidence for sustained wet–dry cycling in early Martian sur-
face environments, strengthening the case for regular episodicity in
earlyMars’hydrologyonshorttimescales,probablyseasonaloreven
shorter. This finding agrees with models14,15
that rule out monotoni-
cally declining water supply in the aftermath of an asteroid impact or
a single volcanic eruption to explain early fluvio-lacustrine activity5
,
and instead favours a more sustained, Earth-like, wet climate regime,
with seasonal or shorter-term flooding. The presented evidence for
subaerialdesiccationindirectassociationwithsaltdepositsalsoadds
to the collective lines of evidence for evaporite deposition within the
fluvio-lacustrinesystemsofearlyMars18,19
.Althoughsulfatescanform
in ultracold climates30
, questioning the link between ancient sulfate
salts and habitability on early Mars31
, our observations at Gale Crater
instead suggest surface temperatures associated with sulfate forma-
tion warm enough for liquid water.
Inaddition,environmentssubjecttowet–drycyclingareconsidered
supportiveof,andperhapsessential,forprebioticchemicalevolution7,8
.
Owing to desiccation, water activity is lowered and the concentra-
tionofsolubleingredientsintheresidualliquidisincreased,boosting
reactionrates,especiallyforhigher-orderreactions7
.Forinstance,the
reactions that form nucleotides from their constituent nucleobases
(ribose and phosphate) produce water and hence are favoured at low
wateractivity.Mostimportantly,thepolymerizationreactionsneces-
sarytoadvancefromnucleotidestoRNAorDNA,andfromaminoacids
to proteins, require dehydration steps that have been demonstrated
to be facilitated by wet–dry cycles32–34
. Also, dioctahedral smectites,
that is, swelling clays, similar to those present at Gale Crater35
, and
ubiquitousacrossMarswithinenvironmentsinterpretedashabitable36
,
arecapableoftightlyadsorbingnucleotidesthroughcationexchange
onbasalsurfacesandhavethusbeenproposedtohelptheconcentra-
tionandpolymerizationoforganics37
.Undertherightenvironmental
conditions, Darwin’s proverbial ‘warm little pond’ could promote the
reactionsformacromoleculepolymerization,andthroughsustained
wet–drycyclingincreasethelikelihoodofchemicalevolutiontowards
the origination of life38
.
Inabroaderregionalcontext,thereiswidespreaddocumentationof
preserved organics in Gale Crater sediments, containing up to about
0.5 kg m−3
, and a variety of other soluble elements2
. The addition of
direct evidence for a series of repeated wet–dry cycles presented
here supports the conclusion that conditions in ancient Gale Crater
were conducive to prebiotic polymerization processes. There is also
potentialforwet–drycyclingtohaveoccurredmorebroadlyonMars
intheperiodwhenbothintrabasinsulfatesaltsandsmectiteclayswere
deposited1
asGaleCraterisastratigraphicsectionofglobalsignificance
forthismineralassemblageneartheNoachian–Hesperiantransition20,39
(Fig.4).Someofthesegloballydistributedstratamaythusharbourwell
preservedevidenceofprebioticchemicalevolution,arecordthatisno
longer available on Earth40
. On the basis of this evidence for wet–dry
cycling within surface 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
a b c d e f
Recharge Flooding Burial Lithification Exhumation
Desiccation
Fig.3|Formationmodelforsulfate-enrichedpolygonalridges.a–c,Repeated
cyclesofdesiccation(a),recharge(b)andflooding(c)formaverticallypropagating
hexagonalpatternofsulfateenrichment.a,Evaporation(greyarrows)desiccates
andcracksnear-surfacesediment,triggeringsaltcrystallization(red)atandnear
crackswherethesubsurfacebrine(purple)concentrates.b,Waterrechargeheals
cracksbysedimenthydration.c,Floodingdissolvesexcesssaltsatthesurface
butsubsurfacebrineandintrasedimentsulfatesaltsarepreservedandsiliciclastic
sedimentisdepositedontop.d,Sedimentisburiedwithsaturatedbrineinpore
spacesandsulfatesaremostlypreserved.e,Laterdiagenesispartiallydissolves
intrasedimentsulfatesaltsandlatediageneticfracturesarefilledwithCa-sulfate
(white).f,Sulfate-cementedpolygonalridgesbecomevisibleduringexhumation
asthesofterhostbedrockispreferentiallyremovedduringweathering.
3.0 2.0
4.0
Time (Gyr) (estimated from cratering record)
?
?
?
?
Smectite clays
Exogenic material flux
Intrabasin sulfates
Subaerial wet–dry cycling (this study)
Gale Crater
Amazonian
Hesperian
Noachian
Pre-
Noachian
Valley networks, fans, deltas
Fig. 4 | Mars’s potential for prebiotic record.Evidenceforsubaerialwet–dry
cyclingcouldberelevanttotheearlyNoachian–Hesperiantransitionperiod
duringwhichbothintrabasinsulfatesandsmectiteclaysformed1
.Thetiming
foraGaleCraterimpact(diamond)withuncertaintycoversthedifferent
reportedages41,42
.Exogenicmaterialfluxisaproxyforthedeliveryoforganics
tothesurface(ExtendedDataFig.5).Thetimingforvalleynetworks,fansand
deltasisderivedfromcratercountingongeomorphicfeatures3
.

4. 302 | Nature | Vol 620 | 10 August 2023
Article
Noachian–Hesperiantransitionperiodcouldhavebeenfavourablefor
theemergenceoflife—possiblymoresothantheearlierNoachianeon
with its potential for perennially wet surface environments21
.
Online content
Anymethods,additionalreferences,NaturePortfolioreportingsumma-
ries,sourcedata,extendeddata,supplementaryinformation,acknowl-
edgements, peer review information; details of author contributions
andcompetinginterests;andstatementsofdataandcodeavailability
are available at https://doi.org/10.1038/s41586-023-06220-3.
1. Ehlmann, B. L. & Edwards, C. S. Mineralogy of the Martian surface. Annu. Rev. Earth Planet.
Sci. 42, 291–315 (2014).
2. Vasavada, A. R. Mission overview and scientific contributions from the Mars Science
Laboratory Curiosity rover after eight years of surface operations. Space Sci. Rev. 218, 14
(2022).
3. Kite, E. S. Geologic constraints on early Mars climate. Space Sci. Rev. 215, 10 (2019).
4. Sheldon, N. D. & Tabor, N. J. Quantitative paleoenvironmental and paleoclimatic
reconstruction using paleosols. Earth Sci. Rev. 95, 1–52 (2009).
5. Wordsworth, R. The climate of early Mars. Annu. Rev. Earth Planet. Sci. 44, 381–408
(2016).
6. Ramirez, R. M. & Craddock, R. A. The geological and climatological case for a warmer and
wetter early Mars. Nat. Geosci. 11, 230–237 (2018).
7. Campbell, T. D. et al. Prebiotic condensation through wet–dry cycling regulated by
deliquescence. Nat. Commun. 10, 4508 (2019).
8. Becker, S. et al. Wet–dry cycles enable the parallel origin of canonical and non-canonical
nucleosides by continuous synthesis. Nat. Commun. 9, 163 (2018).
9. Farley, K. A. et al. In situ radiometric and exposure age dating of the Martian surface.
Science 343, 1247166 (2014).
10. Goodwin, A., Garwood, R. J. & Tartèse, R. A review of the “Black Beauty” Martian regolith
breccia and its Martian habitability record. Astrobiology 22, 755–767 (2022).
11. Guzewich, S. D. et al. 3D simulations of the early Martian hydrological cycle mediated by
a H2–CO2 greenhouse. J. Geophys. Res. Planets 126, e2021JE006825 (2021).
12. Kamada, A., Kuroda, T., Kasaba, Y., Terada, N. & Nakagawa, H. Global climate and river
transport simulations of early Mars around the Noachian and Hesperian boundary. Icarus
368, 114618 (2021).
13. Kite, E. S., Steele, L. J., Mischna, M. A. & Richardson, M. I. Warm early Mars surface
enabled by high-altitude water ice clouds. Proc. Natl Acad. Sci. USA 118, e2101959118
(2021).
14. Turbet, M. & Forget, F. 3-D Global modelling of the early martian climate under a dense
CO2 + H2 atmosphere and for a wide range of surface water inventories. Preprint at https://
arxiv.org/abs/2103.10301 (2021).
15. Steakley, K., Murphy, J., Kahre, M., Haberle, R. & Kling, A. Testing the impact heating
hypothesis for early Mars with a 3-D global climate model. Icarus 330, 169–188 (2019).
16. Stucky de Quay, G., Goudge, T. A., Kite, E. S., Fassett, C. I. & Guzewich, S. D. Limits on
runoff episode duration for early Mars: integrating lake hydrology and climate models.
Geophys. Res. Lett. 48, e2021GL093523 (2021).
17. Grotzinger, J. P. et al. Deposition, exhumation, and paleoclimate of an ancient lake
deposit, Gale Crater, Mars. Science 350, aac7575 (2015).
18. Rapin, W. et al. An interval of high salinity in ancient Gale Crater lake on Mars. Nat. Geosci.
12, 889–895 (2019).
19. Schieber, J. et al. Mars is a mirror—understanding the Pahrump Hills mudstones from a
perspective of Earth analogues. Sedimentology 69, 2371–2435 (2022).
20. Milliken, R. E., Grotzinger, J. P. & Thomson, B. J. Paleoclimate of Mars as captured by the
stratigraphic record in Gale Crater. Geophys. Res. Lett. 37, L04201 (2010).
21. Bibring, J.-P. et al. Global mineralogical and aqueous mars history derived from OMEGA/
Mars Express data. Science 312, 400–404 (2006).
22. Lasser, J., Nield, J. M. & Goehring, L. Surface and subsurface characterisation of salt pans
expressing polygonal patterns. Earth Syst. Sci. Data 12, 2881–2898 (2020).
23. Goodall, T. M., North, C. P. & Glennie, K. W. Surface and subsurface sedimentary
structures produced by salt crusts. Sedimentology 47, 99–118 (2000).
24. Goehring, L., Conroy, R., Akhter, A., J. Clegg, W. & Routh, A. F. Evolution of mud-crack
patterns during repeated drying cycles. Soft Matter 6, 3562–3567 (2010).
25. Goehring, L. Evolving fracture patterns: columnar joints, mud cracks and polygonal
terrain. Phil. Trans. R. Soc. Math. Phys. Eng. Sci. 371, 20120353 (2013).
26. Sadler, P. M. Sediment accumulation rates and the completeness of stratigraphic
sections. J. Geol. 89, 569–584 (1981).
27. Daniels, J. M. Floodplain aggradation and pedogenesis in a semiarid environment.
Geomorphology 56, 225–242 (2003).
28. Kraus, M. J. Paleosols in clastic sedimentary rocks: their geologic applications. Earth Sci.
Rev. 47, 41–70 (1999).
29. Stein, N. et al. Desiccation cracks provide evidence of lake drying on Mars, Sutton Island
member, Murray Formation, Gale Crater. Geology 46, 515–518 (2018).
30. Baccolo, G. et al. Jarosite formation in deep Antarctic ice provides a window into acidic,
water-limited weathering on Mars. Nat. Commun. 12, 436 (2021).
31. Niles, P. B. & Michalski, J. Meridiani Planum sediments on Mars formed through
weathering in massive ice deposits. Nat. Geosci. 2, 215–220 (2009).
32. Becker, S. et al. Unified prebiotically plausible synthesis of pyrimidine and purine RNA
ribonucleotides. Science 366, 76–82 (2019).
33. Higgs, P. G. The effect of limited diffusion and wet–dry cycling on reversible polymerization
reactions: implications for prebiotic synthesis of nucleic acids. Life 6, 24 (2016).
34. Ross, D. S. & Deamer, D. Dry/wet cycling and the thermodynamics and kinetics of
prebiotic polymer synthesis. Life 6, 28 (2016).
35. Bristow, T. F. et al. Clay mineral diversity and abundance in sedimentary rocks of Gale
Crater, Mars. Sci. Adv. 4, eaar3330 (2018).
36. Bishop, J. L. et al. What the ancient phyllosilicates at Mawrth Vallis can tell us about
possible habitability on early Mars. Planet. Space Sci. 86, 130–149 (2013).
37. Pedreira-Segade, U., Feuillie, C., Pelletier, M., Michot, L. J. & Daniel, I. Adsorption of
nucleotides onto ferromagnesian phyllosilicates: significance for the origin of life.
Geochim. Cosmochim. Acta 176, 81–95 (2016).
38. Clark, B. C. & Kolb, V. M. Macrobiont: cradle for the origin of life and creation of a
biosphere. Life 10, 278 (2020).
39. Grotzinger, J. P. & Milliken, R. E. in Sedimentary Geology of Mars Vol. 102, (SEPM Society
for Sedimentary Geology) 1–48 (2012).
40. Knoll, A. H. Paleobiological perspectives on early microbial evolution. Cold Spring Harb.
Perspect. Biol. 7, a018093 (2015).
41. Thomson, B. J. et al. Constraints on the origin and evolution of the layered mound in
Gale Crater, Mars using Mars Reconnaissance Orbiter data. Icarus 214, 413–432 (2011).
42. Le Deit, L. et al. Sequence of infilling events in Gale Crater, Mars: results from
morphology, stratigraphy, and mineralogy. J. Geophys. Res. Planets 118, 2439–2473
(2013).
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this
article under a publishing agreement with the author(s) or other rightsholder(s); author
self-archiving of the accepted manuscript version of this article is solely governed by the
terms of such publishing agreement and applicable law.
© The Author(s), under exclusive licence to Springer Nature Limited 2023

5. Methods
Rover-basedgeochemicaldatasetsanddataprocessing
The Curiosity rover has five instruments that can measure geochem-
istry and mineralogy43
. ChemCam is the primary source of available
chemicaldata.Itisalaser-inducedbreakdownspectrometerthatpro-
vides chemical analyses at a submillimetre scale and detailed images
with the Remote Micro Imager (RMI)44,45
. Major-element contents
were obtained using the current calibration model46
. Water and sul-
fatecontentswereestimatedusingdedicatedcalibrationmodels:for
watercontent,quantificationusesthehydrogensignal47
,andforsulfate
content, the sulfur signal is used with in situ calibration as previously
described for sulfate enrichments observed elsewhere in the Murray
Formation18
. Although the arm-mounted Alpha Particle X-ray Spec-
trometer (APXS) instrument did not sample polygonal ridges owing
to operational constraints, it provided a bulk (about 2-cm-diameter
area)measurementonbrushed,smoothedhostbedrockimmediately
adjacenttoaraised,resistantridge(Fig.1c),highlightingtheabsence
ofsulfateenrichment(SupplementaryTable2).APXSdatafromother
bedrock targets and resistant features in the vicinity of the polygonal
features also support the concentration of Mg- and Ca-sulfates in the
resistant features (Extended Data Fig. 6). Chemistry and Mineralogy
(CheMin)andSampleAnalysisatMars(SAM)analyses(whichprovide
mineralogy and evolved gas analysis, respectively, from powdered,
drilled sample) were not acquired on the polygonal ridges and most
of the nodular bedrock.
Identificationofthenodularlithology
ThenodularlithologywasidentifiedbasedonimagesusingbothMast-
CamandChemCamRMI.Thelithologyischaracterizedbyprotruding,
erosion-resistantformsofnear-centimetrescaleconnectedtothehost
bedrock. They are variably shaped, widespread and abundant in the
examinedsection(ExtendedDataFig.3).TheChemCamRMIwasused
to associate point by point the laser-induced breakdown spectrom-
eter measurements with either smooth host bedrock (for example,
ExtendedDataFig.7)ornodularbedrock(forexample,ExtendedData
Fig.8).ThiscarefulclassificationofallChemCamdataacquiredwithin
theexaminedsectionprovidesanexhaustivelistofgeochemicaldata
fromthebedrock,removingpointswithsoilcontributionorpoorlaser
focus. The full list of points classified from either host or nodular is
shown in Supplementary Table 3.
Lithologyofsulfate-bearingnodularbedrock
Nodules are widespread, highly variable in density in the examined
section(ExtendedDataFig.3),andoccurasthreemorphologictypes:
rounded,dendriticandvacuolar.Roundedformsareupto1 cmindiam-
eter and more resistant to erosion than the host bedrock. Nodules
can be spherical to subspherical, and can show variable clustering
and coalescence to form incipient to well developed nodular ridges
(ExtendedDataFig.9a,b).Dendriticnodulesarejagged,multifaceted,
multicentimetreformsthatareevenlydistributedonbeddingplanes
at certain intervals, and resemble rosette-like aggregates (Extended
DataFig.7a,b,d).Thevacuolarnodulesarelarge(uptodecimetrescale),
polymorphicformsthatapparentlyresultedfromrandomamalgama-
tionofnodules.Theircoalescenceispartial.Nowvisiblevoidsprobably
representspotsonceoccupiedbyuncementedhostbedrockthathas
been removed by aeolian erosion processes48
. Vacuolar nodules are
pervasive at the top of the studied interval (Extended Data Fig. 11).
Outcropswithlaminaeconsistingofvariablycoalescentmicro-nodules
were also observed within the section, although not analysed by the
rover chemical instruments (Extended Data Fig. 9e).
Polygonalpatternanalysis
To generate the map over which the polygonal pattern was analysed
(ExtendedDataFig.4a),weusedtheOnsightsoftware(https://software.
nasa.gov/software/NPO-50830-1), which was developed by the Jet
Propulsion Laboratory for the context of Mars exploration. Onsight
renders a three-dimensional reconstruction and visualization of the
surface of Mars from a variety of images. For our purpose, it allowed
orthoprojectiontothesurface,withorthoprojectedscale,oftheMast-
Cam mosaics. Most of the polygon morphologies are hexagonal, and
resemble honeycomb or hexagon floor tiles. Pentagonal forms are
common. Rare quadrangle and heptagonal patterns have also been
observed.
The size of the polygons is approximated by the diameter of the
circles that are enclosed within the polygons, that is, circles that are
tangential to polygon sides. The circles were manually drawn, and
diameter of the individual polygons was measured using Adobe Illus-
tratorsoftwareappliedtotheOnsightorthoprojectionofthemosaics
acquired on sols 3,152–3,154 (west section) (Extended Data Fig. 4a).
Thepolygonsthatwereselectedforanalysisarethosethatofferafairly
regular shape, that is, the polygons that have equal to subequal side
lengths.Theunmarkedareasarethoseforwhichlatediageneticveins
intersect and disturb the original polygonal ridge pattern, and those
for which the original polygonal ridges have been stripped off after
exhumation by dissolution and/or deflation.
Theaveragediameterofthe467circlesthatwereanalysedbyasingle
operatoroverasurfaceofabout3.75 m2
is3.9 cm,withthesmallestand
largestspecimenat1.73 cmand7.56 cm,respectively(Supplementary
Table 1). To have a notion of the degree of precision of our method, a
second operator made a test by repeating the marking and measure-
ments over block 1 (Extended Data Fig. 4a). The test revealed that: (1)
thenumberofpolygonsdistinguishedbytheoperatorsisminimal,that
is,168against176;and(2)themeansizevariesfrom3.43 cmto3.98 cm,
which is a relative error of only about 16%, at best.
The frequency distribution of diameter centimetre values shows
an unimodal asymmetrical, that is, a Poisson probability distribution
(Extended Data Fig. 4c). Distinct blocks of similar areas yield similar
polygonal pattern’s characteristics (Supplementary Table 1), which
suggeststhepolygon-bearingbedrockoriginallywashomogeneousin
termsoftexture(porosity;grainsize)andmineralogicalcomposition,
laterally at multimetric scale.
ThejunctionanglesweremeasuredontheOnsightorthoprojection
within a larger area to extend the number of samples. Junctions and
ridgesbetweenjunctionsweremappedusingQGISsoftwarefromwhich
angleswerecomputed.Forajunctionwithonlytwoconnectedridges
mapped, only one angle is measured. For triple junctions, with three
connectedridgesidentified,the3anglesarecomputedformingatotal
of360°.Wellexpressedandclearlyidentifiedridgesaredifferentiated
fromridgeswithmorealteredanduncertainfortest,withatotal2,214
angles identified total, 532 of them classified as angles only for well
expressed ridges. The frequency distributions show a similar modal
angle near 120°, then a Gaussian fit was performed on all identified
angles and found centred at 119.1° (Extended Data Fig. 4c).
Data availability
All data are available in the NASA Planetary Data System (https://pds.
nasa.gov/)GeoscienceNodeintheMarsScienceLaboratorydirectory
(https://pds-geosciences.wustl.edu/missions/msl/)orSupplementary
Information.
43. Grotzinger, J. P. et al. Mars Science Laboratory Mission and science investigation. Space
Sci. Rev. 170, 5–56 (2012).
44. Maurice, S. et al. The ChemCam instrument suite on the Mars Science Laboratory (MSL)
rover: science objectives and mast unit description. Space Sci. Rev. 170, 95–166 (2012).
45. Wiens, R. C. et al. The ChemCam instrument suite on the Mars Science Laboratory (MSL)
rover: body unit and combined system tests. Space Sci. Rev. 170, 167–227 (2012).
46. Clegg, S. M. et al. Recalibration of the Mars Science Laboratory ChemCam instrument
with an expanded geochemical database. Spectrochim. Acta Part B 129, 64–85 (2017).
47. Rapin, W. et al. Quantification of water content by laser induced breakdown spectroscopy
on Mars. Spectrochim. Acta Part B 130, 82–100 (2017).

6. Article
48. Schieber, J. et al. Engraved on the rocks—Aeolian abrasion of Martian mudstone
exposures and their relationship to modern wind patterns in Gale Crater, Mars.
Depositional Rec. 6, 625–647 (2020).
49. Hartmann, W. K. & Neukum, G. Cratering chronology and the evolution of Mars. Space
Sci. Rev. 96, 165–194 (2001).
50. Quantin-Nataf, C., Craddock, R. A., Dubuffet, F., Lozac’h, L. & Martinot, M. Decline of
crater obliteration rates during early Martian history. Icarus 317, 427–433 (2019).
Acknowledgements We thank A. Vasavada and A. Fraeman for discussions; and T. Goudge and
J. Bishop for the comments and review. The data used are available in the NASA Planetary Data
System Geoscience Node in the Mars Science Laboratory directory (https://pds-geosciences.
wustl.edu/missions/msl/). This project was supported in the United States by NASA’s Mars
Exploration Program and in France is conducted under the authority of CNES. Mastcam
mosaics were processed by the Mastcam team at Malin Space Science Systems. E.S.K. funding
by NASA grant 80NSSC22K0731. L.M.T. funding as a Mars Science Laboratory team member is
provided by the CSA.
Author contributions W.R. and G.D. equally led the writing of the paper. W.R., G.D., B.C.C.,
J.S., L.C.K., L.M.T., P.J.G., P.-Y.M. and J.L. contributed to methodology, investigation and data
processing. O.G. and N.L.L. are the leads of the ChemCam instrument investigation. J.S. and
E.S.K. provided significant contributions to the writing and reviewing of the paper. All
co-authors provided helpful comments and inputs to the paper.
Competing interests The authors declare no competing interests.
Additional information
Supplementary information The online version contains supplementary material available at
https://doi.org/10.1038/s41586-023-06220-3.
Correspondence and requests for materials should be addressed to W. Rapin.
Peer review information Nature thanks Janice Bishop and Timothy Goudge for their
contribution to the peer review of this work. Peer reviewer reports are available.
Reprints and permissions information is available at http://www.nature.com/reprints.

7. Murray
fm.
Carolyn
Sh. fm.
Sulfate
-bearing
unit
Yardangs
100 m
-4460 m
Sulfate-bearing
unit
Yardangs
Mt. Sharp
Murray fm.
N
1 km
Extended Data Fig. 1 | Context of observations in Gale crater, Mars.
Stratigraphiccontext(left)ofthelowerportionofMountSharpandmap
(right)showingCuriosityrovertraverse(white)ontheHighResolutionImaging
ScienceExperiment(HiRISE)basemapoverlaidwithCompactReconnaissance
ImagingSpectrometerforMars(CRISM)S-index,whichtrackssulfates(shaded
yellow).Redrectangleshowsthelocationofclose-upmapanddetailed
stratigraphy(ExtendedDataFig.3b).

8. Article
Fig. 1e
Fig. 1c
Fig. 1b
Fig. 1a
10 cm
30 cm
30 cm
50 cm
c
a
a
c
b
b
Extended Data Fig. 2 | Larger color image of bedrock with polygonal ridges
for context.MastCamimage(mcam100270)andclose-ups(a,bandc)with
rectanglelocationsofclose-upviewfromFig.1.Close-ups(b,c)showbedrock
10to20metersawaywhereregularlyspacedridgesandnodulescanbe
observedsupportinglateralextensionofthesamepolygonalpatternalthough
cameraresolutionpreventsdetailedgeometricalanalysisatthisdistance.

9. - 4,070 m
- 4,060 m
East
Section
200 m 220 m
Medium
Section
- 4,040 m
- 4,050 m
CCAM
obs.
CCAM
obs.
CCAM
obs.
West
Section
MAHLI
APXS
o o
ox ox
ox -
MAHLI
APXS
o -
ox ox
ox ox
MAHLI
APXS
ox ox
o -
o -
- -
o -
x -
x -
x -
x -
xo xo
Murray
fm.
Carolyn
Sh. fm.
Sulfate
-bearing
unit
Yardangs
Bradbury
group
marker bed
-3860 m
-3350 m
100 m
-4460 m
East Section
West Section
Medium Section
N
Basal sulfate-bearing unit
Mt Mercou
100 m
a
b
Sandstone
Faintly laminated mudst.
Laminated mudst. with minor sandstone
Heterolithic mudstone-sandstone Host bedrock
Nodular bedrock Polygonal ridges
18 m
Extended Data Fig. 3 | Local stratigraphic context of the examined section.
Close-upmapoftheexaminedsectionsalongtherovertraversewithlocation
oftheobservedpolygonalpattern(redcircles)andassociatedgeochemical
measurements(redtriangleanddiamond,seeFigs.2,3)(a);aswellasgeneral
stratigraphiccolumnforcontextwithdetailedlogfaciesforEast(Extended
DataFig.9),Medium(ExtendedDataFig.10),andWest(ExtendedDataFig.11)
sections(b).Otherlocationswithpossiblepolygonalridgesasincipientor
alteredvariantsareannotatedinExtendedDataFig.12.Theapproximate
distanceof200 mand220 mbetweentheadjacentsectionscentersis
annotatedontop(b).Faciesaredrawnbasedonavailableimagesfrom
MastcamM34/M100,MAHLIandChemCamRMI.Nodularfeatureswithin
bedrockinterpretedaschemicaldepositsarehighlightedinred,whereas
physicalsedimentarystructuresareshowninblue.Sedimentarystructures
wererarelyobservedandthehostrockofchemicaldepositswasmostly
smoothandfeatureless.Geochemicaldatafrompolygonalridgesare
representedwithtriangle(aandb).LocationofChemCamLIBSobservations
onbedrockareshowninthe“CCAMobs.”columnwitharedcrosswherepoints
analyzedchemicaldepositsandbluedotforhostrock.LocationsofAPXSand
MAHLIobservationsarealsohighlightedthesameway,withacrossfor
chemicaldepositsandcircleforhostbedrock.

10. Article
4.0 5.0 6.0
3.0
2.0 cm
7.0
N = 477
N = 532
N = 2214
Average
3.9
cm
Average
119.1°
Frequency
(normalized)
Frequency
(normalized)
Diameter (cm)
d
c
b
a
0 20 40 60 80 100 120 140 160 180
Junction angle (°)
20 cm
Rover blocking
foreground
Block 2
Blocks 1-1bis
Block 3
N
a
Ext. Data
Fig. 2b
Ext. Data
Fig. 2c
+1 m 0 m
5 m
-1 m -2 m
Rover
N
Extended Data Fig. 4 | Polygonal ridges pattern analysis.Topviewof
bedrockfromsol3154generatedusingOnsight,withcirclesrepresenting
markedpolygonsizes(blue)andjunctions(yellowdots)connectedto
well-expressed(solidlines)anduncertainridges(dashedlines)formingthe
patternmappedonimage(a).Redcontourrepresentstheareacoveredon
Fig.1e,andwhiterectanglesareasforpolygonsizestatistics(Supplementary
informationTable1).Topviewwithextendedcontextshowstheareaofanalysis
alongwiththelocationofcloseupswherepolygonalridgesareobservedinthe
distance(b).Elevationchangeisindicatedby1 mcontourlines(yellow)to
estimatethethicknessoftheobservablestratafromthislocation.Distribution
ofjunctionanglesforwell-expressedridgesonly(solidred)andallridges
(dashedred)withgaussianfit(c).Distributionofpolygonsizesandfitwith
Poissonprobabilitydistribution(d),seeSupplementaryinformationTable1.

11. 0.00001
0.0001
0.001
0.01
0.1
1
10
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
Cumulave
number
of
craters
D1km
Time (Gyr)
0.0001
0.001
0.01
0.1
1
10
Flux
of
impacts
craters
D1
km
Exogenic material flux
Amazonian
Hesperian
Noachian
Pre-N.
a
b
c
Extended Data Fig. 5 | Flux of impacts on Mars through time.Curves
representingtheimpactratechronologyonMarsbasedoncumulativedensity
ofcraterswithD1 kmperunitarea(a).Thismodelisbasedontheequation
fromthecraterchronology49,50
givingthecumulativenumberofcraterslarger
than1 kmasN(1)=5.44×10−14
[exp(6.93T)−1]+8.38×10−4
T.Thederivativeof
thisfunctiongivesthefluxofimpactcraterslargerthan1 kmovertime(b)is
givenbytheequation:N’(1)=3.77×10−13
[exp(6.93T)−1]+8.38×10−4
T.This
curvepresentsastrongdecreaseinintensityofcrateringwithtimewhichcan
beusedasaproxyforextra-terrestrialmaterialaccretionwithtime.Thelink
betweenthefluxofimpactcratersofagivensizeandtheaccretionfluxofthe
planetisthenshownasagradientbar(c)thatisrepresentedonFig.4inthe
maintext.

12. Article
0
5
10
15
20
25
30
0 10 20 30 40
CaO
(wt%)
SO3 (wt%)
Bedrock
Diagenec (with veins and coangs)
Diagenec (resistant nodules)
0
2
4
6
8
10
12
0 10 20 30 40
MgO
(wt%)
SO3 (wt%)
a
b
Extended Data Fig. 6 | APXS data bedrock and diagenetic features.CaO
(a)andMgO(b)versusSO3 contentforbedrockanddiageneticfeaturesinthe
investigatedsection.Resistantnodules,formingcollectivelynodularbedrock,
showMgOandSO3 enrichment.Thedataonlyincludesrelativelyclean,
dust-freetargets.Listoftargetnamesforbedrock:Gourdon,Bardou_DRT,
Ribagnac_DRT,Chenaud_DRT,Monpazier,Plaisance_DRT,Monsec_DRT.For
veins/coating:Terrasson_Lavilledieu,Festalemps_DRT,Quinsac,Biras,Pezuls.
Forresistantbedrock:Gardonne_DRT,Simeyrols,Rouffignac,Bosset,Bosset_
offset,Nabirat,Sarlande,Salagnac,Le_Bugue.

13. Extended Data Fig. 7 | ChemCam images of fine-grained host bedrock.Thebedrockmatrixtypicallycomposedofalight-coloredsmooth-texturedmudstone.

14. Article
Extended Data Fig. 8 | ChemCam images of nodular bedrock.Salt-bearingconcretionsarewidespreadandabundantintheexaminedsection.

15. - 4,070
- 4,060
3136
3120
3117
3113
East
Section
Sol of arrival
laminar bedded
nodular
laminar bedded
laminar bedded
thin planar band
thin planar band
laminar bedded
vein
laminar bedded
nodular
ridges
Incipient nodular ridges
- a
- b
- c
- d
- e
e
d
c
b
a
5 cm
5 cm 20 cm
50 cm
5 cm 20 cm
5 cm
Extended Data Fig. 9 | Stratigraphic log of East section.Lumpednodules
organizedinpolygonalridges(dottedlines)onbedrockblock(a,sol3137
Montaut).Lumpednodulesformingpossibleincipientridges(dottedlines)on
bedrock(b,sol3137Montignac).Irregularnodulesformingpossibleincipient
oralteredridges(dottedlines)observedwithinnodularbedrock(c,sol3117
drive_direction).Laminatedbedrockwithvariablycoalescentnodulartexture
(d,sol3119allas_les_mines).Aligned,variablycoalescentmicro-noduleswithin
laminarbeddedfacies(e,sol3112garreloup).

16. Article
- a
- b
- c
- d
d
c
b
a
3147
3145
3143
3138
3140
Medium
Section
dendritic nodules
Polygonal ridges
5 cm 20 cm
5 cm
30 cm
5 cm 30 cm
1 cm 20 cm
dendritic nodules
dendritic nodules
Extended Data Fig. 10 | Stratigraphic log of Medium section.Bedrockwith
evenlydistributeddendriticnodules(a,sol3147workspace).Close-upimage
ofadendriticnoduleshowingjagged,multifaceted,multi-centimetertexture
(b,MAHLItargetNabirat25 cmstandoff).Incipientpolygonalridges(dotted
lines)onbedrock(c,sol3139workspace).Dendriticnodules(d,MastCamon
Vayres).

17. - a
- b
- c
- d
- e
- Fig. 1
b
a
c
d
e
3156
3151
- 4,040
- 4,050
3149
3154
3158
3161
3163
West
Section nodular beds
with few nodules
5 cm 50 cm
10 cm 40 cm
5 cm 30 cm
50 cm 1 m
5 cm
30 cm
Incipient polygonal ridges
triple junctions
laminar bedded
vacuolar nodules
nodular beds
laminar bedded
coalescent nodules
Extended Data Fig. 11 | Stratigraphic log of West section.Partially coalescent
nodulesandlaminarbeddedfacies(a,sol3170Organized_nodules).Nodular
bedrockandlaminarbeddedfaciesadjascenttosmoothorlaminatedbedrock
at Pontourslocation(b,sol3163drill_area_context).Large,polymorphic
vacuolarnodules(c,sol3161workspace).Coalescentnodulesalignedinplanar
beds(d,sol3158diagenetic_transition).Incipientpolygonalridges(dotted
lines)onbedrock(e,sol3151workspace).

18. Article
Extended Data Fig. 9b
Extended Data Fig. 9a
5 cm
5 cm
Extended Data Fig. 11e
5 cm
Extended Data Fig. 10c
5 cm
Extended Data Fig. 9c
10 cm
Extended Data Fig. 12 | Polygonal pattern as incipient or altered variants.Detailsofimagesofotherlocationswithpossiblepolygonalridges(dottedlines)
fromExtendedDataFigs.9–11withmarkings(left)andwithout(right),seecorrespondingfiguresforcontext.