Life on Mars: Evidence from Martian Meteorites David S. McKay1 Kathie L. Thomas-Keprta2 Simon J. Clemett2 Everett K. Gibson1, Jr., Lauren Spencer,1 and Susan J. Wentworth2. 1NASA Johnson Space Center, 2101 NASA Parkway, Houston TX USA; 2Jacobs Technology, NASA Johnson Space Center, 2101 NASA Parkway, Houston TX USA 77058 ABSTRACT New data on martian meteorite 84001 as well as new experimental studies show that thermal or shock decomposition of carbonate, the leading alternative non-biologic explanation for the unusual nanophase magnetite found in this meteorite, cannot explain the chemistry of the actual martian magnetites. This leaves the biogenic explanation as the only remaining viable hypothesis for the origin of these unique magnetites. Additional data from two other martian meteorites show a suite of biomorphs which are nearly identical between meteorites recovered from two widely different terrestrial environments (Egyptian Nile bottomlands and Antarctic ice sheets). This similarity argues against terrestrial processes as the cause of these biomorphs and supports an origin on Mars for these features. Keywords: Mars meteorites, ALH84001, Nakhla, Y000593, Life on Mars INTRODUCTION Martian meteorite ALH84001 Our original publication on this subject presented a suite of characteristics closely related in space and time within martian meteorite ALH84001, all of which could be best explained by a hypothesis that they were formed by microbes early in martian history.1 These observations included the presence of chemically zoned carbonates precipitated from water in cracks openings in the meteorite, morphological forms similar to known terrestrial fossils, polycyclic aromatic hydrocarbons (PAHs) associated with the carbonates, and nanophase magnetites embedded within the carbonates. A single hypothesis that all of these features were formed or assisted by early martian microbes was proposed by us. Initially, this hypothesis was challenged on the basis that some of these features were terrestrial contamination, and that the carbonates formed at high temperature.2; 3 Detailed published studies by our group4 and others4-6successfully refuted each of these objections. One alternative hypothesis, that the carbonates formed by inorganic precipitation7; 8 was also shown to be incapable of explaining the chemistry and the morphology of the actual martian carbonates. 4Additionally, it was proposed that the nanophase magnetites were formed by thermal and shock heating of the iron-rich carbonate present in ALH84001 during its early martian history or during the impact that ejected it from Mars.9; 10 To summarize the current situation, most of the scientific community now accepts that nearly every ALH84001 feature that we discussed in the original Science paper actually formed on Mars including the carbonates and the magnetites. Formation of the carbonates by precipitation from water at moderate temperatures is also now generally accepted4. The Recent detailed analysis and thermodynamic treatment has now definitively ruled out the hypothesis that the nanophase magnetite within ALH84001 formed by thermal or shock decomposition of any of the associated carbonate11. ALH84001 Background McKay et al1 characterized a number of closely associated features in ALH84001 and suggested that this suite of features, taken together, could best be explained by a biogenic hypothesis in which early Martian microbes were directly or indirectly involved in producing the features. These features include: Carbonate globules or pancakes found in cracks and veins which we suggested were formed at relatively low temperatures and involved water--formation possibly assisted by microbial action Possible microfossils are present (biomorphs) Polycyclic Aromatic Hydrocarbons (PAHs) are present and closely associated with carbonates Nanophase magnetite (Fe O ) similar to magnetite produced by magnetotactic bacteria is present embedded in the 3 4 carbonates This suite of features taken together provided the possible evidence for the biogenic hypothesis. Note that no single feature was either definitive for biology, or conversely, that showing that no single feature was definitely produced by non-biogenic processes would invalidate the hypothesis, although it would clearly weaken it. A number of early objections to our hypothesis were made at meetings and in publications: The carbonate globules formed at high temperature by volcanic or impact processes on Mars The meteorite was contaminated in Antarctica Carbonates grew in Antarctica PAHs were deposited from Antarctic melt water Magnetites were from terrestrial sources such as wind-blown dust Magnetotactic bacteria would not develop on Mars because Mars had no magnetic field The “microfossils” were coating artifacts added during preparation for SEM studies The microfossils were too tiny to be real We addressed most of these objections in published rebuttals and papers. In addition, some of the earliest criticisms were essentially retracted by more careful analysis of the available data. Some3; 12 proposed that the carbonates formed at high temperatures. If true, this would rule out microbial involvement and seriously weaken our hypothesis. These papers were given wide play in the media and in some scientific meetings. The consensus began to develop that we had been discredited. However, Treiman and RomanekWarren 1998, and McSween and Harvey ,2000 5 published papers supporting a low temperature aqueous precipitation hypothesis for the carbonates. Note that McSween and Harvey did a complete reversal from their earlier interpretation of the carbonates.5 The majority of scientific papers published in the past 10 years now accept that carbonates formed on Mars at low temperature by precipitation from water. Terrestial contamination was another issue. However, detailed isotope studies proved that the carbonates formed on Mars, not in Antarctica. Careful search for PAHs in Antarctica melt waters near the collection site failed to find detectable PAHs casting doubt on whether the ALH84001 PAHs resulted from contamination13. Laboratory results showed that PAHs are relatively insoluable in water and would not be concentrated in carbonates. It is now generally accepted that the PAHs in this meteorite are Martian in origin. Similarly, if the carbonates are Martian, the embedded magnetites must also be Martian. No mechanism has been proposed to embed terrestrial magnetite in the Martian carbonates. One objection to the hypothesis that the Martian magnetites were produced by Martian magnetotactic bacteria is that Mars did not have a global magnetic field, thought to be a requirement for bacteria to develop magnetic inclusions within their cells. However, after the original paper was published, mapping of the Martian surface by orbiting spacecraft revealed that early crustal rocks did have strong remnant magnetism which could only be explained if Mars had an early strong magnetic field, now gone. While some researchers 14have supported our biogenic hypothesis for the origin of many of the magnetites, a number (Golden et. al, Treiman, Brearly, and others) 2; 8; 15; 16have proposed an alternative hypothesis that the magnetite was formed totally nonbiologically by thermal decomposition of the iron-rich carbonate during or following an impact shock event.7; 8; 15 This alternative non-biologic hypothesis has been used for the past decade as the primary argument against our hypothesis. In a recent paper, we have now addressed this alternative hypothesis in detail and have shown that it cannot explain the pure chemistry of the most of the magnetites and their lack of other cations such as Mn and Mg considering that surrounding carbonate is mostly a mixed carbonate containing Mg, Mn, and Ca as well as Fe virtually all laboratory studies have shown that thermal decomposition of a mixed carbonate produce a mixed composition spinel (magnetite), not the pure Fe magnetite common in the carbonates11. We conclude that the nanophase magnetites could not have been made by thermal or shock decomposition of the carbonate and therefore these magnetites had a separate origin not directly related to the carbonates. Our original hypothesis that they were produced by martian magnetotactic bacteria and introduced and trapped in the precipitating carbonate pancakes remains a viable explanation. The unique properties of these magnetites (elongated along the c-axis, single domain grain size, extremely pure Fe oxide, tightly sorted grain size distribution) remains a suite of properties absolutely unique to magnetotactic magnetites on Earth17. This suite of properties has been used for decades as certain biosignatures when found in terrestrial sediments or water. Application of this biosignature concept to these martian magnetites remains a viable and credible approach. BIOMORPHS IN ALH84001 Biomorphs are physical features and textures that resemble features known to be biogenic from terrestrial environments. Biomorphs are not proof of a biological origin, but can serve as pointers for further investigation such as chemistry, isotopic characterization, and mineralogy. Additional biomorphs have been found within ALH84001. Figure 1 shows a series of molds within the carbonate pancakes of this meteorite. These molds are lined with a lower Z lining. The uniform ~1 micrometer diameter of these molds is characteristic of molds of many kinds of coccoid bacteria cells where the impression of the cell is left in the matrix after the cell has decayed away. Figure 2 is a filament from ALH84001. Filaments and filament fragments of this size are common in modern microbial biofilms and are a product of microbes much larger than the filament. Figure 1. SEM view of carbonate pancake near the border with the host Orthopyroxene in ALH84001. These depressions are biomorphs suggestive of cavities left after embedded coccoid bacteria have partially or completely decayed away. The uniform ~1μm size of these depressions is typical of a biological origin. Figure 2. A filament from ALH84001. Filaments and filament fragments of this size are common in modern microbial biofilms and are a product of microbes much larger than the filament. This filament is clearly intergrown with the matrix and is unlikely to be random microbial contamination. NAKHLITE METEORITES A number of martian meteorites in museum collections all have similar mineralogy, chemistry and ages One group is named after the meteorite Nakhlite and this and similar meteorites are termed Nakhlites which are all about 1.3Gy old. Nakhla was seen to fall in Egypt June 28, 1911 and Yamato 000593 recovered recently by the Japanese Polar Expedition in Antarctica . Most of the Nakhlites that we have examined have biomorphic features, resembling known terrestrial microfossils. Many of these biomorphs appear to be integrally embedded in the iddingsite crack filling material of these meteorites. Iddingsite is a mixed phase including smectites and oxide minerals. In Nakhlites, iddingsite has a distinctive texture and distinctive chemical compositions. It is commonly accepted to be a low temperature martian aqueous alteration feature.18 All Nakhlites examined in detail show low-temperature aqueous secondary mineral alteration which occurred on Mars. Furthermore, all Nahlites examined in detail contain carbonates formed on Mars, usually closely associated with the iddingsite.. Furthermore, all Nakhlites examined closely by us and by other also show biomorph features. Figure 3. A fragment of Nakhla. This original piece had a nearly complete and continuous fusion crust. The light gray area, mostly pyroxene with lesser olivine, was revealed only after the original piece was cracked open. BIOMORPHS AND RECENT BIOLOGICAL CONTAMINATION The possibility that meteorites may become contaminated with microbial features such algae, mold, fungi, bacteria, or even detectable viruses must always be considered19; 20. We have been particularly alert for such terrestrial contamination in both meteorite falls and meteorite finds. The next figures present biomorphs that we considered to be terrestrial contamination from three separate martian meteorites. They all have in common the presence of abundant low Z composition, high C, the lack of typical elements from natural rocks and minerals, and the tendency to charge up in the electron beam. In addition they have diagnostic textural features that are different from the biomorphs that we describe later in this paper which we consider to be genuine indigenous features. The terrestrial contaminants are always found covering or partly covering the substrate without any indication that they are intergrown or embedded with the substrate. Clear contamination may consist of filaments that extend through existing cracks and may disappear into holes or cracks, but in no case are they intergrown with indigenous matrix, mineral crack filling, or mineral coatings. Particularly critical is textural evidence that a biomorph is intergrown or embedded with matrix material such as iddingsite which is known to be indigenous based on independent evidence such as non-terrestrial isotopes. So specific criteria to determine terrestrial contamination includes both composition and texture of the biomorph. The most reliable evidence for indigenous biomorphs is the textural evidence showing that it is embedded and completely intergrown in a mutually reactive manner with a matrix such as iddingsite which can be firmly documented as nonterrestrial. The chemical evidence of high organic composition and the lack of rock-forming chemistry is added support for contamination. From a practical approach, features that have obvious low Z, are electron transparent, and show abnormal charging properties are likely contamination. If we have any question about any of these characteristics we do not include the features in our category of possible Martian biomorphs. Figure 4. Likely microbial terrestrial contamination found in three separate martian meteorites. These feature consist of filaments and biofilm-like coatings on martian mineral grains in Nakhla (A), in Shergotty (B), and in ALH84001 (C).. NAKHLITE BIOMORPHS When Nakhla is broken, the chips sometimes part along existing fractures and reveal surfaces with iddingsite. Attached to these surfaces or partly embedded in the iddingsite are a number of features which can be correctly called biomorphs because of their close resemblance to features known to be created by biology in the form of microbial bodies, casts of microbes, molds of microbes, or remnants of biofilms or debris from microbial colonies. Most of these features are associated with iddingsite Iddingsite is a common crack and void filling in Martian meteorites, particularly Nakhlites. The iddingsite in Nakhla has been roughly dated at about 700my21; 22 The measured D/H ratios is +500 to +800 per mil for minerals associated with the iddingsite23. This isotopic composition is much to heavy to be the result of terrestrial water and its presence shows that the iddingsite was formed on Mars.23 All of these biomorph features are common in recent geologic samples of material known to be rich in biology, and some of these features are found in ancient rocks where they have been specifically identified as microfossils or trace fossils. All of these features fail the test for terrestrial contamination as discussed previously. The following figures are examples of biomorphs in Nakhla and the Nakhlite Yamato000593. Figure 4 is an example of biomorphs in Nakhla. These forms resemble closely many coccoid fossils. SEM/EDX chemical analysis of these features reveals that they are dominated by iron and oxygen and the substrate is iddingsite. Figure 5 shows several coccoid biomorph forms partially wrapped in a layer of iddingsite. Clearly, these forms were present during the interval when the iddingsite precipitated. They are contemporaneous with the formation of the iddingsite Biomorphs may resemble typical bacteria but may show weathering, alteration, and partial replacement. Figures 6-8 show some of the complex iddingsite texture and presents the concept of microstratigraphy of the iddingsite which can be used to provide relative time information on the biomorphs. Figure 9 is a rod-shaped biomorph that displays a basic iron oxide chemistry but also shows a significant carbon peak in the EDX spectrum. This biomorph is not completely embedded in the adjacent iddingsite substrate, but is firmly attached to it. Figures 10-11 show biomorphs from the Columbia River Basalt. Such biomorphs from known biologic regions within rocks provide examples and standards for comparison with the martian biomorphs. Figure 4. Biomorphs in Nakhla attached to a substrate of iddingsite. The composition of these objects is mainly iron oxide. Figure 5.Coccoid biomorphs partially wrapped by iddingsite supporting the interpretation that they are indigenous and contemporaneous with the precipitation of the iddingsite. Figure 6. Multiple layers (at least four) of iddingsite in Nakhla showing the complexity of this crack filling sequence. A biomorph similar to a collapsed cell is present in the center of the image and is partly imbedded in the layer of iddingsite. Figure 7. Another region of multiple iddingsite showing several thin layers (bottom), a fragmental layer, and a massive partly stratified layer, and another fragmental layer. Figure 8. Nakhla biomorphs consisting of coccoidal and rod-shaped forms partly embedded in iddingsite,