Mars microbes may make methane: the Viking view
The discovery of extraterrestrial life could impact biology in a manner not seen since the discovery of the double-helix structure of DNA. Yet to date, only one life-detection mission, the 1976 Viking lander, has ever traveled to another planet. The Viking Labeled Release (LR) experiment has been a source of controversy for more than 30 years. In these experiments, a small amount of a 14C-labeled nutrient solution was added to a Martian soil sample and the evolution of 14C-labeled gas was monitored with a beta detector. The rapid evolution of gas and the failure to see such evolution in sterilized soil was considered presumptive evidence for Martian microbiology,1 although alternative explanations have dominated the interpretation of this data until recently. Our more recent analyses have suggested that the presence of circadian rhythmicity, a reliable biosignature, in the temporal structure of the gas release data further support the idea that the gas was of biological origin.2
The discovery of methane in the Martian atmosphere,3 and its steady replacement from equatorial sites that seem to host substantial subsurface ice deposits,4 suggests a possible biological origin, methanogenic microbes. An apparent absence of Martian volcanic activity precludes one explanation for the presence of methane. These findings suggest that methane, in addition to CO2, was evolved in the LR experiment.5 Here, we consider how much LR gas could have been methane and whether extrapolation of the amount released per gram of Martian soil could produce the measured levels of methane in its atmosphere. Finally, assuming terrestrial rates of methane production by methanogens, we estimate the necessary size of the microbial population and the amount of water necessary to maintain that population in an aqueous state.
For a first approximation, we will largely ignore temperature, pH (a measure of acidity or basicity of a solution), and atmospheric-pressure differences, and any correction for doubling rates of the microbes. Terrestrial rates of methane production by methanogens could easily account for the observed levels of methane in the Martian atmosphere. For instance, in a study of hyperthermophilic methanogen at 85°C and 0.4MPa pressure (1ml of medium),6 first doubling occurred at 6hr with 4×106 cells and 37μmol/ml of methane released. Similarly, Kral7 produced results in terrestrial experiments using a wetted Mars soil stimulant such that methane release was approximately 7μmol/ml with methane doubling times of 1–5 days. Estimates of methanogen density and methane production varied from 104 to 4.3×105 microbes/ml, depending on the particular microbial species and the amount of water mixed with the Mars soil stimulant. These estimates are similar to various reports of terrestrial methanogen soil densities of ~105 microbes/g.
In the Viking LR experiment (e.g., VL2c3: see Figure 1), evolved gas increased slowly over the first two weeks, resulting in approximately 30nmol of a radio-labeled carbon-containing gas. However, approximately 1/3 of the gas was resorbed following second administration of an aqueous nutrient (0.115ml) to a 0.5cm3 soil sample. Methane is not soluble at the temperatures and pressure of the LR experiment but CO2 is. If the nonresorbed fraction were entirely composed of methane, and in terms of methane/ml of added nutrient (for comparison), it would be approximately 174nmol/ml. This is ~40× lower than the maximum Kral result and might thus scale down to a population of 104 microbes/ml, still well within the range of terrestrial methanogen soil densities.
The total amount of methane in the Martian atmosphere is 2.5×108 kg. Without a renewal source, that much methane would be broken down by UV photolysis in some 300yr. However, more recent estimates suggest that the actual methane lifetime is ~600 times shorter than that required by photochemical processes,8 implying a global methane-production rate of 1018 nmol CH4 per 2 weeks. Extrapolating the Viking LR numbers suggests that 5×1015ml of a probably briny, high-osmolality medium would be needed to support an active global methanogen population of approximately 1020 microbes in 2.5×1015g of Martian soil. This amount of mass could be provided by a 10m-deep shell around the planet. Kral's data suggests that water is rate-limiting for methane production, which may have been true for the Viking LR experiment as well. If so, there could be even larger numbers of inactive (possibly sporiform) methanogens in dry Martian soil.
The water content of Mars is now estimated at approximately 1/1000 of that of the Earth (4.4×106 km3 or 4.4×1021 ml). The aqueous requirement for a putative, slow-growing methanogen with kinetics inferred from the Viking experiments would only require roughly a millionth part of Mars water in a liquid, presumably in a high-osmolality form. Liquid water can exist in microdomains in ice on Earth and possibly on Mars. Alternatively, at a sufficient depth below the Martian ice layer, temperatures may be high enough to allow water, particularly a brine, in liquid state. Prospects would be further improved if growth and cell-division kinetics were assumed to follow those of Kral or Takai.6
The remarkable result is that a population of methanogens existing at depths from the surface to no more than 10m below the surface on average, and in a low concentration of Mars water similar to the Viking LR experiments, could easily replenish the methane content of the Martian atmosphere on the timescale required. These estimates are global averages by necessity. If methanogens exist, they may be concentrated in geologically restricted patches or oases. We will improve our simulations and extrapolations, and replace them by more realistic estimates of Mars soil pH, water content, temperature, atmospheric pressure, and constituents. But considering the available data, we believe that after an interregnum of some 35 years, it is time for NASA to again fly an explicit life-detection mission to Mars.
Joseph Miller received his PhD in 1979 from the University of Texas (UT) and did postdoctoral work in neurophysiology, neuropharmacology, and circadian biology at UT Health Sciences Center, the University of California at Riverside and Davis, and Stanford University. He is currently an associate professor of cell and neurobiology.
Marianne Case is a graduate student. She is a computational modeler and studies spatial processing in the entorhinal cortex and hippocampal formation of rats. In 2005, she graduated from the University of Southern California with a BS in biochemical engineering.
Patricia Ann Straat received her AB (1958) and PhD (1964) from Oberlin College and Johns Hopkins University, respectively. She was co-experimenter of the 1976 Viking LR Life Detection Experiment (LDE). From 1980 to 2001, she worked at the National Institutes of Health. Although retired, she continues collaborating on interpretations of the LR data.
Gilbert Levin, experimenter of the LR LDE on NASA's Viking mission, has led many research projects in public health, science, and engineering. An adjunct professor, he is also a Johns Hopkins University PhD and serves on its engineering and library advisory councils.
