CP-2156 Life In The Universe


Atmospheres and Evolution



[79] Bacteria have dominated Earth's biosphere since early Precambrian times, altering surface and atmosphere, maintaining and evolving the environment of all life. To understand the phenomena on which our survival depends, we must view Earth in a unified context.


Our purpose is to summarize some work of many people concerning regulation of the atmosphere and its relation to the evolution of microbial life. Remarkably, looking back, one recognizes a strange unconscious kind of cooperation among millions of species of organisms over at least 3 billion years. This cooperation, itself a product of evolution, has altered the entire surface of planet Earth.

We will have five main points. First, from a planetary perspective-that is, comparing Earth, Mars, and Venus - Earth's atmosphere is totally improbable. It has too much oxygen in the presence of too many gases that react with that oxygen. Furthermore, for a long time-hundreds of millions of years at least - Earth's atmosphere has not been obeying the rules of physics and chemistry alone. Perhaps we are beginning to understand why not.

Second, life on Earth is far older than most people had reckoned until recently. Life extends as far back as the oldest sedimentary rocks, perhaps as far back as the oldest rocks on Earth, which are metamorphic (Schidlowski et al., 1979).

The third point is that the living world is not divided primarily into plants and animals. Green plants and animals are in fact closely related organisms. The living world is rather divided into two other groups: procaryotes such as bacteria, whose cells lack nuclei, and eucaryotes such as animals, plants, and fungi, which are nucleated organisms. The differences between these groups are far greater than the differences between animals [80] and plants. The oldest life on Earth consisted of bacterial cells, which tend to be small. Bacteria have dominated Earth since the early Precambrian, and despite our anthropocentrism, they dominate Earth today. Bacteria far more than eucaryotes have altered the planet and its atmosphere. We must understand the interactions of bacteria with the atmosphere and the surface of Earth to perceive why Earth is so different from Mars and Venus.

Fourth, regarding the origins of the eucaryotic cells, they probably evolved from bacterial cells by mechanisms other than the simple accumulation of random mutations and that the establishment of intracellular symbioses played an important part in this process. Different species of bacterial cells established associations and coevolved, forming new kinds of units that became ancestors to animal, plant, and fungal cells. The appearance of organisms composed of nucleated cells led to many changes on Earth and was a prerequisite for the superficial "explosion" of invertebrate animals at the Phanerozoic boundary.

Finally, we will discuss the idea that the gases of the atmosphere are kept in balance mainly by bacterial cells. The argument will be that species diversity is absolutely required for current maintenance of conditions hospitable to life. Furthermore, without species diversity, the living system on the surface of Earth would not have originated and been maintained for more than 3 billion years.

The results of the Viking mission to Mars are having a profound effect on our thinking about Earth, which now appears to be the only planet in the Solar System that harbors life (Mazur et al., 1979). An important scientific goal is therefore to understand the factors that make our planet and its living system unique. It is our belief that, unless scientists of various disciplines employ their tools in a joint effort to look at Earth in a planetary context, we will never understand the phenomena upon which our survival depends.

Let us start by comparing the atmospheres and surfaces of Mars, Earth, and Venus. The surface of Mars, pockmarked with craters, probably records very ancient events in the history of the Solar System. The Viking lander found a rubble-strewn, extraordinarily dry terrain. The atmosphere is primarily composed of carbon dioxide, although it contains almost 2% nitrogen and less than 1% oxygen. Mars now is dry, oxidized, and, relative to the Earth, unchanging.

Before the hot, acidic conditions on Venus burned up the camera in less than a minute, Venera 9 took photographs of another dry, rubbly surface. The Soviet missions to Venus have studied the surface of that planet. They reveal a place in some respects very much like Mars. The atmosphere of Venus, too, is composed of carbon dioxide and small quantities of nitrogen, traces of oxygen, and very little water.

[81] In what way is Earth's surface improbable? Superficially, it appears to be a normal planet between Mars and Venus. Without life, Earth's atmosphere would probably be far more like that of its neighbors, containing carbon dioxide, nitrogen, trace amounts of oxygen, and water vapor. But as it is, Earth's atmosphere has far too little carbon dioxide and far too much oxygen.

Earth's atmosphere contains the hydrogen-rich gases that are found in the atmospheres of the outer planets, but it also contains oxygen, which tends to react with hydrogen and hydrogen-rich gases. There are other chemical anomalies in Earth s atmosphere: it contains sulfur compounds such as dimethyl sulfide as well as highly improbable compounds such as terpenes, disparlure, eugenol, myoporum, nepetalactone, and butyl mercaptan. Unlike any of the inner planets, Earth's atmosphere also contains phosphate, but it is in particulate forms we call spores, seeds, birds, and bats! The presence of these complex organics can only be understood on the basis of their informational role among the biota (Margulis and Lovelock, 1975).

Table 1 documents the chemical improbability of the atmosphere by answering the question: Given that the atmosphere is 20% oxygen, what quantities of the other reactive gases would we expect to be simultaneously present?

Hydrogen and oxygen, of course, react to form water; nitrogen and oxygen react to form nitrate. Ammonia and oxygen also react, as do methane and oxygen. Earth's atmosphere seems to be composed of a combustible mixture. From a chemical point of view, it is far from equilibrium, yet it is maintained in a steady state with a composition unlike that of the atmospheres of Mars and Venus. It is rather more like fuel than like exhaust. Given 20% oxygen and standard conditions, table 1 shows what the approximate equilibrium concentration of the other gases should be. Nitrogen is too abundant by about a factor of 10, but methane is too abundant by a factor....




Present concentration

Expected equilibrium concentration a

Approximate discrepancy

Residence time, yr


106 tons/yr






3 x 106



1.5 x 10-6





Nitrous oxide

3 x 10-7











Methyl iodide







5 x 10-7





a Given that oxygen comprises about 0.20 and is released into the atmosphere at the rate of about 110 x 109 tons per year.



[82] ...of about 1035 ! Scientists often make errors of a few percent, and are occasionally off by factors of 100, but when we make errors of 1035, we are very poor scientists indeed. Models that attempt to explain the composition of Earth's atmosphere by physics and chemistry alone tend to sustain errors of 1035, a strong argument that such models do not suffice. The atmosphere also contains far too much nitrous oxide, ammonia, hydrogen, and hydrocarbons, given the quantity of atmospheric oxygen. Furthermore all gases listed in table 1 move through the atmosphere. From a geological point of view they have very short residence times: they cycle in years, months, or even days in some cases. Of the nonnoble gases, the residence time of nitrogen is the longest, of the order of a million years. Those of methane, nitrous oxide, and the others are far shorter-only 10 years or less.

The annual output of these gases-the production rates-are enormous, measured in billions of tons per year. These large production rates and short residence times imply that gaseous emissions occur continuously. What produces and maintains such reactive gases at anomalously high concentrations? The tectonic and volcanic contribution is minor. In fact, the vast quantity of cycling gas is produced by microorganisms and plants, with animals contributing very little. Furthermore, neither plants nor animals qualitatively contribute to the production of many gases, such as methane and nitrous oxide; however, for each gas listed in table 1 there is always a microbial source (Margulis and Lovelock, 1974).

Our second point is that life is very much older than previously thought. Figure 1 shows a "time line" measured in billions of years. Little is known about Earth before 4 billion years ago. However, starting about 3.5 billion years ago, a fossil record of life is abundantly available. Most scientific effort has gone to develop a paleontological perspective that does not begin until the Phanerozoic, only about a half a billion years ago. Extensive fossil reconstructions of life during the past half billion years are available, as museum lovers are well aware. But what occurred prior to half a billion years? Why is the record for life so different during the first 3 billion years of Earth's history?

By half a billion years ago, there were well-developed communities of invertebrate animals and algae. The "dawn of fossil records" is marked by trilobite fossils that appear about 600 million years ago at the Cambrian boundary. By the lower Paleozoic era, many kinds of skeletalized animals had appeared, some squid-like, some starfish-like. Most look generally familiar even though they were members of species now extinct. An immense body of paleontological evidence exists for communities of reptiles, tropical forests, aquatic invertebrates, coral reefs, and so forth by Mesozoic times. Ecosystems during the Phanerozoic, the last half billion years, although differing from modern ones in detail, are recognizable and changed in familiar ways. They are quite different from those of the pre-Phanerozoic.



Figure 1. Times of origin of major events during past eons.

Figure 1. Times of origin of major events during past eons.


[84] During the late Cenozoic, populations of woolly rhinoceroses and mammoths and other familiar but now extinct beasts roamed the Earth. Forest, lake, savannah, alpine, reef, and a multitude of other well-known ecosystems to support these animals had long since been established. In the last few million years, human-like australopithecines and members of our genus Homo have thrived. On the scale of billions of years, whatever else he is, man is but an upstart on Earth, some 0.004 billion years old. Furthermore, Homo sapiens came into a well prepared scene in which complex interacting microbial and plant ecosystems were highly developed.

Bacterial ecosystems preceded those dominated by animals, plants, and fungi (Reimer et al., 1979). Familiar animal, plant, and fungus species are, in fact, closely related to each other, and the eucaryotic cells of which they are composed appear relatively recently in the fossil record, during the Proterozoic. Some differences between them and the smaller, nonnucleated bacterial cells that greatly preceded them are indicated in figure 2.


Figure 2. Comparison of the structure of procaryotic and eucaryotic cells.

Figure 2. Comparison of the structure of procaryotic and eucaryotic cells.


[85] All bacteria, including the blue-green algae, are morphologically relatively simple. Chemically, however, bacterial cells are extremely complex and diverse. From the point of view of gas metabolism, bacteria are far more versatile than animals, plants, and fungi, which take in oxygen and release carbon dioxide. Plant cells also take in carbon dioxide for photosynthesis, but animal, plant, and fungal aerobic metabolism is generally rather uniform. The point we wish to make here is that nearly every major metabolic feat performed by these eucaryotes is also represented in the bacterial world. Moreover, bacteria are able to carry out many other gas transformations such as nitrogen fixation and methanogenesis. It is generally conceded that bacterial cells initially changed the atmosphere; for example, the blue-greens first produced large quantities of atmospheric oxygen (Cloud, 1974). We will argue that the metabolic uniformity of eucaryotes can be understood in the context of a symbiotic theory of the origin of animal, plant, and fungal cells, which asserts that certain partnerships among bacterial cells eventually led to the origin of nucleated cells (Margulis, 1981).

Bacterial cells lack chromosomes, whereas plants and fungi tend to have chromosomes so like those of animal cells that often the three types of cells are difficult to distinguish. The ultrastructure of bacterial cells, including the blue-greens, as revealed by electron microscopy is so different from that of nucleated organisms that there is no ambiguity in the distinction.

What was the course of evolution during the enormous stretch of time from 3.5 until about 5 billion years ago? The Precambrian era was once considered to be nonfossiliferous, that is, to have left scant or no evidence of life at all. But in the last 20 years or so, it has been shown that life was abundant during the Archean and especially during the Proterozoic. Fossiliferous rock samples from all over the world and from various points in time before the Phanerozoic are now available. Part of the difficulty that hampered recognition of pre-Phanerozoic life is related to techniques and to scientific expectations. One aspect of this is the gradualistic view of cell evolution that prevailed before the symbiotic theory was considered.

A comparison between an ancient rock sample, a carbon-rich crypto-crystalline silicate rock called a chert, and a modern sediment is shown in figure 3. Until recently such commonplace reservoirs of life have been ignored by biologists and paleontologists. In fact, upon close examination, both the ancient and modern sediments are filled with evidence for life. A great part of the revolution in our thinking about rocks and soft sediments such as these in the context of the early history of life is attributable to the insight of Elso Barghoorn of Harvard and his late colleague Stanley Tyler. These investigators realized that the conspicuous fossil record of large organisms must have been preceded by something smaller. Such ideas led them to microscopic studies of thin sections of unaltered sedimentary rocks.



 Figure 3. Comparison between an ancient thinly laminated chert from the 3.4-billion-year-old Swaziland system in South Africa with banded sediments from alive community of microbes at Laguna Figueroa, Baja California, Mexico.

Figure 3. Comparison between an ancient thinly laminated chert from the 3.4-billion-year-old Swaziland system in South Africa with banded sediments from alive community of microbes at Laguna Figueroa, Baja California, Mexico.


[87] If some of the black and smooth areas of cherts such as those shown in figure 3 are sectioned with a diamond knife and examined with optical microscopy, inclusions interpreted to be remnants of the oldest organisms on Earth can be observed. Whether they are indeed fossil bacteria is still debated; we think that the evidence for fossils in rocks older than 3 billion years is excellent (Knoll and Barghoorn, 1977). By 2 billion years ago, a multitude of microbes thrived; these are detected as fossils in thin sections and macerations of rocks (Schopf, 1975).

A further important clue to ancient microbial life is given by the stromatolites, which are layered sedimentary rocks produced by the trapping, binding, and precipitation of sediments by actively metabolizing communities of microorganisms. They may be thought of as the ruins of ancient "cities" of microorganisms, which began as microbial mats but then became lithified and preserved. Examples from all over the world are now available (Walter, 1976).

Some ancient cherts reveal upon microscopic examination large numbers of well-preserved, fossil microorganisms, including filamentous ones that resemble modern blue-greens so closely that they probably belonged to genera that persist today and that still produce atmospheric oxygen by photosynthesis. Direct microfossil evidence for the potential to produce atmospheric oxygen 2 billion years ago and probably far earlier is abundantly available.

Are there living ecosystems dominated by microorganisms that might tend to lithify and thus might be used for comparison with ancient stromatolites and fossiliferous cherts? Figure 4 shows Laguna Figueroa in Baja California del Norte, Mexico, some 250 km south of San Diego. Very little life seems to prevail in such hypersaline lagoons; in the absence of fresh water input, animals and plants are virtually absent. However, we are beginning to recognize that not only are such sediments thriving with microbial life, but they may provide realistic analogs to fossilized microbial communities. Laguna Figueroa is a closed, generally dry lagoon dominated in some regions by conspicuous evaporatic polygons (Horodyski, 1977). When these are sectioned, sediments of differing colors and textures are revealed. The colored laminations are due in part to stratified microbial communities, for example, mats dominated at the surface by the sheathed cyanobacterium, Microcoleus vayinatus, and underlain by purple photosynthetic bacteria (Margulis et al., 1979a). Smelly gases are emitted from the sediments; gas bubbles are seen everywhere. Whatever scums these mats and sediments contain, they certainly are sources of some of the "out-of-equilibrium" gases presented in table 1. The mud-bound microbes produce hydrogen sulfide, ammonia, methane, and many other gases-none of which have yet been measured in the field.



Figure 4. Panoramic views of the microbial mat communities at Laguna Figueroa. (Courtesy of Robert Horodyski).

Figure 4. Panoramic views of the microbial mat communities at Laguna Figueroa. (Courtesy of Robert Horodyski).


In dry seasons this lagoonal complex extends for about 8 km in the north-south direction and 1 km across. Is life depleted in this hypersaline hot harsh region? No. Complex patchy colored surfaces prevail, composed of dense communities of microbes, the vast majority of which have not been identified and cultured. Some of the microbes in the mats at Laguna Figueroa resemble those preserved in cherts several billion years old (Knoll and Barghoorn, 1977); these communities are also similar to the inferred environment of deposition of the carbon-rich shales of South Africa, which are over 3 billion years old (Reimer et al., 1979).

In communities such as those at Laguna Figueroa, which may serve as analogs to Precambrian microbial ecosystems, the photosynthetic bacteria are usually well represented. Sheets of purple anaerobic photosynthetic bacteria- organisms that remove hydrogen sulfide but do not produce oxygen [89] - may even form extensive surface colonies. These bacteria produce food for other forms of bacteria thriving below the surface. The conspicuous layers, at least near the surface, represent the stratifications of the naturally growing microbial ecosystem (fig. 3). In many places the surface layer is made of evaporites such as gypsum and halite; sand and sometimes aragonite are present. The second layer often tends to be dominated by blue-greens, which produce oxygen. The lower layers usually harbor anaerobic photosynthetic bacteria, with nonphotosynthetic bacteria of many kinds below them. Figure 5 shows some photosynthetic and other bacteria isolated from the microbial mats. Electron micrographs of material from surface scums of these mats reveal many photosynthetic bacteria, some probably new to science (Margulis et al., 1979a).

The sediments of the Laguna Figueroa community are compared in figure 3 with a 3.4-billion-year-old banded chert from the South African Swaziland system. Again, the community of live microorganisms is hypothesized to be analogous to some ancient one that was silicified. The modern microbial community probably lives under conditions of higher salinity than the Archean one. In today's world, if there were less salt and less extreme conditions generally, animals and plants would tend to dominate. This sort of invasion by eucaryotes was seen in the spring of 1979 after a long period of heavy rains and an influx of fresh water. Conspicuous microbial communities such as these are restricted to quiet waters in hot saline tropical and subtropical locations such as the Bahamas, Baja California, and western Australia, and to hot springs and geysers. In the Precambrian, when competition by animals and plants was lacking, they were probably far more widespread.

Evidence for ancient communities of microorganisms abound in the Precambrian fossil record from 3.5 until 0.5 billion years. Stromatolites made of calcium carbonate or, less frequently, of silica are especially common from 1.0 to 0.5 billion years ago. Some have striking modern counterparts. Examples of stromatolites from tropical western Australia have been described in detail (Walter, 1976). If these laminated calcium carbonate rocks are sectioned again, stratified communities of microbes will be found with them. The species comprising the community in large stromatolites differ from those found in Laguna Figueroa sediments. In carbonate-accreting environments, sheathed filamentous blue-greens grow, trapping sand, calcium carbonate particles, and other bits of sediment. By trapping particulates in their mucous polysaccharide sheaths, blue-greens often cover themselves with clasts until the light they require for photosynthesis no longer penetrates (Golubic, 1973). Many of these filamentous blue-greens exhibit a kind of gliding motility. Thus they are capable of active movement, emerging from beneath the surface layer of particles and growing up through it. This type of growth pattern forms layers rich in the remains of organisms:



Figure 5. Ultrastructure of some microorganisms taken from Laguna Figueroa. P= phage (Courtesy of David Chase).

Figure 5. Ultrastructure of some microorganisms taken from Laguna Figueroa. P= phage (Courtesy of David Chase).


[91]....organic debris alternating with layers rich in clastics. Such recent stromatolites found at Hamelin Pond in Shark Bay, western Australia, are similar to ancient ones formed some 2.3 billion years ago found near the Arctic Circle in northwest Canada (Walter, 1976).

The direct evidence for pre-Phanerozoic life also includes "chemical fossils," refractory organic complexes found in the keragen fraction, material, by definition, not extractable from rocks by standard aqueous or organic solvents. Thus there is substantial stromatolite microfossil and geochemical evidence that early Earth was dominated by bacterial communities.

From where, then, did communities of animal, plants, and fungi come? Eucaryotic cells contain organelles ("little organs" or cell parts) called mitochondria that are the sites of oxygen respiration (fig. 2); the mitochondria are absent from all bacterial cells. In addition, plant and algal cells contain plastics, membrane-bound compartments within which photosynthesis takes place. It is our opinion that both the photosynthetic plastics of plant cells and the mitochondria of plant, animal, and fungal cells were once independent bacteria. New biochemical data on the sequence of amino acid residues in certain proteins are in agreement with the concept that plastics and mitochondria were once free-living bacteria, thus suggesting a symbiotic origin of eucaryotic cell components (Schwartz and Dayhoff, 1978). A plant cell can be thought of as a composite formed by partnerships among bacteria that could convert sunlight into food, bacteria that became plastics, and aerobic bacteria that generated energy by burning oxygen and became mitochondria.

One striking difference between eucaryotic and procaryotic cells is the nature of the whiplike organelle of motility, the flagellum. In eucaryotes, the flagellum has a peculiar complex ninefold symmetry. It is made of nine sets of doublet tubes of protein and two central tubes, all called microtubules. These so-called 9+2 structures (fig. 2) are absent in procaryotes. The organelle of flagellated bacteria is far smaller, solid, and simpler, and is composed of an entirely different protein than those found in eucaryotes. The name "undulipodium" (Smagina in Corliss, 1979) has been suggested to refer to the eucaryotic but not the procaryotic flagellum. The uniformity of the 9+2 undulipodia is remarkable; for example, the ultrastructure of both the gill cilia of marine animals and the oviduct cilia of women display this ninefold symmetrical pattern of microtubules. Furthermore, not only do the sperm tails of mammals and amphibians, indeed of nearly all animals, have the 9+2 motif, even male ginkgo trees have the same detailed structure in the tails of their sperm! This sort of observation has led to a suggestion that both animals and plants evolved from cells bearing undulipodia, which may themselves also have been free-living once (Margulis et al., 1979b). The hypothesis we arc now testing is that undulipodia were once highly motile and originally on their own as spirochaete bacteria that formed symbioses with the rest of the cell.

[92] Regardless of the precise details, if the symbiotic theory is correct, all animal cells-including our own, of course-formed from partnerships of two or maybe three members of different procaryotic species. Plant cells with their photosynthetic organelles evolved from these same partnerships with yet another set of symbionts-photosynthetic ones that became plastids. In summary, then, the concept is that bacteria of several very different species came together in certain sequences to form new complex units: eucaryotic cells. Natural selection acted on the complexes - associated microbes left more offspring than their unassociated free-living counterparts. The theory states that the first step in the origin of animal, plant, and funga cells was an association between two types of bacteria, one fermenting and one respiring. The respiring bacteria were efficient in using oxygen. A second step involved the acquisition of motility through the association with spirochaete symbionts that eventually evolved to be organelles of motility, the undulipodia. The third step, which occurred only in the ancestors of algae and plants, was the acquisition of the capacity for photosynthesis, that is, the formation of hereditary symbioses with procaryotes that became plastids. Thus animals, plants, and fungi-complex organisms made of complex cells -have not only evolved by natural selection acting on organisms in which favorable single mutations have accumulated (neo-Darwinian evolution) but are also products of symbiotic associations.

The power of associations can be impressively illustrated by analogy to live modern symbioses. New symbioses still occur that lead rather abruptly to new complexes with traits different from either of the individual partners. For example, bacteria are known that always invade other bacteria (Stolp, 1979); this may explain how bacteria that became mitochrondria first entered fermenting host cells. Live examples of motility symbioses are also known: some spirochaete bacteria attach to the surface of host cells, for example, to Mixotricha paradoxa, and move them around.

Mixotricha is a huge eucaryotic microorganism that lives in the hindgut of dry wood termites. The single cell is about half a millimeter in length. It ingests pieces of wood and digests the cellulose. It has four small undulipodia that it uses as rudders to change its direction. But in order to swim it uses its symbionts: about 500,000 highly motile surface spirochaete bacteria anchored to its surface. There are other types of symbiotic bacteria inside the Mixotricha cell (Cleveland and Grimstone, 1964). M. paradoxa is a symbiosis of at least three different kinds of microorganisms, and living inside termites, it is itself a symbiont. Without the wood-digesting abilities of microbial complexes such as M. paradoxa, wood-eating termites could not survive, since only a few bacteria, protists, and fungi contain cellulases and other enzymes required to digest wood.

Some motility symbioses can be seen on films of microbes from Pterotermes occidentis and Kalotermes schwartzi, dry wood termites from [93] southern Arizona and southern Florida (Margulis et al., 1978). These termites break. wood into manageable pieces and ingest it. Their soldiers, however, have mouthparts so modified for defense of the colonies that they cannot even ingest wood: they must therefore be fed by other termites through the rear. Soldier termites of these species thus depend on hindgut microorganisms supplied by members of other castes. E. O. Wilson of Harvard University has suggested that termites became social insects primarily because they must transmit their wood-digesting hindgut microbes to each other.

The microbial community is sensitive to oxygen. In dry wood termites, the anaerobic microbial community digests wood while the hypertrophied termite hindgut maintains the appropriate environmental conditions for microbial growth and cellulose digestion.

The dry wood termites can live on a dry, hard wood diet-billiard balls or cellulose fibers and water! If by chemical or heat treatment they are deprived of their community of microbes, they die within weeks. Dry wood termites, like all organisms, need sources of nitrogen and sulfur in their diets. Wood has very little nitrogen and sulfur; cellulose fibers have none. At least some termites therefore harbor other bacterial symbionts that fix the nitrogen from the air. Only a few other bacteria are capable of fixation of atmospheric nitrogen. Thus, a single termite is in fact a complex ecosystem harboring, in some cases, over 30 different species, many of them anaerobic.

The dry wood termite hindgut community usually contains a variety of species of spirochaete bacteria. The role of these spirochaetes in the community is unknown; despite enormous effort, it has not been possible to grow them (and indeed most of the other termite microbes) outside the hindgut ecosystem. Production and removal of gases is probably involved in providing the proper milieu for the microbes, but which gases and in what quantity are not known.

Spirochaetes of termites tend to form symbiotic associations, which is why we have studied them (To et al., 1978). Some large spirochaetes contain microtubules, an extremely rare feature in procaryotic cells (Hollande and Gharagozlou, 1967; Margulis et al., 1978). The presence of microtubules at least superficially like those found in undulipodia suggests that spirochaetes and undulipodia may have common ancestors.

Spirochaetes often move or beat together, generating forces that can move other, larger organisms. Some show behaviors such as might be expected of organisms ancestral to undulipodia, if indeed undulipodia were once free-living (Margulis et al., 1979b). Individual spirochaetes beat in complete synchrony, apparently by virtue of their physical proximity.

Is the protein making up the spirochaete microtubules the same protein, tubulin, found in the microtubules of animals, plants, and fungi? Microtubules composed of tubulin have been discovered in nerve and sensory cells, [94] and in the mitotic spindles of all animals, plants, and fungi. Where did microtubules come from? Were they brought into cells symbiotically by once freeliving spirochaetes that became undulipodia? Within a decade perhaps we may be able to prove or disprove the hypothesis of spirochactal origin of undulipodia and other organelles composed of microtubules.

Microbial symbioses often occur in anaerobic microenvironments. Many wood-digesting protists in termite guts are covered with hundreds of surface bacteria. In lake muds, in the digestive tracts of animals, and in the anaerobic environments found below the evaporitic surface in the microbial mats at Baja California, for example, symbiotic relationships between microorganisms seem to be common. Without such complex microbial ecosystems, there would be no breakdown of wood and many other organic products. Without microbial degradation and gas release, there would be immense accumulations of dead and diseased carcasses and a failure to cycle through the gas phase the elements required for life to continue.

Some extant symbioses provide analogies for the acquisition of photosynthesis. In many distinctly different cases, animals or other nonphotosynthetic but motile organisms have established stable symbioses with photosynthetic partners; the complex tends to survive, grow, and leave more offspring because it has the double advantage of motility and photosynthesis. In nutrient-poor waters such associations are particularly common. For example, coral animals and giant clams of the south Pacific are routinely symbiotic with photosynthetic algae called dinoflagellates. Several species of worms such as Convoluta paradoxa routinely harbor algae in their tissues. Some snail-like molluscs have acquired photosynthetic partners, including many if not most of the radiolarians and foraminiferans. In all of these cases, nutrient transfer occurs from the photosynthetic to the nonphotosynthetic partner. For example, Convoluta paradoxa does not feed through its closed mouth; it sunbathes. The algae living between the cells of the worm's tissues produce food for the worm by photosynthesis. Furthermore, the continuity of the partnership between algae and worm is assured. When the flatworm dies, its tissue algae are released into the water. The swimming algae are attracted by some chemical signal to the surface of the worm eggs, and all normal members of the species Convoluta paradoxa are green due to the chlorophyll in the symbiotic algae. The partnership has been selected because it has advantages relative to each separate partner alone.

The origin of such algae-animal associations is analogous to the origin of plant cells. Plant cells can be thought of as animal cells that early in their history trapped photosynthetic microbes inside them. Such photosynthetic, oxygen-releasing microbes evolved into organelles and with time became the totally dependent plastids.

How does this discussion of cells relate to our earlier remarks about atmospheres? The major relationship between the larger groupings of organisms [95], the kingdoms plotted as a function of time, is shown in figure 6. The outlines of the history of life over time can be roughly drawn for the last 3.5 billion years. A remaining mystery is how life in the form of anaerobic bacterial cells arose from nonlife so quickly. The details of the origin of life, especially of the immediate precursors to procaryotic cells, are still obscure. However, unbroken threads of life. The earliest organisms-anaerobic bacteria of many kinds-are still with us today. All produce and remove atmospheric gases, and many strongly interact with the clastic particles, minerals, and humic acids of the sediments. These and other bacteria, fungi, and protists make soil what it is: a far more complex material than the martian or lunar regolith. Bacteria originated early, have diversified a great deal on the metabolic level, and have modulated Earth's surface environment until the present.

After the diversification of anaerobic bacteria, another revolution occurred: oxygen became plentiful in the environment, at first because it was a waste product of the photosynthetic activities of the blue-greens. Since oxygen in large quantities is required by virtually all eucaryotes, it was probably plentiful in the environment by the time eucaryotic cell organization from about 3.5 billion years ago until the present, we can trace the....


Figure 6. Temporal relationships between the five kingdoms of organisms. (Schwartz and Margulis, 1981).

Figure 6. Temporal relationships between the five kingdoms of organisms. (Schwartz and Margulis, 1981).


[96] ....emerged. Many new sorts of cells probably originated by symbiosis, and some populations of cells that bore undulipodia with microtubules eventually evolved. If cell symbiosis theory is correct, serial microbial symbioses preceded the origin of skeletalized invertebrates with hard parts. In any case, eucaryotic cells must have evolved before animals and plants composed of this type of cell.

In fact, the dramatic appearance of calcium carbonate, phosphate, and other mineralized hard parts was probably only a very conspicuous manifestation of many other evolutionary innovations at the cell and tissue levels that are not preserved in the fossil record (Lowenstein and Margulis, 1980).

The frontispiece of Sachse de Lowenheimb's book, Oceanus Macromicrocosmicus, published in 1664, is shown in figure 7. De Lowenheimb was a champion of Joseph Harvey's concept of circulation of the blood. Apparently, literate people at that time agreed that water flowed through the environment as through a closed circulatory system. The water raining into the rivers and flowing down the mountains eventually entered lakes and oceans. Water coming down as rain after evaporation was not lost: it cycled. Sachse de Lowenheimb used the concept as an analogy. Blood leaving the heart and passing through the arteries came back through the veins; it must therefore flow in a closed system.

Now we use the analogy again, but with an extension. All scientists agree that vertebrate blood vascular systems are closed. The only time blood is in equilibrium with its surroundings is when a person or animal is dead! Furthermore, the blood is a highly modulated fluid system controlled by and for a whole functioning living system-the animal. In blood there are many regulated deviations from equilibrium, including, for example, temperature and the amounts of bicarbonate, salt, and oxygen. These deviations are purposeful in that they maintain the organism. They are products of relentless and continuous natural selection. To regulate deviations from equilibrium at the ultimate expense of solar energy is an intrinsic property of living systems.

We have named our extension of the analogy between the blood vascular system of animals and the lower atmosphere of Earth, the Gaia hypothesis (Margulis and Lovelock, 1974). The Gaia concept is that Earth's atmosphere is a highly modulated circulatory system produced by the biosphere for the biosphere. What aspects of the atmosphere are modulated? Probably the composition and related features such as the acidity. Our atmosphere is too alkaline, relative to the other inner planets, and it harbors far too much oxygen. Earth's temperature may be actively modulated as well. That certain aspects of the atmosphere arc actively regulated by life is relatively easy to show, although how this regulation works in detail is an extraordinarily difficult problem to solve. Recently, we have considered the role of anaerobic gas release processes such as methanogenesis in the stabilization of the oxygen....



Figure 7. Frontispiece to Sachse de Lowenheimb Oceanus Macro-microcosmicus (1664). (Courtesy Trustees, Wellcome Institute for the History of Medicine, London.)

Figure 7. Frontispiece to Sachse de Lowenheimb Oceanus Macro-microcosmicus (1664). (Courtesy Trustees, Wellcome Institute for the History of Medicine, London.)


.....content at its current 20% (Watson et al., 1978). Far more work needs to be done. The analogy to the blood circulatory system is again instructive. Although there is little doubt that blood ion ratios, protein composition, temperature, pressure, and so forth are not in equilibrium with the environment of the organism but must be actively regulated to support the organism, the physiological mechanisms by which regulation is achieved are still not entirely understood. It has taken more than 300 years to get this far [98] with vertebrate blood regulation. The atmosphere is even more complex and even less is known. However, until we recognize that Earth's atmosphere is out of equilibrium for a purpose, we doubt if a profound understanding of either climates or paleobiology is possible.

The composition and other aspects of the lower atmosphere are purposefully regulated in the same way that the atmosphere inside a beehive is purposefully regulated. Neither the disequilibria of the troposphere nor the atmosphere in a beehive is constructed according to a divine plan; both are products of many years of evolution. Natural selection has acted strongly to optimize complex systems by optimizing their component parts. This has led to Earth's dynamically stable yet chemically improbable Gaian atmosphere in which the temperature and oxygen content are optimal for the microbial animal and plant ecosystems on the planet's surface. Understanding how Earth's atmosphere and surface are maintained in disequilibrium is a crucial scientific goal whose achievement is aided inordinately by comparison with our flanking planets Mars and Venus. Earth's surface has apparently followed a peculiar course for hundreds of millions of years at least.

As we have discussed mud-dwelling bacteria, termite hindgut microbes, and the atmosphere of Earth, we have shown how symbioses and other interactions between organisms of very different species have been continuously important in environmental regulation. Without such interactions in which some organisms produce gas and carbon compounds that others utilize and remove, for example, Earth might also have a hot or cold dry planetary surface with a dead CO2 atmosphere. It might be much more like Mars and Venus. In fact, without the gas-exchanging microorganisms we would never have evolved at all.

We are grateful to NASA (NGR-004-025), the Boston University Graduate School, and the Guggenheim Foundation for support of this work. Some of this paper was based on an oral presentation given to the Boston University community (by L. M.) as a part of The University Lectureship, Boston University, 1978.




- Corliss, J. O.: The Ciliated Protozoa. Characterization, Classification and Guide to the Literature, 2nd edition. London and New York: Pergamon Press, N.Y., 1979.

- Cleveland, L. R.; and Grimstone, A. V.: The Fine Structure of the Flagellate Myxotricha paradoxa and Its Associated Micro-organisms. Proc. Roy. Soc. London B, vol. 159, March 17, 1964, pp. 668-686.

[99] - Cloud, Preston: Evolution of Ecosystems. Amer. Sci., vol. 62, Jan.-Feb. 1974, pp. 54-66.

- Golubic, S.: The Relationship Between Blue-Green Algae and Carbonate Deposits. In: The Biology of Blue Green Algae, N. S. Carr and B. A. Whitton, eds., Oxford, 1973, pp. 434-472.

- Hollande, A.; and Gharagozlou, I.: Morphologie Infrastructorale de Pillotina calotermitides praecox. C. R. Acad. Sci. Paris, vol. 265, Oct. 1967, pp.1309-1312.

- Horodyski, R. J.: Lyngbya mats at Laguna Mormona, Baja California Mexico - Comparison with Proterozoic stromatolites. J. Sedimentary Petrology, vol. 47,1977, pp.1305-1320.

- Knoll, A. H.; and Barghoorn, E. S.: Archean Microfossils Showing Cell Division from the Swaziland System of South Africa. Science, vol. 198, Oct.1977, pp. 396-398.

- Lowenstein, H. W.; and Margulis, L.: Calcium Regulation and the Appearance of Calcareous Skeletons in the Fossil Record. Proceedings, Third International Symposium on the Mechanism of Biomineralization in the Invertebrates and Plants. Kashikoyima, Mie, Japan,1980, pp. 289-300.

- Margulis, L.: Symbiosis in Cell Evolution. San Francisco: W. H. Freeman Co.,1981.

- Margulis, L.; Barghoorn, E. S.; Banerjee, S.; Giovannoni, S.; Francis, S.; Ashendorf, D.; Chase, D.; and Stolz, J.: The Microbial Community in the Banded Sediments at Laguna Figueroa, Baja California: Does It Have Precambrian Analogues? Precambrian Research, vol. 11, 1979a, PP. 93-123.

- Margulis, L.; Chase, D.; and To, L. P.: Possible Evolutionary Significance of Spirochaetes. Proc. R. Soc. Lond., vol. B204, April 11, 1979b, pp.189-198.

- Margulis, L.; and Lovelock, J. E.: Biological Modulation of the Earth's Atmosphere. Icarus, vol. 21, April 1974, pp. 471 -489.

- Margulis L.; and Lovelock, J. E.: The Atmosphere as Circulatory System of the Biosphere: the Gaia Hypothesis. Co-Evolution Quarterly, Summer 1975, pp. 30-40.

[100] - Margulis, L.; To, L.; and Chase, D.: Microtubules in Prokaryotes. Science, vol. 200, June 1978, pp. 1118-1124.

- Mazur, P.; Barghoorn, E. S.; Halvorson, H. O.; Jukes, T. H.; Kaplan, I. R.; and Margulis, L.: Biological Implications of the Viking Mission to Mars. Space Science Reviews, vol. 22, June 1979, pp. 3-34.

- Reimer, T. O.; Barghoorn, E. S.; and Margulis, L.: Primary Productivity in an Early Archean Microbial Ecosystem. Precambrian Res., vol. 9, 1979, pp. 93-104.

- Schidlowski, M. P.; Appel, W. U.; Eichmann, R.; and Junge, C. E.: Carbon Isotope Geochemistry of the 3.7 x 109-year-old Isua Sediments, W. Greenland: Implications for the Archean Carbon and Oxygen Cycles. Geochim. Cosmochim. Acta, vol. 43, Feb.1979, pp. 189-199.

- Schopf, J. W.: Precambrian Paleobiology: Problems and Perspectives. Ann. Rev. Earth Planet. Sci., vol. 3,1975, pp. 213-249.

- Schwartz, R. M.; and Dayhoff, M. O.: Origins of Prokaryotes, Eukaryotes Mitochondria and Chloroplasts. Science, vol. 199, Jan. 1978, pp. 395

- Schwartz, K. V.; and Margulis, L.: Phyla of the Five Kingdoms: a Guide to the Diversity of Life on Earth. San Francisco: W. H. Freeman Co.,

- Stolp, H.: The Interaction Between Bdellovibrio and Its Host Cell. Proc. R. Soc. Lond., vol. B204, April 11,1979, pp. 211 -217.

- To, L. P.; Margulis, L.; and Cheung, A. T. W.: Pillotinas and Hollandinas: Distribution and Behavior of Large Spirochaetes Symbiotic in Termites. Microbios, vol. 22,1978, pp. 103-134.

- Walter, M. R., ed.: Stromatolites (Developments in Sedimentology, vol. 20). Amsterdam: Elsevier, 1976.

- Watson, A.; Lovelock, J. E.; and Margulis, L.: Methanogenesis, Fires and the Regulation of Atmospheric Oxygen. BioSystems, vol. 10, Dec. 1978, pp. 293-298.