Genesis of Eden

Genesis Home

The Gaia Hypothesis James E. Lovelock

When James Lovelock was first working out his Gaia concept of the physicochemical regulation of the Earth's surface, he visited the Environmental Evolution class several times. The disparate data and observations that led him to suggest a new view of life and the environment are revealed in this early lecture (from 1973) and in his responses to questions posed more than 10 years later. We have presented them here both for historical interest and to illustrate the process of the development of scientific thought.

What has come to be called the Gaian view considers the atmosphere to be an integral part of the biosphere; the atmosphere here is not just a separate physicochemical system which interacts passively with life on Earth. Many will regard this as mere speculation, but I will try to prove that it is not. Even if I fail in this attempt, I think you will find that the Gaian view elicits new questions which otherwise might never have been asked.

Where are we? The small thatched building depicted iii the foreground of figure I is my former laboratory at Bowerchalke in South Wiltshire. Built over 400 years ago, it is, as far as I know, the only thatched space laboratory in the world. I am not introducing this territory as a bit of cozy folk science; the environment, as always, entails the organism.

The ground on which this laboratory was built is a well-known inorganic chemical substance calcium carbonate, or limestone: CaCO3. The air-which, as you can see, is quite clear-is made up of oxygen (02), nitrogen (N2), carbon dioxide (CO2), water vapor (H20), and other inorganic chemicals. The plant segment of the biosphere colors the biosphere green. The animals are not so conspicuous, but they exist. Life amid inanimate surroundings, life in its inert environment: this is the ordinary, traditional view of the biosphere. By contrast, the picture I'm developing is that the air and the ground are not independent inorganic chemicals, but that the sediments and atmosphere are part of a living system. On the other hand, despite its seeming position as part of pristine, unspoilt nature, even the green that surrounds my laboratory is not "natural": it is nearly all man-made, biology wittingly or unwittingly ordered by man over the course of time. Without people, this part of southern England would probably be primeval scrub forest. From this point of view, air pollution on a global scale might perturb not just the atmosphere but the biosphere. The possibility of air pollution on a global scale began to attract professional interest in the early 1970s. My interest was sparked by the reports Man's Impact on the Global Enz)ironment and Inadvertent Climate Modification, published by the MIT Press in 1970 and 1971. When I read them, it occurred to me that a special view of the Earth was denied to the distinguished groups that produced the reports, a view that was more consonant with those of my colleagues working in the planetary sciences. This special view came from a need to look at the planet in its entirety, and in an interdisciplinary manner. By contrast, the view of these otherwise wholly excellent books is limited by the division of science into arbitrary disciplines. For example, the meteorofogists state explicitly that they do not at all consider the chemistry or the biology of the Earth. The atmospheric chemists, for their part, say that meteorology lies beyond their territory; they make no reference to biology. This recognition of territorial rights is instinctive to most male animals, and this includes scientific experts and university professors especially. Such a comment may seem glib, yet it accurately accounts not only for the limitations of the MIT reports but also for those of current science in spite of the growing interest in and sympathy toward more interdisciplinary approaches.

Earth's Atmosphere: Evidence for Life

The problem of detecting the presence (or, niuch more likely, the absence) of life on other planets demands a less divided view. The search for life elsewhere brings biologists and engineers, for example, together in constructive conversations. For some years I have, with colleagues such as these, been interested in the possibility of detecting the presence of life ()n other planets merely from the knowledge of the chemical composition of the planet's atmosphere. I hypothesized that Mars would be without nitrogen, since it was probably without life; at least, this is what we surmised after seeing the cratered, moon-like surface revealed by the first Mariner mission. To understand the reasons for this educated guess, let us consider the relationship of nitrogen in the atmosphere to the presence or absence of life on Earth.

The cosmic abundance of elements is fairly constant, and (apart from an absence of hydrogen, which may have escaped to space) Earth is fairly representative of the general distribution of elements. These elements tend to combine to the state at which the lowest potential energy is reached; this is a law of chemistry. Comparison of the major constituents of Earth's atmosphere with those of Mars and Venus (table 1), however, reveals that Earth's atmosphere is anomalous with respect to these gases.

Table 1 Major features of the planetary atmospheres compared: percent by weight of the reactive gases carbon dioxide, nitrogen, and oxygen; water in precipitable meters over the planet if all vapor precipitated out of the atmospherelpressure; and mean annual surface temperature.


Venus Earth Mars

C02 98% 0.03% 95% N2 1.7% 79% 2.7% 02 trace 21% 0.13% H20 0.003 m 3,000 m 0.00001 m Pressure 90 bars I bar 0.0064 bar Temperature 477'C 17'C -47C


With the chemical mixture present in the Earth's atmosphere, the element nitrogen is expected to form its most stable compound, which is not N2 but the nitrate ion (NO3 ). One would expect N03 to be present either on the surface or in the seas as a potassium or sodium salt. Conversion of the nitrate ion to nitrogen gas is an "uphill" process which requires the presence of life. The expected chemical conversion of nitrate makes the presence of nearly 80 percent molecular nitrogen in Earth's atmosphere an indication of the presence of life.

The next Mariner mission found only 2.7 percent nitrogen in the Martian atmosphere. The large amount of nitrogen on Earth relative to Mars supported my view that life on Earth shows itself as a global chemical phenomenon.

When I was a young man, the atmosphere of Earth was said to have originated in the primeval outgassings from the Earth's interior; outgassing is a process, particularly important in the early history of the Earth, whereby gaseous and volatile compounds escape from the interior of the planet and help produce the early oceans and atmosphere.

Oxygen, for example, was explained to have come from the photo-dissociation of water vapor by sunlight followed by hydrogen escape. Nitrogen was said to have been a stable, inert constituent throughout Earth's history. Until a few years ago, this view was largely unchallenged. Indeed, it is widely held, and you will still find it in many textbooks on the Earth's atmosphere. In spite of the distinguished science done by such workers as L. V. Berkner and L. C. Marshall, who proposed a wholly biological origin for atmospheric oxygen, most aeronomists still believe that life, responsible for no net increment of oxygen, merely recycles oxygen gas. Oxygen, that is, just happens to be in the air at exactly the right concentration for most life because of blind inorganic processes. Or, if it is your preference, oxygen was arranged by some beneficent providence. To me this view of the air as a product of wholly inorganic processes is the most magnificent nonsense; indeed, one of the most intriguing puzzles in the history of science is how it has managed to persist for so long. G. N. Lewis and M. Randall in the 1920s, G. E. Hutchinson in the 1950s and, most recently, L. G. Sillen all showed that at the pH and the redox potential of the Earth molecular nitrogen is thermodynamically unstable. The element nitrogen, given these conditions should be present on Earth not as the gas but as nitrate ion in seas. From aii inorganic viewpoint, the presence of oxygen is, in fact, equally anomalous when Earth is compared with its neighboring planets, Mars and Venus, which lack atmospheric oxygen. That Mars and Venus have no free oxygen is not unexpected. It is, rather, the presence of free, extremely reactive atmospheric oxygen on Earth that is a complete anomaly. The once-reducing atmosphere of the Earth is today an oxidizing one.

The opposed terms reducing and oxidizing, roughly equivalent to "hydrogen-rich" and "hydrogen-poor," come from old-fashioned chemistry. A reducing substance, such as hydrogen, will combine with the oxygen of a metal oxide, such as iron oxide (rust), to give the oxide of the reducer (in this case water) and also free iron as the metal. In a reducing atmosphere, iron would remain as metal. In an oxidizing atmosphere, it rusts-that is, iron recombines with oxygen to give its oxide again.

Another shibboleth which has held up progress in this branch of science is that the climate and the chemical composition of the Earth are uniquely favorable for life. This is the one you will find in most science fiction stories, sad to sav. I say it is sad because, on the whole, science fiction tends to be a little less blind that science itself. Indeed, it is not commonly appreciated that seemingly quite small changes would render Earth unsuited to contemporary life.

Life Regulates the Environment

What would happen, for example, if oxygen were to increase in concentration? If it were just over 25 percent, the probability of a fire starting by a lightning flash would be so high that even tropical rain forests would be at risk. For each 1 percent increase in oxygen concentration over the current level, the probability of initiating combustion doubles. Interestingly, if the oxygen concentration were to fall to 13 percent one could not start any sort of fire at all. Nearly no difference exists in flammability between 02 concentrations of 40 percent and 100 percent. Indeed, the current 21 percent 02 concen tration is just about ideal for the existence of trees, which participate in making oxygen: at higher levels trees would be burnt up; at lowe ones there might well be too few animals and other oxygen consum ers to maintain an ecological balance.

For another example of a quite small change that would have drastic consequences for present life forms, consider that a change in atmospheric pressure of merely 10 percent, assuming that the composition of air was unaltered, would result in a 4'C change in the world's mean surface temperature! Such a change would set Earth on a highly unfavorable climatic course. These examples, which show just how well suited the present atmosphere is to the present for of life on Earth could be multiplied. I think the biota, the sum of al living organisms, interacts actively with its environment so as to maintain the environment at values of its own "choosing." The notion that blind chance led to such a perfectly adjusted atmosphere by contrast, seems untenable.

Early in its evolution, life acquired the capacity, I believe, to control the global environment to suit its needs. The capacity for environmental maintenance has persisted; it is still active. The sum total of all the species that go to fill up the biosphere is far more than just catalog; like other associations in biology, this global biota is an entity with properties greater than the sum of its parts. Such a large creature, with the powerful capacity to air-condition the whole planet may be only hypothetical at the moment; nonetheless, it needs name. I am grateful to Mr. William Golding, who lives in my village for the suggestion of the word Gaia: the Greek personification o Mother Earth. It has various advantages, not least of which is it status as a four-letter word with the capacity to focus the attention of my scientific colleagues; certainly "Gaia" is a lot easier to say than "a biological cybernetic system with homeostatic tendencies."

Any theory stating that the Earth's surface is wholly a product of biological processes must be considered wrong. Comprehensive biological explanations of atmospheric composition are difficult to formulate. At the root of this problem is the fact that there exists no formal scientific statement of life as a process. I know of no single and exclusive test that could prove or disprove the existence of Gaia as a living entity. Biologists, fortunately, usually are not deterred by such a lack of rigor. Even if eventually prepared, such a formal physical statement of life is liable to be statistical, mechanical, very mathematical, and quite unsuited to the design of simple experiments to test for the presence of Gaia.

Most biologists, indeed most people, upon seeing a giraffe even for the first time, and especially if it moved, would be able to pronounce it alive without any conscious use of chemistry or physics. Life is still much in the realm of phenomenology. One scientific approach attractive to interdisciplinary biologists would attempt to prove the presence of life by seeing if the entity tested were able to maintain a constant temperature and compatible chemical composition in the face of environmental change or perturbation. From such a phenomenological basis, what evidence points to the existence of Gaia, a creature made up of the biosphere but more than just the sum of its parts?

We believe that during the period in which life has existed on Earth, from at least 300 million years ago to the present, the reduction-oxidation (redox) potential of the atmosphere has shifted from a pE of -5 (reducing) to a pE of +13 (oxidizing). (A useful way of indicating the oxidizing or reducing tendencies of a system, pE, the logarithm of the reciprocal of the electron concentration, is expressed in gram-molecules per liter.) During the time in which the pE changed from -5 to +13, the atmospheric composition changed concomitantly. The change from a reducing to an oxidizing atmosphere led to many correlated changes in atmospheric composition and total pressure. The early atmosphere was probably rich in hydrogen and ammonia. At the same tinie, as it moves along the standard course for average stars, the energy output of the sun has increased at least 30 percent. This increase in solar luminosity is one of the few relatively certain facts of astronomy. In spite of extensive atmospheric chemical changes and changes in output of radiant solar energy, the geological record, with its demonstrable persistence of life, indicates that at no time in the last 3 billion years did the Earth's mean temperature change more than a few degrees from what it is now. What sort of remarkable coincidence might account for such physical constancy, which is exactly what is required for the continued existence of life? Indeed, I am doubtful that this is coincidence; I think it very much more likely that a biological regulatory system has been and is working, ensuring planetary homeostasis at physical and chemical states appropriate to the global biota at a given time.

Life Regulates Global Mean Temperature

The most important evidence for Gaia is found in the constancy of the Earth's mean temperature through time. Since liquid water has always been present, the average temperature is unlikely ever to have exceeded 50'C or to have decreased much below the mean temperatures during the Pleistocene ice ages. Our sun, like all main sequence stars, has been increasing in luminosity since its origin. Near the origin of the Earth and life, some 4000 million years ago, the sun is thought to have been 30 percent fainter than at present. Quite clearly, even a comparatively short time ago, the Earth's temperature if it responded passively to solar luminosity in the same atmosphere would have been much lower than it is now. This paradox-the fact that the solar output was much weaker in the past, and yet Earth' mean temperature seems to have remained within certain boundaries-has been called "the Faint Young Sun Paradox."

To maintain a constant surface temperature, assuming current atmospheric composition, a decrease in atmospheric pressure would have to have occurred over time. On a mountain at a point wher the pressure has dropped by 10 percent, you will find that the temperature has fallen by 4'C. Similarly, if you descend 1,000 feet below sea level (as in a deep depression at the Dead Sea), you will find a equivalent rise in temperature. A 10 percent increase in atmospheri pressure corresponds to a rise in temperature of about 4'C. Meteorologists refer to this as the adiabatic lapse rate. Atmospheric pressure would have had to decrease to compensate for the increasing output of the sun with time. What atmospheric composition provided higher atmospheric pressure in the past, leading to maintenance of constant temperature over geologic time? A greater amount of oxygen in the past is precluded. Even small changes in oxygen levels would render disastrous effects. Drastic C02 rise is precluded by inorganic equilibria in water which hold this gas fairly near present level. The only gas that changes easily is nitrogen. About 1.5 billion years ago, about 20 percent more nitrogen may have been in the atmosphere than now. We can check this by seeing if younger rocks contain more nitrogen than older ones, since some of this extra nitrogen would have to have been buried to reduce the amount of nitrogen in the younger atmosphere. In fact, younger rocks do, I believe, by and large, contain more nitrogen than older ones. I view this as one possibility of "gaian" pressure and hence temperature regulation of the atmosphere, via manipulation of the quantity of atmospheric nitrogen. But affairs are more critically balanced. If the mean surface temperature of Earth falls by more than 2 or 3'C, then positive feedback mechanisms associated with the increase of snow cover hasten the failing temperature, as is well established in meteorology. The Earth cools to a point where even the oceans may freeze. Earth's temperature is poised between hot and cold extremes. A 2 or 3'C rise would increase greenhouse gases in the atmosphere and, by a similar positive-feedback mechanism, cause further increased heating. With a sufficient rise in temperature, a thermal "runaway effect" would occur. Yet the temperature has remained relatively constant in the face of perturbation. To me this is the best evidence of the existence of Gaia: a system that controls the conditions at the surface of the planet. If all of the life on Earth were deleted at one stroke, many inorganic atmospheric reactions-from electrical discharges and ionizing radiation to solar ultraviolet rays-would permit oxygen and nitrogen to react with one another. A predominant end product of these reactions Of 02 with N2 would be the nitrate ion, which would wash into the sea. Nitrogen so removed would not return to the air: biological processes alone perform nitrate reduction to molecular nitrogen. Whereas the burying of nitrogen from the sea in plate-tectonic processes involves relatively slow recycling, we are concerned here wit relatively short periods of about a million years. Small quantities oxygen might be supplied by water or C02 photolysis (the breakup of carbon dioxide via lightning), ensuring the removal of the remaining nitrogen after the last of the oxygen has reacted. The end result of the steady-state inorganic equilibrium, in the absence of life, woul be an atmosphere Of C02, water vapor, carbon monoxide, and rare gases; only trace amounts of oxygen and nitrogen would persist. The scenario for the chemical composition of the atmosphere of a lifeless Earth is a very reasonable interpolation between the atmospheres present-day Mars and Venus. The redox Earth, compared with its imaginary sterile twin, as well as oxidizing Venus and Mars on th one hand, and reducing Jupiter on the other, is portrayed in the bar graphs in figure 2.

Biotic Emissions Have Global Effects

Now let us consider: Do the activities of man have atmospheric effects, adverse or otherwise, on a global scale? The various gases the atmosphere, their concentrations, their emission rates from the biosphere, their atmospheric residence times in years, and their emission rates from man-made sources are shown in table 2. The first thing you will notice when you look at this table is the overwhelming dominance of the biota. The gases turn over quickly most of them have short residence times. Oxygen, for example although fully 21 percent of the total atmosphere, turns over once every thousand years. The residence time for N2 is measured on the order of a million years-longer than other elements but still short compared to the time that life has been present on Earth. Carbon monoxide and oxides of nitrogen are produced not solely from the exhaust of cars but by the biota as well. Many creatures in the sea find carbon monoxide an essential part of their everyday business Siphonophores float using little bladders filled with 80 percent carbon monoxide. Even Fucus, a common brown seaweed, has 0.1 percent carbon monoxide in its bladders. Carbon monoxide is not a noxious toxic emission, but part of the vital existence of these marine organ isms. At least 1.2 billion tons of ammonia are annually produced by the biota. I say "at least" because the figures are probably underestimates; living organisms as new sources of gases are continually discovered.

Figure 2: Concentrations of reactive gases in the atmospheres of the inner planets of the solar system on a log scale (from 10 ' to 100%). From most reducing (left) to most oxidizing (right): hydrogen, methane, ammonia, nitrogen, carbon monoxide, hydrogen chloride, carbon dioxide, oxygen. "Sterile Earth" was calculated by interpolation between Mars and Venus or by letting "Earth" gases react with each other to chemical equilibrium.

This very large production of ammonia has an important bearing on the maintenance for life of planetary pH. I think biogenic ammonia compensates for the tendency of the planet toward acidity. In regions such as northern Sweden, the northern United States, and Canada, ammonia production is apparently deficient; rain as acid as pH 3 sometimes falls.

Methane is produced in quantities of more than 2 billion tons per annum. I believe the huge production of methane, representing some 8 percent of all of the energy of photosynthesis, is very important for the regulation of the atmosphere. One function of methane is the formation of a very convenient "molecular balloon"; this gas passes the atmosphere to regions where photolysis occurs and hydrogen escapes. Water vapor does not easily enter the upper atmosphere: the tropopause is very cold, and water vapor freezes out to a concentration of 0.5 ppm. Methane, in the same region, has a concentration of 1.5 ppm. Since each methane (CH4) molecule carries twice as much hydrogen as water, six times as much hydrogen is carried up and outward by methane as by water. In addition, the Earth maintains a net oxidizing state by expelling hydrogen to space. If methane is involved in oxygen regulation, a mature ecosystem is justified in squandering as much as 8 percent of its energy to produce it.

The atmosphere is not a static mixture of gases preserved by Earth's gravitational field. It is a system in dynamic and contemporary balance (table 2). Undoubtedly the immediate origin of the atmosphere is the biota. I am certain that the relative constancy of atmospheric composition over time is actively maintained by sensing and control mechanisms within the biosphere. The human industrial sector, with the possible exception of carbon dioxide emission, contributes relatively little.

Can large-scale atmospheric effects, for example combustion emissions, be used as perturbations to test for Gaia? We first notice air pollution on a large scale by the presence of smoke haze, such as the atmospheric turbidity of a locale such as the British industrial city of Sheffield, a small steel town. By contrast, Bantry Bay in southwestern Ireland is far removed from industry and pollution sources. When the wind blows from the North Atlantic, the air is sparklingly clear.

Visibility at ground level may be more than 40 miles. When the wind blows into Bantry Bay from continental Europe, a source of all sorts of pollution, the picture changes; the visibility range may be reduced to fewer than 1.5 miles. Alan Eggleton and his colleagues at Harwell Laboratory inform us that the rural summertime haze of northern Europe, like that at Sheffield, is a form of photochemical smog frequently associated, like Los Angeles smog, with comparatively high levels of ozone. When the wind blows from continental Europe, a level of about 0. I part per million of ozone would be expected during the day, to be compared with about one-tenth of that in clear air conditions.

Atmospheric turbidity is a measure of the scattering of incoming sunlight by any particulate matter in the air. Dust particles are blown from the desert by wind or from farm lands under dust bowl conditions. But particulate matter, including fine droplets, may also be produced by reactions among the gases in the atmosphere. Sulfur compounds react with oxidants to produce sulfuric acid droplets or ammonium sulfate aerosols. Trees produce unsaturated hydrocarbons like pinene (a terpenoid; see chapter 14), which polymerizes to form pinene polymer particles, an aerosol that scatters incoming sunlight. Dust particles can be generated ill situ; they need not have been stirred up. Atmospheric turbidities plotted monthly at Bowerchalke (southern England), Greensborough (North Carolina), and San Bernardino (in the heart of the California smog basin) are indistijiguishable. They all show the same high-level seasonal increase. Few people realize that the rural regions of southern England, the Appalachian Mountains, and Los Angeles all share the same density of turbid aerosol in the summertime. Though this turbid aerosol is of photochemical origin at all three sites, the haze at Bowerchalke and Greensborough is quite different from that of Los Angeles. At its worst in southern England or Greensborough, there is little or no odor or eye irritation; the aerosol appears to be principally composed of natural ammonium sulfate or sulfuric acid droplets.

A puzzling feature of smog in rural regions away from industry is that it is most marked with winds from directions which also lack industry. In southern England the densest smogs come with winds from a southeasterly quarter, which traverse many miles of open sea and rural areas before arriving. By contrast, air masses coming from the north over the densely populated industrial regions of the United Kingdom are comparatively free of this sort of haze. Seasonal change in atmospheric turbidity can be measured for air masses from the principal directions at Bowerchalke. Continental air masses have high levels of summertime haze. Air masses coming from a maritime tropical direction off the southern part of the Atlantic, or a maritime polar direction coming from the north, either over the sea or over the industrial regions of the United Kingdom, carry little smog. No significant smog occurs with winds from northerly directions! The densest turbidity in the United States, too, is found not in the densly populated industrial regions, but in the southern Appalachian Mountains, where urbanization is light but trees abound. A compilation of the areas of turbidity in the northern hemisphere can be made from satellite photographs. The regions of denser turbid aerosol are not the industrial regions of western Europe, the United States, or Japan, but the tropical and desert regions of the world near the equator! The whole of Africa in its middle region lives in a state of permanent haze; this appears to be true also of much southeast Asia. A sunset on the harbor at Dakar, Senegal, has the same golden look seen in Los Angeles. The Senegal haze, however is not industrial, for Dakar is quite a small city with little industry rather, it is the natural haze of Africa. The Canary Islands, photographed from a ship 5 miles offshore, can barely be seen, so dense is the turbidity. Whatever causes this atmospheric turbidity potentially affects the Earth's surface temperature in one way or another. Haze and atmospheric particulates come from the nonhuman biota as well as from industrial and domestic sources. In the last century the concentration of atmospheric carbon dioxide has increased as a consequence of the ever-increasing burning fossil fuels. It has risen from 280 parts per million to 350 parts million, and the increase is beginning to accelerate. This increment corresponds to the retention in the atmosphere of about half of the carbon dioxide from fuel that has been burned. This increase anticipates a warming of the planet by the well-known greenhouse phenomenon: the absorption of infrared radiation by carbon dioxi which lessens the heat loss of the Earth into space. These atmospheric changes may be due largely to human active or maybe not. Changes in climate might be associated with the

Figure 3 Changes in global mean temperatures since 1860.

measured atmospheric changes. Since about 1925 the climate of the northern hemisphere has changed, and up until the winter of 1970-71 the change had continued in the same direction (figure 3). In the second of the MIT publications I mentioned (Inadvertent Climate Modification), a diagram shows the reduction in temperature in the northern hemisphere plotted against latitude for the period 1960-1965 and the period 1965-1971. A clear-cut, quite unequivocal decrease in temperature occurred for all regions higher than about latitude 50' north. A slight warming trend may have occurred near the equator, but it is nowhere near as marked as the apparently continuing decrease in mean temperature in the northern regions. A reverse of the warming trend seems to have occurred. Since this time, however, average global temperatures seem to have been increasing. Averages, however, are often misleading. A meteorologist friend, Dr. James Lodge of the National Center for Atmospheric Research in Boulder, Colorado, continually reminds himself of this fact with a cartoon on the wall of his office. This cartoon shows a man with his head in an oven and his backside in a refrigerator and a rather dim technician measuring the temperature of his navel. The caption reads "His average temperature is fine, Doc." So it is with the world. The world's mean surface temperature may not have changed much, but large fluctuations are evident from year to year on a local scale. Not only is the mean surface temperature in the northern hemisphere changing; other properties are also changing. One is the frequency of westerly wind drift across the North Atlantic. Hubert Lamb, of the British Meteorological Office, has compiled records of westerly wind frequencies going back to as early as records were taken in the past century. The decline of the westerlies in the early 1970s was greater than had ever been observed before. The decline in temperature at this time and the decline in westerly winds would expectedly be associated because the warm winds traversing the Atlantic toward the Arctic Basin are westerlies, and if such winds become infrequent, northern hemispheric temperatures, at least in the regions above the British Isles, become lower over time. The extent of these climatic changes seems significantly greater than "the climatic noise level." Explanations fall into three categories. The first is "natural/organic." Some authorities suggest that the decline in northern hemispheric temperatures since World War 11 and through the mid 1970s is largely attributable to increased volcanic activity. Others attribute changes to solar output. Volcanic dust veils had occurred, but fewer were present then than in the nineteenth century. They seem to me insufficient to account for this episode of rapidly declining temperature. The decline was occurring considerably earlier than the most severe of the volcanic dust emissions, e.g., from Mount Agung in Bali (1963-64). The sun's output, which changes little over the sunspot cycle (11 years), is quite insufficient to account for surface changes on Earth.

Another category of explanations is anthropogenic; it includes factors such as carbon dioxide from fossil-fuel emissions, jet-aircraft contrails (which reflect sunlight back into space), and human dustraising activities in general, including farming, slash-and-burn agriculture in the tropics, and the production of aerosols as a result of the burning of sulfur compounds. 'anthropogenic changes are our own contributions to the environment, and climatic changes may be a response of the global biota to anthropogenic perturbations. Controversy is intense around the issue of possible effects from industrial, domestic, and agricultural activity. By the early 1970s more than 15 billion tons of carbon dioxide had been injected into the atmosphere, and at least half of this increment still circulates. The rate of increase of carbon dioxide concentration has accelerated so much that it cannot be accounted for merely by emissions. Meteorologists are in unanimous agreement that C02 increase leads to a warmer climate, especially in northern latitudes. Exactly the reverse could happen, however. Both dust and haze reflect sunlight back into space, reducing the heat received from the sun and causing the Earth's surface to cool. Cooling has occurred, but this simplistic explanation does not hold. Hazes in the troposphere absorb more energy than they reflect; a haze in the stratosphere may warm that region also by absorbing energy, but this results, by a strange meteorological sequence, in the cooling of the Earth's surface such as happens when volcanoes inject haze into the stratosphere. Haze near ground level should have a slight warming effect. Almost every man-made effect, from jet contrails to carbon dioxide emissions and haze generation, should cause a rise in mean surface temperature. Obstinately, though, the temperature during this time fell. The Challenger expedition in the 1950s showed that biological systems readily add methyl radicals to a wide variety of elements, including sulfur. I thought that dimethylsulfide (DMS) might be the major biological sulfur compound emitted into the atniosphere. Challenger had already shown that marine algae and certain land plants , emit dimethylsulfide. The production of dimethylsulfide is ubiquitous in the biosphere. Marine algae, soil, and almost all plants emit it. The output of DMS is strongly light-dependent; it may be the missing component of the natural sulfur cycle. lt is far more stable than hydrogen sulfide, and so could survive transfer to the stratosphere, where it might oxidize to give sulfate and perhaps a methane sulfonate aerosol. Such an aerosol is well worth seeking; if found, the source will be biological. Another product of the biota tantalizingly suggestive of important atmospheric change is nitrous oxide (N20). N20 is emitted, mostly by soil microorganisms, at the huge rate of 2.5 billion tons per year. The concentration of N20 in the troposphere is about 0.5 ppm. In the stratosphere the destruction of N20 by solar ultraviolet light leads to the production of, among other things, nitric oxide (NO), which modifies ozone production. Nitrous oxide production may affect thedensity of the ozone layer; it may be another biological climate modifier (figure 4).

Figure 4 Nitrous oxide (N20), produ:ed by soil microorganisms, can modify the ozone layer, causing changes in atmospheric circulation patterns.

Regarding recent climatic change and the possible responsibility of man's activities, facts are few and opinions many. But while I think it will be a long time before the complete system that determines the climate is understood, the answers are unlikely to be found if the biota is neglected. The biota has continued to survive and modify its environment for over 3 billion years. Changes which have occurred since the evolution of Homo sapiens, and especially more recently since the industrial revolution, seem huge to us; the consequences seem dire and immediate. But devastating environmental change due to rapidly growing populations of young species is a recurring theme of the evolution of life. A strong response of exponentially increasing organisms to their own "pollutants" has happened before. The enforcement of gas-emission standards on internal-combustion engines, the building of "biospheres" by people in Arizona and in the USSR, the announcements both by NASA and by Earthwatch of their plans for "missions to Planet Earth," and the great rise of environmental concerns on the part of industry and educators all begin to exercise a negative feedback on the tendency of our global human population to make its immediate environment unfit for our kind of life. This behavior is typically "gaian"; indeed, the hope I want to leave you with is that Gaia, in fact, truly exists.

Gaia: What's New?

The first presentation on the Gaia hypothesis for our Environmental Evolution course was made in 1973. Now everyone wants to know what's new. What is its current status?

A great deal of new evidence has accumulated. We now have a quite plausible theoretical model of the way Gaia works. A good example of this is the close coupling between carbon dioxide and climate regulation. There is little doubt that increasing carbon dioxide in the atmosphere increases the absorption of outgoing infrared radiation through the greenhouse effect and tends to warm the planet. The evidence for this has strengthened over recent years. Exciting new evidence in the last couple of years has come mostly from analysis of ice cores from Greenland and Antarctica. A geologic record of the carbon dioxide concentration of the atmosphere from the present to tens of thousands of years into the past (well into the last glaciation) can be documented from gas samples from ice taken at different depths. The exciting aspect of this new information is the correlation between C02 and temperature. During the last glaciation, the C02 concentration fell to somewhat below 200 parts per million; temperature-CO2 correlation is very close to that of the model predictions made beforehand. Even more exciting: at the end of the last glaciation the carbon dioxide concentration rose close to its present value in a period just short of a hundred years! Geophysical and geochemical processes that presumably control atmospheric C02 concentration cannot operate at that speed. In my view, this change in C02 concentratioii was a consequence of the growth and change in populatiori density of the nonhuman biota.

What is the current status of your theoretical work on the Gaia hypothesis?

When the hypothesis was introduced, we felt some biological system must regulate the chemical composition and climate of the planet, but we did not know how. Most of us imagined a very complicated, intricate affair. I thought it might involve something built into the genetic structure of organisms. I am now happily confident that Gaiaii regulation is a natural and simple consequence of intrinsic properties of life on this planet.

Three fundamental aspects of life determine its tendency to establish a Gaian regulating system. The first is the important fact that life on Earth is strongly constrained by its environment. Life does not flourish when water is frozen, nor when conditions are too hot. More favorable conditions exist between these extremes. Such a constraint applies to all manner of other planetary variables, like acidity; life will not flourish if conditions are either too acid or too alkaline. Life prefers neutrality. Ocean salinity is another constraint. If water is too saline, life cannot grow; if water is so fresh that nutrient salts are lacking, equally life will not continue. Between the extremes lie the best growth conditions. The most important property determining Gaia is the existence of life in the universe of constraints. The second crucial fact is the tendency of all life to grow exponentially whenever or wherever a niche is open, and whenever the environment becomes favorable. The third property relates to diversity. When different organisms emerge, they use opportunities when a new niche opens in different ways, or exploit old niches when other organisms fail to occupy it. So my theoretical approach is based on life's tendency to grow exponentially, limitations to this growth, and organismal diversity. From these assumptions my colleagues and I have been able to produce a simple model of the workings of Gaia. The numerous forms of life interact with their environment in an unbelievably intricate manner in the real world. lt is quite impossible, even with the largest computers available, to adequately build a model of the entire world. However, we can investigate the Situation by a process of reduction. I have reduced the environment to a single variabletemperature-and the species to a single type-a daisy plant-in order to produce an imaginary world, Daisyworld.

Imagine a planet very like the Earth in many ways, although with less ocean. It is well watered, and plants grow almost anywhere on its surface. It also has a very clear atmosphere, uncomplicated by clouds or greenhouse gases. The surface temperature of the planet is very dependent upon one property only: its albedo, i.e., reflection of sunlight back into space. This imaginary planet, Daisyworld, is at the same orbital distance from its star, identical to our own sun, as the Earth is froi-n the sun. Daisyworld's sun shares a universal property: with age it grows warmer and its output of heat increases. I want to demonstrate liow the temperature of Daisyworld varies with and without life as itS SLIII increases its ()utput of heat. The relationship between the growth of daisies and temperature is represented as a parabolic curve. Growth begins at a temperature of about 5'C and increases steadily to a maximum at about 20'C, room temperature. As the temperature rises beyond that, the growth rate declines until all growth ceases at a temperature of 40'C. This choice of growth curve is not arbitrary; it adequately describes growth as a function of temperature for most vegetation. Growth of daisies has an effect on the environment of Daisyworld. A "s ' )ecies" of daisy that is light-colored and reflects sunlight tends to lower the planetary temperature. When a maximum number of light-colored daisies covers a large proportion of planetary area, the temperature of the planet is at its lowest. Conversely, when very few or no light-colored daisies exist, the planet is darker because of lack of the light reflection tendency, and the planetary temperature is much higher. Small temperature changes take place on Daisyworld depending on the population of daisies. The daisies have the capacity to regulate planetary temperature. We can simulate a lifeless world by holding the daisy population constant, not allowing it to vary. In this case large temperature changes occur with changes in solar output. A lifeless planet does not have the capacity to lessen changes in temperature brought about by increases in solar luminosity. The mean temperature of a lifeless planet can be compared with one inhabited by dark daisies. Assuming that Daisyworld were a lifeless planet covered with just bare rocks, mean planetary temperature would increase as solar luminosity increased. Temperatures would rise from below freezing to about 50-70'C as the solar output steadily rose. Growth of dark daisies would have quite a different effect on temperature as solar luminosity rose. As the planetary temperature reached 5'C, dark daisies would start to grow. Imagine dark daisies that start to grow; because the stand is dark, the daisies will be warmer than their environment. This results in more growth and a little faster spreading. Before long dark daisies will cover a whole area; the temperature of that area will be warmer. This warmth adds to the extra warmth of the daisies. So, with positive feedback, daisy growth rapidly explodes until dark daisies cover a sizable proportion of the planet. The planetary temperature zooms up to close to the most favored value for daisy growth (20'C). It does not continue to rise because daisy growth is discouraged when temperature rises too high, so the planetary temperature remains more or less constant over a wide range of solar luminosity. Dark daisies alone can regulate planetary temperature to a considerable extent.

We now add light daisies and see the effect of competition between two different daisy types. Light daisies prefer to grow at warmer temperatures than dark daisies. When the mean temperature of the planet reaches about 5'C daisy growth commences. The dark daisies grow like mad until the temperature has risen to just above the optimum for growth of dark daisies on the planet. The planet is now warm enough for light daisies to grow, in spite of the fact that their tendency to reflect sunlight makes the environment somewhat cooler. This early stage, favorable for dark daisies, is less favorable for the light ones. As the sun warms up further, the two daisy populations change in number. Dark daisies decrease iii numbers as white daisies increase until their growth curves intersect: a point is reached where the light daisies are just as numerous as the dark ones. As the sun warms further, white daisies are more and more favored. Still, they regulate the planetary temperature, and the mean is held very close to the most favorable value for plants. Eventually the sun becomes too hot and the entire system suddenly dies. A sudden rise in planetary temperatures occurs as all life ceases. That is the end of Daisyworld.

Daisyworld is contrived and artificial; it bears little relationship to a real planet. I do believe, however, that in principle the operation of Gaia as illustrated on Daisyworld is a close parallel to what is occurring on our own planet. Indeed, some recent evidence on the possible significance of tropical forest ecosystems and the regulation of planetary and environmental temperature is like that of Daisyworld. In the early days, I wondered why tropical forests were so dark. Their very darkness would make them absorb sunlight, like the dark daisies. Since these are the hottest regions of the Earth, it seemed to me a counterproductive trait to have evolved. Quite recently, what we should have known from the beginning has become obvious. If seen from space, these tropical forest regions of the Earth are not dark at all, but blindingly white. They are covered by white clouds that are the product of evapotranspiration from the tropical forest trees beneath. These clouds stabilize forest temperatures. Moreover, at night, when the sun is not shining, the clouds

Figure 5 Patrick Holligan's photograph of a coccolithophorid bloom.

tend to disperse, and the very dark color is useful in dissipating the heat that is gathered during the daytime. The tropical forest-dark trees and white clouds-acts toward the regulation of the planet as both dark and white daisies simultaneously. My suggestion is speculative, and I mention it to illustrate how Daisyworld might be extrapolated to Earth. In addition to this tropical forest cloud effect, satellite photographs of oceans show blooms of coccolithophores, appearing to act in the manner of white daisies, that may affect the sea temperature (figure 5). I expect this story to develop more subplots when I come back and repeat it in 10 years.

We should consider more than the effect of albedo in the Daisyworld model. This same sort of environmental and growth feedback model helps explain regulation of concentration of atmospheric carbon dioxide. The growth of the biota continuously pumps carbon dioxide out of the atmosphere into the soil. The concentration of soil C02 is 30 times greater than that of the atmosphere. Similar C02-pumping processes occur in the sea. The biota continuously pumps C02 and maintains current atmospheric levels, which is probably a major climate-regulating mechanism.

In your books Ages of Gaia (1988), and even earlier in Gaia: A New Look at Life on Earth (1979), you mention the possibility of a direct connection between the Gaia hypothesis and the phenomenon of plate tectonics. Please explain.

When I prepared this material, I admit the idea was even beyond the category of speculation. Nevertheless, I stuck my neck out . nd included it. I am glad to report that no less a figure than Don Anderson, Professor of Geology at the California Institute of Technology, in an article in Science (Anderson 1984), stated quite specifically that we should consider the possibility that biological influence on the production of eclogite and liniestone made plate tectonics possible. We had one closely related notion 10 years ago: that limestone can be a fluxing agent. For movement of the massive tectonic plates, the region of molten material (the magma beneath them) must be fluid and capable of motion. The nature of limestone is such that it lowers the melting point of the rock mixture into which it is drawn. This extra fluidity, present as a result of the subduction of limestone, lubricates the movements of the plates, making the process possible. The limestone-nearly all of which is biological in origin-acts as a fluxing agent, keeping the rocks below the crust molten so that heat is more readily transferred. Convection currents persist. I can't give geological details of the eclogite argument, because it is not my expertise, but I strongly advise reading Anderson's article.

Please explain the relationship between atmospheric CO2 and limestone.

The amount Of C02 in the atmosphere depends entirely upon its lithospheric sources and sinks. The ultimate source Of C02 iS OUtgassing from volcanoes, and the sink for C02 is calcium silicate in rocks. In a process called weathering, calcium silicate (a very common mineral in igneous and metamorphic rocks) reacts with atmospheric C02 in the presence of water and forms calcium bicarbonate and silicic acid, both of which are soluble. These are transported by rivers to the ocean, where the calcium bicarbonate dissociates. The bicarbonate is taken up mostly by organisms in the formation of calcium carbonate shells, skeletons, scales, and other biogenic structures. After death, carbonate sediments form from the rain of dead coccolithophores, foraminifera, marine animals, etc. Eventually, under pressure, limestone-primarily CaCO3-forms from the former skeletal material. Later it is subducted to become the fluxing agent I just mentioned. For a better idea of the role of the biota in the formation of minerals I strongly recommend the book by Lowenstam and Wiener (1989). They give us great insight into the mechanisms of Gaia. What new ideas have you been entertaining?

Perhaps Earth's water has been retained by Gaia. Ocean salinity, water retention, and lateral movement of crustal plates are the ideas concerning us these days. Maybe you can help us work on them!


How the seabed saves the world New Scientist 7 Feb 96

Lou Bergeron

OXYGEN is a good thing, in moderation. Too little, and oxygen-breathing organisms would all suffocate. But if the atmosphere contained too much, firestorms would sweep the world. Now two geochemists believe they have unravelled the feedback mechanism that prevents a global catastrophe of either kind. The atmosphere's oxygen content is regulated by bacteria living in sediments on the seabed, they say. The amount of oxygen in the atmosphere is always roughly in balance with the oxygen content of the oceans-if one of these reservoirs of the gas becomes depleted, oxygen diffuses into it from the other. Much of the oxygen in the atmosphere is produced by photosynthetic algae floating near the surface of the sea. Over millions of years, excess oxygen is consumed by the process of weathering, in which the gas combines with minerals in rocks. Weathering also releases phosphorus, which eventually finds its way into the sea, where it is a vital nutrient for the same photosynthetic marine algae. "There is only one source of phosphorus for the ocean, and that is weathering," says Philippe Van Cappelien of the Georgia Institute of Technology in Atlanta. This connection with the major oceanic source of oxygen suggested to Van Cappellen and his colleague Ellery Ingall of the University of Texas at Austin that phosphorus could be the key to the oxygen feedback mechanism. They have now developed a computer simulation of the Earth's bkogeochemical cycles that supports their case. The model assumes that phosphorus which has accumulated in marine algae will end up in the faeces of animals further up the marine food chain. Some of this faeces reaches the seabed, where the phosphorus is incorporated into sediments by bacteria that consume oxygen. Van Cappellen and Ingall had already shown that this phosphorus "burial" increases when oxygen is abundant in the ocean. Low levels of oxygen, on the other hand, allow phosphorus to escape back into the sea. The new computer model has revealed that this balance between the burial and leakage of phosphorus may be the central cog of the Ear-th's oxygen feedback mechanism. The researchers tinkered with the inputs to their model to simulate what hap pens when the oxygen con tent of the atmosphere starts to rise or fall. During episodes of mountain for mation, for instance, the amount of weathering inc reases, pulling oxygen out of the atmosphere-and in turn from the ocean. Left unchecked, this process would asphyxiate the planet. But the model showed that as oxygen-depleted surface water circulates to the ocean floor, more phosphorus is released from the sediments. When it reaches the surface the phosphorus causes a bloom of algae that pump out oxygen, replenishing the gas lost through weathering. If an algal bloom were to occur when little weathering was taking place, however, it could boost the Earth's supply of oxygen to dangerous levels. But again, the model showed that the seabed and its population of bacteria would come to the rescue. As oxygen-rich water reaches the ocean's depths, the bottom-dwelling bacteria work overtime, taking phosphorus out of the water and depositing it in the sediments, starving the surface algae of hie nutrient (Science, vol 271, p,493). other researchers in the field are impressed. "What we call the sedimentary microbial community is really what rules the Earth," says John Hayes, a geochemist at Indiana University in Bloomington.