Paul J. Sise

Evolutionary Biology

February 17, 1996

In 1772, Joseph Priestly discovered that there is a mutual life support system between animals and plants. This is because, as he realized, animals inhale oxygen and exhale carbon dioxide, while plants do essentially the opposite. As we now know the atmosphere consists of several different kinds of gases. These gases remain in the atmosphere in a certain mixture of 21% oxygen and 78% nitrogen, and smaller proportions of other gases like carbon dioxide. This is especially important because that combination is highly reactive. Something must be regulating the Earth's atmosphere or else it would be more like the atmosphere of Mars or Venus (almost all carbon dioxide). Almost two hundred years after Priestly, an atmosphere chemist name James Lovelock wondered why the ratios of the different gases remained constant not only across the globe, but also throughout vast periods of geologic time. He concluded that, much like the relationship between animals and plants, there is a relationship between the planet's inhabitants and the nonliving aspects of the planet like the air, soil, and oceans. This system, and Lovelock's hypothesis, are called Gaia.

The Gaia hypothesis states that life can influence, regulate, or even optimize such aspects of its environment as the temperature and chemical composition through feedback loops between life and the non-living environment. According to Lovelock in an article for Nature, his own colleague Andrew Watson "observed that almost everyone now accepts that life...influences the environment...life and the environment are a coupled feedback system...The real debate, then, is how important and how tight the coupling" (Nature, 1990). Life both alters and adapts to the environment, but is the relationship tight enough to produce emergent properties like stability? People like Lovelock and microbiologist Lynn Margulis think that it is.

Lovelock and Margulis met through NASA, where they both worked in the 1960's. Lovelock was working for NASA as a consultant when NASA was preparing to launch the Viking spacecraft on a mission to see whether or not life existed or had existed on Mars. The atmosphere of Mars is dominated by carbon dioxide and has very little oxygen and methane. Lovelock believed that the gases which are produced during photosynthesis would be rare on a lifeless planet. According to Stephen Schneider in his article Debating Gaia, "Lovelock deduced that the probability of life existing on Mars was extremely small" due to the composition of the atmosphere of Mars (Schneider, 1990). Lovelock and Margulis eventually proposed that the Earth was a self regulating system that could maintain the ratios of gasses in the atmosphere due to the life that inhabited its surface. The name for this system, Gaia, was suggested to Lovelock by his neighbor William Golding, the Nobel Prize-winning novelist (Schneider, 1990).

Lovelock and Margulis proposed that the system by which the Earth regulates itself is more like a biological process, such as homeostasis, than it is like a mechanical process. From this comes the idea that the Earth is an organism. Much like how the specialized fields of study such as biochemistry and anatomy, as well as others, combine to form the branch of science known as physiology, the different Earth sciences can combine to form another similar systems approach field. This field is called geophysiology. James Kirchner, in his article "The Gaia Hypothesis: Can it be Tested?" states that he agrees that it may be useful to think of the Earth as an organism, "but the question of whether the Earth actually is an organism is neither scientifically meaningful nor scientifically answerable" (Kichner, 1989). Thus come the criticisms of Gaia. Is it testable?

There are three main criticisms of the Gaia hypothesis. Each are proposed by a variety of individuals who consider themselves good scientists, among them include many evolutionary biologists. They are: 1) The Gaia hypothesis cannot be scientifically tested. 2) The Gaia hypothesis is actually several distinct hypotheses. and 3) If the planet is an organism, what gave birth to it? What did it evolve from?

Kirchner, in his article which focuses on the first criticism outlines the criteria for whether or not an hypothesis is testable or not. He says that the hypothesis must be "clear and...unambiguous." He also says that it must "generate" confirmatory and falsifying predictions (Kirchner, 1989). Apparently the Gaia hypothesis has failed his criteria for being a true hypothesis. He continues by exclaiming that even false hypotheses are more useful than one that is untestable. Others, however, have argued that Gaia is helpful as a metaphor because it has seeded the interest in certain research and has fueled debate within the natural sciences.

The idea that the Gaia hypothesis is actually five hypotheses was proposed by Kirchner in 1988. They are according to an article by Stephen Schneider:

1) Influential- "the biota has a substantial influence over certain aspects of the abiotic world"

2) Coevolutionary-"the biota influences its abiotic environment and that the environment in turn influences the evolution of biota by Darwinian processes"

3) Homeostatic- "the biota influences the abiotic world and that it does so in a way that is stabilizing"

4) Teleological- "the atmosphere is kept in homeostasis, not just by the biosphere, but by and for the biosphere"

5) Optimizing- "the biota manipulates its physical environment for the purpose of creating biologically favorable, or even optimal, conditions for itself" (Schneider, 1990)

It may be said that these hypotheses become more Gaian as one reads down the list, and that the first seems obvious to be true, but that does not necessitate the other more extreme forms of the hypothesis to be true. We must remember, then, to clarify what form of Gaia we are discussing when we say "Gaia." In this paper I will tend to refer to the Homeostatic form of the hypothesis since it is not only the midpoint between the extreme views but also because it is the Gaia that I (and I believe many others) have had the most contact with.

In response to these criticisms Lovelock was compelled to find a way to defend his hypothesis and to find a way to test it. Lovelock designed a model to illustrate the way in which he thought that Gaia would work if it truly did exist. This model is called Daisyworld. Since I do not have the data or formulas used to produce the actual models I am reluctant to critique both the models themselves beyond a general depth and the methods used to produce the models. For a more complete description of Daisyworld see the accompanying article by Lovelock and the related graphs. The mathematics, which as I understand to be crowned by partial differential equations, can be found in an article published in 1983 by Lovelock and Watson, titled "Biological Homeostasis of the Global Environment: The Parable of Daisyworld."

(Analysis of two current works)

The Ages of Gaia

In his second book on the Gaia hypothesis, The Ages of Gaia, Lovelock attempts to refute the claim that Gaia is untestable by discussing his Daisyworld model in chapter three, titled, Exploring Daisyworld. Daisyworld is a planet that is essentially identical in property to the Earth, it is the same size, maintains the same orbit, and is the same age as the Earth. Its star, like our sun, is gradually becoming hotter as time progresses. The one difference is that it is populated solely by daisies. There are light colored daisies and there are dark colored daisies. Simply put, the dark color of the dark daisies tends to warm the air around them, while the light colored daisies reflects light and heat, thereby cooling the area around them. This produces a regulatory affect on the temperature of the planet as the daisies compete against each other and as the solar luminosity of the star increases. There is "no foresight, planning, or purpose...invoked" (Lovelock, 1988).

Lovelock soon found that the Daisyworld model worked just as well with more than two degrees of daisy darkness. Lovelock made Daisyworld models with up to twenty different variations of daisy, each darker than the previous. (See graphs labeled "A" and "B.") He wrote that the only limits were the "speed and capacity of the computer used and by my patience" (1988). Invariably, the effect of the inhabitants of Daisyworld was a stable form of temperature regulation.

Does this model actually reflect what we see in the natural world? The fact that life has persisted throughout its three and a half billion year history, and the fact that the atmosphere has remained close to its current mixture of gases for so long appears to correlate with what the Gaia hypothesis predicts. Also, take note that when there is the least stress on the system there is the greatest diversity in both nature and in Daisyworld, but when there is the greatest stress, when the solar luminosity is at its peak, for instance, there is the least diversity. (See Daisyworld graphs)

At this point Lovelock is convinced that he has answered the criticisms of many scientists, but he continues by writing "there is much more to this new model than an answer to the criticisms of skeptical biologists" (1988). Within Daisyworld is the key to understanding how a complex system can become and remain stable. This is due to the feedback relationship between the biota and the environment. Has there ever been devised a model that is similar to Daisyworld that we can use to prove this? "In the 1920's, the mathematical biologists Lotka and Volterra introduced their...model of the competition between rabbits and foxes. It differed (from Daisyworld) in that the environment was taken as infinite and neutral" (1988). The relationship between the animals was as follows: "Foxes increase in numbers as the numbers of rabbits grow, but rabbits decline as the foxes increase" (1988). In their model the animals had no effect on the environment and the environment, excluding the rabbits and foxes, did not effect the animals. Since there was no feedback the result was instability. Any change in the populations of the animals "dooms this simple world to cyclical fluctuation from which it can never return to a stable ratio" (1988). Lovelock then experimented to find the cause of Daisyworld's stability. "The answer is that in Daisyworld the species can never grow uncontrollably; if they do, the environment becomes unfavorable and growth is curtailed" (1988). To prove this he added foxes and rabbits to Daisyworld, as well as four plagues, to see if the model could regain its stability. (Graph "C") The system, even with the plagues, quickly regained stability.

Lovelock then continues the trend in his experiments with Daisyworld by exploring the ways in which periodic drops in population by "plague" effect the ability of the model to regain stability. In one experiment (Graph "D") Lovelock designs a model that has solar luminosity at a constant throughout the model. At an arbitrary point in time a plague kills sixty percent of the daisies. With the decrease in the population there is an increase in the temperature, but once the daisy population returns to normal the temperature increase is reversed. At no point in the model does the temperature reach the point in which it is too hot for life to survive. Lovelock finally combines a model that makes use of plagues with a model that shows stability in relation to the ability of ten kinds of daisies to regulate the temperature of Daisyworld in the presence of a star with increasing luminosity. (Graph "E") Of all the Daisyworld models, this model most closely represents what has occurred throughout the history of life on this planet. Though solar luminosity increases, the temperature has remained fairly constant, much like in this model. Sometime in the distant future the sun will be far too hot for life to continue existing on this planet and just before this happens there will be a dramatic period of selective pressure. This pressure should be the cause of a significant decrease in the diversity of life, ie. extinction. "This experiment graphically illustrates the way that stability, as measured by the capacity to regulate the climate, correlates with diversity" (1988).

The chapter concludes by stating that "It matters little whether Gaia theory is right or wrong; already it is providing a new and more productive view of the Earth and the other planets" (1988). Lovelock also lists what seem to be some major points that summarize his views on Gaia. These points are not necessarily tied directly to arguments in the chapter, but they are relevant as answers to the criticisms by other scientists. The list is as follows:

1. "Life is a planetary phenomenon. On this scale it is near immortal and has no need to reproduce."

2. "There can be no partial occupation of a planet by living organisms. This would be as impertinent as half an animal...Where there is incomplete occupation, the...forces of physical and chemical evolution would soon render it uninhabitable."

3. The concept of adaptation has been altered. Not only does life adapt to the environment, but it affects it as well. The evolution of life and the abiotic environment are coupled.

4. "We can, for the first time , build ecological models that are mathematically stable...increased diversity among the species leads to better regulation" (1988)

So why is it that after all of this mathematical evidence are scientists still unwilling to accept Gaia? According to Kirchner, "Daisyworld cannot be directly tested against real world data." The main problem here is that Daisyworld compares the theory to its own mathematical model, in other words it compares itself to itself. Because of this, there is no way for the experiment to fail, therefore it neither actually tests nor means anything (Kirchner, 1989).

Since Lovelock's models incorporate both stability and instability, or how stability can come forth from instability, then "one wonders what conceivable events could not be interpreted as supporting the Gaia hypothesis. If there are none, Gaia cannot be tested against the geologic record" (1989). These are cutting words indeed especially when we remember what Kirchner said about hypotheses that cannot be tested. (They are not really hypotheses and they are more harmful than helpful.)

Kirchner has continued his attacks on Gaia and on Daisyworld for some time. In 1989 he challenged Lovelock and Margulis to devise a falsifiable version of the Gaia hypothesis. Though Lovelock questioned the value of requiring his hypothesis to be strictly falsifiable he did meet Kirchner's challenge. Lovelock predicted that the oxygen levels in the atmosphere have remained within 5% of their current level during the history of land plants. Someday we will see whether Lovelock is right, and if he is, Gaia will grow in strength as a theory and should become more accepted in the scientific community.

In closer relation to evolutionary biology, Gaia hypothesis conflicts with the most fundamental rules of evolution, especially if we consider the planet to be an organism. Variation and natural selection, which are responsible for the existence of every living individual to date, cannot be responsible for the origin of the Earth as a living entity. This is because natural selection requires a population to select from, but there is only one planet Earth. So either the Earth is not alive, or there is a singular exception to the rule that all organisms are a result of natural selection. Whether or not the Earth is alive is a very tricky question. It does display some of the most important characteristics of a living organism, but it does not self-replicate. Margulis believes that self-replication is not as important, in this case, than the fact that the Earth has a metabolism, so to speak, and has the properties of self construction and repair. (Barlow and Volk, 1992) Arguing over the origin of Gaia, as the origin of a global living organism, can be too much like an argument concerning the origin of life itself. If one is not careful it can turn into a theological discussion and stray helplessly far from good science.

The proponents of Gaia refute even this criticism concerning natural selection by insisting that it does have a major role in the origin of the planet as an organism. Jonathan Schull argues that "species intelligence is an emergent property of lower level selection" (1990) so why cannot there be a global lifeform that has emerged in the same manner? According to Webber "Those populations are fittest that best enhance the autocratylitic behavior of the reward loops in which they participate" (1989). This means that if an organism can better contribute to the feedback loop which produces the stability of Gaia than another organism, then it has higher fitness. It will pass on its genes and because it is selected for we can say that Gaia is an indirect product of natural selection.

One year after The Ages of Gaia Lovelock published a paper titled "Geophysiology, the Science of Gaia" in Reviews of Geophysics. As the title suggests, this paper energetically defends Gaia hypothesis as real science, boldly stating that it is it has "matured" over the years and that it makes "risky predictions," which are after all, the best kind. (Lovelock, 1989) (To predict that it will snow next winter somewhere in the United States is a much less risky, and much less useful prediction than predicting that it will snow in Appleton WI on the 26 of November, 1996.)

As with most papers this one starts with a general introduction to the Gaia hypothesis. Lovelock specifically draws attention to the fact that it is often useful to combine the values of distinct disciplines in order to accomplish a common goal. Medicine, he claims, benefitted from the existence of the discipline of physiology, which combined anatomy, biochemistry, etc. He then writes "The purpose of this paper will be to put forward analogous Earth science, geophysiology, as the transdisciplinary environment for planetary scale problems...It may be useful...to consider the Earth as if it were a living organism" (Lovelock, 1989).

There it is again, the notion of a living Earth. Lovelock defends this idea by reminding the reader that this is not a new or revolutionary concept. Even James Hutton, one of the fathers of modern geology held this view. Though this view has not been common since the beginning of the last century due to the divergence of biology and geology, there is reason to believe that we should go back to it. Gaia hypothesis bridges the gap between the now separate biology and geology, so to understand it better we should find a way to think of the Earth in a way that both geologists and biologists can feel comfortable working with. To view the Earth as if it were alive could accomplish this goal.

Once Lovelock completes his introduction he begins a section called "Earth as a superorganism." Organisms and closed loop self-regulatory systems (I assume he mentions the closed loop system because Daisyworld is a closed loop self-regulating system) are "expected to show emergent properties" (1989). He continues by saying that these systems are dreadfully difficult to explain and that the use of Popperian falsification tests on them may not be very useful, as compared to an hypothesis in physics or chemistry. A major part of the big Gaia debate lies in the fact that many scientists believe that no good hypothesis is beyond the scope of Popperian falsification. Perhaps Lovelock realizes that he cannot falsify his hypothesis so he just denies the need for falsification. Then again, perhaps Gaia really is beyond the scope of simple Popperian falsification. If this is true then we must reevaluate what good science is and accept Gaia for its useful qualities. Whether fortunate or unfortunate, many scientists haven't budged on the need to falsify Gaia hypothesis. Lovelock concludes this section by saying that if the Earth is alive it "requires physiology, as well as chemistry and physics" just like any other living organism does (1989).

The next section is called "Models of Gaia." He begins by discussing the origin of the hypothesis and the prediction he made concerning the lack of life on Mars. Concerning the gases found in the Earth's atmosphere he writes "this disequilibrium reveals the presence of life" (1989).

To illustrate this point he includes a bar graph that contrasts the atmosphere of a planet with life and an atmosphere for a dead planet. The differences are very pronounced and it is easy to see that there is little chance of confusing the two. A planet with life, for instance, should have a substantial amount of both CH₄ and NH₃ whereas a planet without life has neither one.

Lovelock then explains that in order to understand feedback loops the numbers of variables must be reduced. This is where Daisyworld comes into play. In this paper he goes into more detail about the assumptions about the daisies on Daisyworld and explains how the temperature is calculated. The daisies "do not grow below 5.0C or above 40.0C, but grow best at 22.5C" (1989). The mean surface temperature is determined by the amount of heat that the planet receives from the star and the amount of heat lost by radiation. The albedo, or solar reflectivity of the planet affects the surface temperature. This leads us to the daisies. If the planet is covered by light daisies then the albedo is high, thus cooling the planet. If the planet is covered by dark daisies then the albedo will be low and planet will absorb more heat, thus allowing for a warmer surface temperature than without the dark daisies. This assumes that the light daisies are lighter in color then the surface color without daisies and that the dark daisies are darker than the ground without the daisies. Figure "G" shows Daisyworld with dark and light daisies and their effect on global temperature in the presence of a star with ever increasing solar luminosity.

The next Daisyworld model includes a third kind of daisy, one that is colored with the same degree of darkness as the surface color. As figure "H" shows, the dark daisies are selected for while the planet is in its warming up stage during its youth, the light daisies are selected for while the planet it striving to remain cool enough to maintain life toward the end point of the model, but the neutral colored daisies are selected for in the interim. This is because there is no need to regulate the temperature of the planet when it is already at the appropriate temperature.

Finally, figure "I" shows Daisyworld with ten variations on color within the species of daisy. This model more accurately reflects the natural world. As in the other publication on Daisyworld, Lovelock stresses that unlike similar models of competition, the Daisyworld model is stable, and can recover stability even if half of the population of daisies is destroyed in a plague.

Lovelock suggests that the Daisyworld has no true thermostat, but that the system moves to a stable point determined by the relationship between the organisms and the environment. "The system always moves to a stable state where the relationships between the daisy population and planetary temperature and that between temperature and daisy growth converge" (1989).

Lovelock moves away from Daisyworld with his description labeled figure six, or the graph labeled "F" in this paper. This graph reminds us that there is supposed to be more to Gaia than temperature regulation. There is also the effect life has on the chemical makeup of the environment. Here, we see the relationship between plants and the amount of oxygen in the atmosphere. As we know, plants release oxygen when they photosynthesize. The oxygen they release increases the rate in which rocks are weathered, or broken down into soil. Soil provides the nutrients that the plants need to survive. Too much oxygen is toxic, so there must be a way of limiting the amount of oxygen in the atmosphere. This mechanism is found in the fact that the plants produce oxygen, and benefit from its effect on rock weathering, but will die if there is too much oxygen. Once the optimum amount of oxygen in the atmosphere is surpassed then the detriments of oxygen (the toxicity) outweigh the benefits and the that which produces the oxygen begins to die, thereby reducing the amount of oxygen added to the environment. The excess oxygen is then removed from the atmosphere through oxidation and the respiration of animals.

Notice that once this model has reached a stable point then it will tend to remain stable. Also note that on the graph the point that the mathematics predicts to be stable coincides with the actual percentage of oxygen in the Earth's atmosphere. Unlike the Daisyworld model, this is comparing the hypothesis to nature, instead of comparing the hypothesis to itself. So according to this graph, Lovelock's prediction, that the relative amount of oxygen in the atmosphere has remained within five percent of what it is currently, during the history of land plants, should be correct.

As in the chapter from The Ages of Gaia, Lovelock concludes by stating that "Life is a planetary phenomenon...Gaia theory adds to Darwin's vision...(and that we now have) a way to look at the Earth mathematically" (1989).

Lovelock and Gaia hypothesis have stoked the fires of a debate that has lasted for decades and is sure to last for several more. Despite Gaia's flaws, or should I say, because of them, scientists across the world have been forced to reexamine their views of the planet Earth. Often the debate has inspired research that would have gone undone. The very definition of science and the use of Popperian falsification has been questioned. Without a reason to question things, or to reexamine them, we would continue in whatever trend is fashionable and we may miss very important clues. At the base of it, Lovelock's goal is to discover what he can about the world. He and his compatriots have used Gaia as a tool to reach this goal. Through Gaia they have learned about the Earth and have forced other people to do so as well. Whether or not Gaia is a proper hypothesis it has served science well as an hypothesis generator.

(Update) It is now the summer after my first year of graduate school, 1998. This year I had the benefit taking "Environmental Evolution" with Lynn Margulis. As you may guess we went to great lengths to understand Gaia better. We decided that there were ways of testing Gaia. Both involve life on other planets. First, if life were discovered on a planet that had gasses in that atmosphere that were not reactive with each other then the Gaia hypothesis would be disproved. The second deals with terraforming Mars. If we changed the atmosphere just enough to have life on Mars, and then find that life itself was regulating the atmosphere then the Gaia hypothesis would be proved valid.

Works Cited

Barlow, Connie and Tyler Volk. 1992 Gaia and Evolutionary Biology. Bioscience. vol. 42

No. 9 October

Kirchner, James W. 1989. The Gaia Hypothesis: Can it be Tested? Reviews of Geophysics. May, pg 223-235

Lovelock, James E. 1988. The Ages of Gaia. Chapter 3. "Exploring Daisyworld." W.W. Norton and Company, Inc. New York

Lovelock, James E. 1989. Geophysiology, the Science of Gaia. Reviews of Geophysics.

May, pg 215-222

Lovelock, James E. 1990. Hands up for the Gaia Hypothesis. Nature. vol 334. March 8

Schneider, Stephen H. 1990 Debating Gaia. Environment. vol 32, no. 4. May

Lovelock, James E. 1979. Gaia: A New Look at Life on Earth. Oxford University Press New York

Williams, George C. 1992. Gaia, Nature Worship and Biocentric Fallacies. The Quarterly Review of Biology. vol. 67, No. 4 The University of Chicago. December

Marcia Bjornerud- Lawrence University Department of Geology

"Of course life can affect the temperature of a planet, but only to a degree."

Christopher C. Schmidt, 1996