Pond Weeds and Daisies
Aquatic weeds are a nuisance. They’re irritating enough when they wrap themselves around my rudder when I’m dinghy racing, but more seriously, they also cause substantial ecological damage when they take over too much of a river system or lake. Even this, though, is a very minor problem compared to events 49 million years ago when a single species of pond weed may have brought an end to the warm climate our planet had been enjoying and kick-started global temperatures on a downward spiral towards the cold and highly glaciated planet we inhabit today. Prior to this catastrophe the early Eocene Epoch, which began about 56 million years ago, was a time of high carbon dioxide levels and very warm temperatures. Indeed, this period of time is possibly the best ancient parallel for the world humans will create if we persist with our greenhouse gas emissions. However, after about 7 million years of warm weather, carbon dioxide levels and temperatures started to fall. Many explanations have been advanced for why this cooling began but one of the most interesting involves weed infestation on a truly monumental scale.
All the action took place in the Arctic. During the Eocene the Arctic was a very different place from the frigid wasteland of today because the high carbon dioxide levels allowed the polar region to be about as warm as modern France and at least as wet as the lush and famously emerald-tinted island of Ireland. The heavily forested lands that then fringed the ancient Arctic Ocean were so warm, even in the permanent dark of the polar winter, that there were lizards, tortoises and alligators living inside the Arctic Circle. The geography too was rather different from today, since the Arctic Ocean’s link to the North Atlantic was then much narrower and, on the other side of the ocean, Alaska and Siberia were joined together to close off the Bering Strait. The Arctic waters were almost completely isolated from the rest of the world’s oceans and this allowed fresh river water and rain water to accumulate on the ocean surface as a light, brackish layer floating on denser, saltier water a few metres below. Even today the upper 50 metres or so of the Arctic is less salty than water in the rest of the world’s oceans, but in the Eocene, the Arctic was more isolated than it is now, rainfall was heavier, and stirring by waves was less intense so that this effect was greatly magnified. As a result, the Eocene surface layer on occasions became completely fresh. Those were the conditions that allowed rapid growth across the Arctic of Azolla, a freshwater fern normally incapable of living on a salty sea but able to thrive on this very odd, almost salt-free ocean.
Azolla is a fascinating plant. Unlike the land-loving ferns most of us are familiar with, Azolla grows as free-floating mats on the surfaces of lakes and rivers. It can do this because of its symbiotic relationship with bacteria that are able to break apart the strong molecular bonds in atmospheric nitrogen to make more plant-friendly compounds. Azolla carries these useful companions around in specially formed nodules that therefore provide it with the nitrate fertiliser that a free-floating plant can’t get by the usual approach of sticking roots into soil. In return for this invaluable service, the plant provides its bacterial guests with food in the form of glucose and this partnership allows Azolla to thrive in nutrient-poor waters where other aquatic plants struggle. In the case of the Eocene Arctic Ocean with its very unusual freshwater surface layer, this ability occasionally allowed ocean-sized Azolla blooms to form in the polar summer. The northern polar seas were then covered by a mat of vegetation from the coast of North America to the shores of Russia on the far side of the ocean more than 2,000 kilometres away. The entire Arctic Ocean became plant-green rather than the ice-white you might have expected. But in the long, dark polar winter, these weeds died or possibly just became dormant and much of the vigorous summer growth sank into the deep ocean. This could well have been the cause of the mid-Eocene cooling, because, as discussed in an earlier chapter, large-scale burial of plant matter in ocean sediments removes carbon dioxide from the atmosphere. Levels of the greenhouse gas would have fallen as Azolla remains accumulated at the sea floor and the resulting temperature drop may have been large enough to begin the descent towards the icy world of the last 2.5 million years: a world of regular ice ages interspersed by relatively short but warmer interglacial phases such as the one we have been living through since about 9000 BC. Many other factors played important roles in these changes to global climate but Azolla blooms may have been the straw that broke the camel’s back – a final trigger that set off a complex chain of events flipping the world from the warm climate of the early Eocene into the colder climate of more recent times.
This ‘Azolla event’ was, geologically speaking, a brief episode lasting little more than a million years. Even during that period Azolla blooms occurred rarely and seem to have been associated with occasional drops in global sea level that cut the Arctic Ocean off from the rest of the world even more than it already was. The world’s oil industry is currently very interested in exploiting the hydrocarbons that can be found in the Arctic Ocean and an Azolla layer, which is one of the principal sources for this oil, has been found everywhere that drills have penetrated marine sediments of the right age. This discovery of the remains of a freshwater plant across an entire salt-water ocean was more than a little surprising, but the evidence for ocean-scale blooms of this particular aquatic weed is strong. The sediments show too much Azolla detritus spread across too wide an area for it to simply be the result of run-off from the land. The sediments also show no other terrestrial remains, such as wood fragments, that would be expected in land-derived deposits.
The Azolla event was a rather unusual chapter in Earth’s history, but it is far from being the only time when there may have been life-generated climate change. Perhaps the best known additional examples are the Proterozoic cooling associated with the oxygenation of our atmosphere, discussed earlier, and the cooling that occurred 400 million years ago when roots of the newly evolved land plants accelerated weathering of rocks and increased the rate at which carbon dioxide was removed from our atmosphere. Events like these leave little doubt that living organisms have had significant effects on our climate but are these effects beneficial, catastrophic or some mixture of the two? Whatever the answer to this question, the impacts of life on climate are important to this book, because they almost certainly play a key role in the central problem I am tackling: What has kept our climate suitable for life throughout the 4 billion years of life’s existence?
In previous chapters I’ve tried to show that, even for worlds that begin well, there are no guarantees that climate will remain life-friendly for billions of years. As Mars and Venus both show, planets with initially fairly benign environments can later become too hostile to sustain a complex biosphere, even allowing for life’s extraordinary adaptability. In contrast, the Earth has somehow remained habitable across the eons and, if anything, has become more life-friendly as it has aged. So, why has terrestrial climate history been so different to that of our planetary neighbours? There are plenty of small differences, as discussed in earlier chapters, that may account for this happy circumstance. The absence of magnetic fields, the slightly different distances from the Sun, and the smaller sizes of our sister worlds all seem to have a role to play. However, there is one additional and particularly striking difference between the Earth and its neighbours – life itself! Perhaps life helps to stabilise climate. If this is true, and if life never became firmly established on Venus or Mars, this would explain the better climate outcomes on our own planet. The idea that life helps to create and sustain conditions suitable for life is, as I’ve mentioned before, called the Gaia hypothesis and much of the rest of this chapter will be concerned with this fascinating idea.
When it comes to discussing Gaia, everyone starts with a world full of daisies because there is no better illustration of how Gaia might work than ‘Daisyworld’. This imaginary planet was invented by Gaia’s originator, James Lovelock, and his collaborator Andrew Watson who we met in the last chapter. Daisyworld is a wonderfully simple hypothetical planet inhabited solely by daisies of just two varieties: black and white. Black daisies are better than white daisies at absorbing sunlight and so thrive when Daisyworld is cool, while the white variety stays cool in sunlight and hence thrives if Daisyworld becomes hot. As a direct result of this arrangement, Daisyworld becomes darker in times of cold climate, as more dark flowers grow across its surface, and lighter when it’s warm as the overheating black daisies die off to be replaced by their paler cousins. The resulting global colour changes stabilise Daisyworld’s climate, since the darker planet of cooler times absorbs more heat from the Sun and a whiter world reflects more heat into space when the climate becomes warmer.
While Daisyworld is an inspiring idea, no one pretends that it’s anything but a greatly simplified toy-model designed to illustrate a concept rather than realistically represent the behaviour of the highly complex real Earth with its millions of interacting species. However, it’s quite easy to come up with equally simple alternative hypothetical planets that behave in the opposite manner. Take ‘Mouldy Pineworld’ for example. Like Daisyworld, this planet has just two species: pine trees that turn carbon dioxide into wood, and mould that grows on the trees. Mould breathes oxygen in and carbon dioxide out and grows better in the warm, so on Mouldy Pineworld an increase in temperature encourages mould growth, which thrives at the expense of its hosts. The growth then increases the amount of carbon dioxide in the atmosphere and this, in turn, increases temperatures yet further. A cooling of Mouldy Pineworld, on the other hand, kills off the mould and encourages tree growth, which decreases carbon dioxide levels and enhances cooling. In contrast to Daisyworld, Mouldy Pineworld suffers from positive feedback, and so its biosphere magnifies rather than moderates changes in climate.
Daisyworld and Mouldy Pineworld are both imaginary, but one genuine biological climate feedback found on Earth, the response of northern forests to warming or cooling, is very nearly a combination of Daisyworld and Mouldy Pineworld – and it shows that both are wrong! The average temperature of our world has changed repeatedly over the last 2.5 million years, as the ice ages have cycled through the long glacial and short interglacial periods, and the coniferous forests of northern Europe, Asia and America have responded to this by spreading north during warm times and retreating southwards when the cold returned. This is the exact opposite of Mouldy Pineworld, where the trees spread when it was cold rather than when it was warm. On the other hand, as the forests moved northwards they grew over what was formerly tundra and those invading trees were substantially darker than the frozen wastelands they replaced. Thus, as the world warmed, the spreading northern forests made the Earth darker and more heat-absorbent. This is the exact opposite of Daisyworld, which became lighter as the world warmed. The climate response of the northern forests therefore beautifully shows how misleading simplified models can be if they are taken too seriously.
My invention of Mouldy Pineworld was, of course, highly contrived to ensure that it showed living organisms interacting in a way that necessarily produced the climate instability I was trying to demonstrate. However, Daisyworld is just as vulnerable to a criticism of being highly contrived, because the petal colours it assumes are far from inevitable. Black is indeed the warmest colour in bright sunshine, but it is the coolest colour at night or on a cloudy day. Dark colours are best both for gaining heat in sunshine and for losing heat in the shade, which is why, unlike mad dogs and Englishmen, many inhabitants of hot countries wear dark clothes and then stay out of the mid-day sun. I saw another example of this basic physical property of dark colours a few days ago as I was travelling by train through the snowy countryside of a wintry, but slowly thawing, Essex. As I looked out of the window I was struck that some fields remained blanketed in white while immediately adjacent fields had completely lost all snow. It took me a few minutes to work out what had produced this remarkable dichotomy but the key was that some fields had been ploughed and others had been left as grassy pasture. The relatively dark colour of tilled ground where it stuck out above the snow-filled furrows had allowed it to cool more overnight and the snow had not yet melted from these slightly colder fields. As these examples show, black is often the worst colour for staying warm and so it is far from obvious that dark daisies would have an overall survival advantage on a cold Daisyworld. If white daisies thrived instead in chilly weather, disaster would ensue. Daisyworld would become whiter as the climate cooled and this would enhance rather than retard cooling. The resulting anti-Daisyworld would look and behave very like the snowball Earth we met earlier in this book; a world that lurched uncontrollably between a brilliantly white but near-lethally cold state and a darker but near-lethally warm one.
Daisyworld can also be modified to illustrate another difficulty that the supporters of Gaia are working hard to resolve: the problem of cheating discussed in Chapter 9, which arises because all of life’s adaptations have costs as well as benefits. The human brain is an excellent example of an adaptation with both good and bad effects, because its development was key to our success as a species but it requires significant extra food to grow and maintain such an extravagant organ. A little less dramatically, there is also a cost to producing black or white pigment if you are a daisy. The biological cost of pigment production comes because it takes nutrients and sunlight to produce them, resources that could have been used for other purposes such as growing. This cost causes a problem for Daisyworld because, under such conditions, it would be beneficial for daisies to cheat at the natural selection game in exactly the sense I discussed in Chapter 9. Specifically, a mutant, unpigmented daisy can keep itself warm in cold times by growing next to sunshine-absorbing black daisies or cool in warmer times by growing next to sunshine-reflecting white daisies because the pigmented flowers affect not only their own temperature but also that of the surrounding air. Unpigmented daisies therefore benefit from the climate control of their neighbours without paying the cost of producing pigmentation. These cheats will out-compete their pigmented cousins even when the black and white daisies become rare, since they will still benefit from any marginal remaining climate control. Eventually, pigmented daisies might even be driven to extinction and Daisyworld would lose its Gaian stability.
In fairness I should admit that there is an ongoing and mathematically sophisticated scientific debate about whether the climate-stabilising behaviour of Daisyworld really is destroyed by such cheating but, nevertheless, the discussion does illustrate a genuine problem for Gaia’s supporters: a problem widely known as ‘the tragedy of the commons’. This parable was originally devised to illustrate a fundamental conflict between laissez-faire economics and good husbandry of our planet but it equally well demonstrates why anti-Gaian tendencies might be expected to evolve in ecosystems. Common lands, in England, are shared fields where all the inhabitants of a village have grazing rights. In the tragedy of the commons it is imagined that a particular field can support ten cattle and that ten villagers are entitled to use it. It would therefore make sense for the farmers to graze only one cow each, but a selfish villager could gain a distinct economic advantage over his neighbours by grazing two cows. There would then be eleven marginally malnourished cattle on the common, but because two slightly hungry cows are worth more than a single well-fed one, the selfish farmer would do better than his more community-minded colleagues. The other farmers might attempt to compensate for the reduced value of their own cattle by also grazing two cows each and, at that point, the common would become so over-grazed that all the cows would starve. Every villager (let alone their cattle) would end up worse off than if they had agreed among themselves to stick to a single cow each. Entirely rational, if selfish, behaviour by an individual produces catastrophe for all.
Humans are, at least in principle, capable of realising what is happening and of making agreements to prevent such insane over-use of limited resources. A good example is fisheries policy in which countries agree to limit the amount of fish they each take from the ocean rather than having a free-for-all that drives fish stocks to extinction. However, the same kinds of problems also occur within evolving ecosystems and, in this case, the uncompromising rules of natural selection do not permit agreements that limit the behaviour of individuals in the interest of everyone. Natural selection takes place at the level of single organisms and there is no mechanism to allow selection of attributes ‘for the good of the species’ let alone for the good of an entire ecosystem. This deeply anti-Gaian principle at the heart of life’s dynamics is well illustrated by a very stark real-world example: the near-absence of life across most of the Earth’s surface.
We are used to thinking of the Earth’s oceans as teeming with marine organisms but the open ocean is actually quite desolate and typically has about twenty times less biological productivity than those parts, such as coral reefs, that tend to get shown on television. The problem is that the seas are over-grazed in exactly the manner discussed for the common land above. In the ocean’s case the grazers are microscopic organisms rather than cows and they have over-consumed nutrients rather than grass. Iron, for example, is in very short supply in many ocean waters since it can only get a long way from the coast by being blown off the mineral-rich land in dust particles. Unfortunately, a little bit of dust doesn’t provide much metal and there is increasing evidence that photosynthetic marine micro-organisms compete for what little iron there is – and they may even hoard it. Any easily accessible iron dissolved in the sea is therefore rapidly removed to leave an environment that cannot support the biodiversity it might otherwise have sustained. The contrast between the barren oceans covering much of the Earth’s surface and those parts of the ocean fortunate enough to receive nutrient-rich water upwelling from the dark, and photosynthesiser-free, waters of the deep sea is striking. All the world’s major fisheries are either in areas of upwelling or close to land where nutrients can be supplied by air-blown dust.
The effect of iron deficiency on ocean life also plays a role in another interesting example of the multi-faceted interactions between life and the physical environment: the role of marine aerosols in our climate. Aerosols are solid particles or liquid droplets in air that are small enough to remain suspended for long periods. They undoubtedly play a direct role in climate by making the air less transparent, thus reflecting solar heat back into space and Earth heat back to the ground. Furthermore, aerosols play an indirect role by ‘seeding’ clouds. They provide particles around which water vapour can condense more easily than it does in their absence. Aerosols are produced by many processes such as volcanic eruptions or human activities, or simply by spray from the sea. The suggestion that sea-spray might, indirectly, encourage rain was first made by Glen Shaw, a geophysicist at the University of Alaska who had noticed that even the pristine atmosphere of the Antarctic has a very slight sulphurous haze that he later tracked down as originating from chemicals sprayed by breaking waves into the atmosphere. Shaw suggested that this could produce a significant negative feedback in the climate and his ideas were expanded on in a classic scientific paper outlining what is now known as the CLAW hypothesis (named after the paper’s authors Robert Charlson, James Lovelock, Meinrat Andreae and Stephen Warren). In this proposal, the marine aerosols contain a chemical called dimethyl sulphide (DMS), which is generated by microbes in the ocean. The DMS aerosol helps clouds to form and might be produced in larger quantities on a mild world with less ice cover and warmer, more microbe-friendly water. Thus, clouds should form more easily when the Earth is warm and these could act to reflect heat into space, counteracting the warming.
This was a nice idea, but further work has thrown up three major difficulties for the hypothesis. Firstly, the microbes are held back not by cold weather but by the iron shortage I discussed earlier. DMS is produced in greater quantities when it is dusty, and detailed examination of ice cores from Antarctica and the Arctic show the Earth to be dustier during ice ages than it is during warmer interglacials. More microbially generated DMS is therefore produced when the world is cold than when it is warm. The second problem is that clouds have a complex effect on the climate and frequently warm it rather than cool it. Clouds can hold warmth in as well as reflect sunlight away, and that’s why, for example, cloudy winter nights are warmer than cloudless ones. It is also now thought that much of the DMS is associated with sea salt rather than microbes, so it’s not even clear that biological productivity is the main control on its atmospheric concentration. So, once again, we see that the relationship between life and climate is complex and difficult to untangle. The DMS story may illustrate a negative feedback in the climate but it could equally well be a positive feedback. It may even be a negative feedback for exactly the opposite reason to that given by CLAW. DMS-generated clouds may be more common in cold times than warm times but may have a warming rather than cooling effect!
To me, the picture that emerges from all this is that the Gaia hypothesis lacks unambiguous observational support and has significant theoretical difficulties. That seems pretty damning but I’d nevertheless regard my views on Gaia as being broadly positive. My attitude can, I hope, be gleaned from events surrounding a poster entitled ‘Gaia or Goldilocks?’ which I presented at a conference concerning ‘Life and the Planet’ in London a few years ago. The text for my poster began:
Pseudo-Gaia
Imagine that everything you think you know about Gaia results from coincidence. Every cancellation of a physical effect (e.g. an inexorably warming Sun) by biological evolution (e.g. the rise in atmospheric O2) happened by luck at just the right time rather than because Gaia required it. Phenomenally good fortune would be needed but there are ~1022 planets in the observable Universe (and vastly more beyond our cosmic horizon) so that even highly improbable things will happen sometimes. Imagine further that a stable environment is a necessary precondition for a complex biosphere and that a complex biosphere, in turn, is necessary for the emergence of observers. Observers would then only emerge on rare worlds where all the coincidences required for a stable environment have occurred. How would the World look to such observers? It would look like Gaia and would have always behaved like Gaia, but it would not be Gaia. Is this our World?
As I stood by my poster discussing its contents with other conference attendees, I was approached by a sweet little old lady who politely enquired whether my poster might be ‘a little discourteous’. ‘Oh, I do hope not’, I replied, ‘it’s certainly not meant to be.’ Despite this reassurance I could see that, on second thoughts, producing a poster with the sub-title of ‘Pseudo-Gaia’ at a conference organised largely by Gaia enthusiasts might indeed be seen as unnecessarily aggressive so imagine my horror, half an hour later, when that sweet little old lady climbed onto the lecture theatre stage to give a conference keynote talk. My interlocutor had been Lynn Margulis, at the time one of the world’s greatest living biologists (sadly, she died later that same year) and a long-term collaborator with James Lovelock on Gaia theory. My point here (and on my poster) is that I regard the Gaia hypothesis as a fascinating and productive idea despite my opinion that it is probably wrong. Indeed, at the end of this chapter and in the next one, I will discuss a mechanism that I believe really could produce Gaian biospheres on inhabited worlds. However, before that, I want to look at a rather different, non-Gaian, way in which life could play an important role in keeping the Earth life-friendly.
I think that what is needed is a slightly different perspective on the problem of how biology, geology and astronomy interact to produce a broadly stable climate. Perhaps the emphasis on feedback, whether positive or negative, is misplaced. Perhaps an evolving planet naturally imposes a general trajectory on climate: a tendency to cool a planet and hence counter the warming influence of solar evolution. Steady continental growth, for example, has slowly increased the area available for weathering and also represents a permanent store for the carbon.
Furthermore, David Schwartzman, a professor from Howard University in Washington, DC, believes that life’s major innovations have tended to reduce greenhouse gas warming. An important example for Schwartzman is the emergence of primitive lichens that colonised the continents and helped to break down the rocks that they grew upon, perhaps as early as 700 million years ago. This increased weathering would, as we have seen, have led to more rapid removal of carbon dioxide from the atmosphere as acid rain dissolved the fragmented rocks and flushed the resulting bicarbonate ions into the oceans. Colonisation by lichens also had the effect of opening up the continents for other life forms and this general trend for the biosphere to expand through time is another central feature of Schwartzman’s proposals. Land plants, when they emerged hundreds of millions of years later, increased the rate of weathering and the degree of colonisation still further and so ratcheted up cooling by another notch. These examples illustrate a tendency for life to evolve a succession of increasingly efficient mechanisms for conquering the land, at each step removing more carbon from our atmosphere to give the otherwise quite unexpected downward trend in temperatures seen through geological history.
Micro-organisms, lichens, fungi and plants certainly strongly enhance weathering by helping to break up the solid rocks of the Earth and so it is very possible that the weathering feedback mechanism would have been too weak on a lifeless Earth to help stabilise the climate. It is also quite probable that more complex organisms could not have evolved unless increasingly efficient weathering, due to increasingly effective land organisms, had allowed global temperatures to fall. Continental growth and successively more efficient species of rock-digesters (e.g. trees) therefore acted to bring temperature down. We could be living on a world where the warming effect of an evolving star happened to be roughly cancelled by the cooling effect of geology and biology with no overall control from feedback mechanisms at all.
The best evidence available to support this idea of broad cancellation of solar heating by biological cooling comes from the well-constrained history, discussed earlier, of equatorial sea surface temperatures over the last half-billion years. As I mentioned, these sea surface temperatures show a very slight cooling of a couple of degrees centigrade superimposed upon more random, background fluctuations of about 10°C. Feedback, whether resulting from geochemical or biological processes, cannot explain the overall cooling, because feedback dampens change but cannot reverse it and so at best could only have reduced the expected warming of about 10°C that should have been produced by the steadily evolving Sun. In contrast, the combination of two opposing processes that roughly cancel each other would leave a small residual temperature fluctuation that could be either minor warming or minor cooling. On our world it just happens to be a slight cooling.
Of course, this rough cancellation requires a coincidence. The amount of cooling has to roughly equal the amount of warming and there’s no obvious reason why this should happen. Why should the timescale and amplitude of biological cooling be anything like the timescale and amplitude for the very different and completely independent processes causing solar warming? Schwartzman’s answer would be that this is Gaia in action but, it seems to me, there is a much simpler and more straightforward explanation. If bio-geological evolution occurs at different rates on different worlds, intelligent observers will emerge only on those planets that, by chance, have just the right rate of geological and biological cooling. This is the anthropic principle again, the observational bias that inevitably results from the fact that the history of our planet must be compatible with our existence as observers. On this interpretation, such cancellation does not occur on the vast majority of Earth-like planets. The usual situation is either that biology freezes itself to death or that solar physics fries the world well before complex organisms ever evolve. We just happen to be living on one of the rare, lucky worlds where neither of those things happened.
I should add that combining Gaia and Goldilocks in this way is not really a new idea. Rather, it is a twist on a proposal called ‘Lucky Gaia’, which suggests that only those planets that, by chance, produce Gaian mechanisms will give rise to observers. However, to me, this idea rips the heart out of Gaia. Is it really Gaia at all when the anthropic principle has done all the heavy lifting?
This is all, perhaps, a little too negative and I can hear the spirit of Professor Margulis politely requesting that I tone things down a little. So, I’d like to finish this chapter on a more positive note by discussing one possible way forward for Gaia. For all the reasons I have been through above, I have difficulty believing that Gaia can emerge simply as the result of a preponderance of negative feedback over positive feedback among the innumerable interactions that occur in our biosphere. Rather, I agree with those who have suggested that Gaian stability can be understood only as an emergent property of a very complex system. Emergent properties can be illustrated using a heap of sand as an analogy. If you slowly add sand to a pile, one grain at a time, its height gradually grows along with the slope of its sides. However, at a critical point, the sand pile stops getting any steeper and any additional grains simply slide down the sides to make the pile wider rather than taller. The maximum slope on the sides of this pile is a property of the sand but not a property of a single sand grain. It simply doesn’t make sense to talk about the maximum stable slope for a single grain – and, indeed, this is a property that doesn’t emerge until there are at least a few tens of grains in the heap. The maximum stable slope is therefore an emergent property of sand grains. Similarly, environmental stability may be a property of ecosystems that emerges only once there are large numbers of interacting biological and geochemical processes. On this interpretation, the life-generated stability of our environment simply cannot be understood as the crude sum of individual species–climate feedbacks.
However, why should the laws of nature do this? It seems just too good to be true. Personally, I can think of only one plausible answer and it comes back, once again, to the anthropic principle but now played out on an immensely larger stage. Perhaps we live in a lucky Universe, rather than on a lucky planet. Perhaps the entire cosmos is fine-tuned for life.