Many of us, scientists included, harbour contradictory intuitions about Mother Nature. We can see that ecosystems often have an inherent ability to self-stabilise, and we know we wouldn’t be here if the planet hadn’t maintained conditions suitable for life for almost 4 billion years. One reaction is to claim that some Earth-wide equilibrium, though fragile, does exist, and reflects the fact that species have evolved to cooperate with one another. Another is to say that the first response is nonsense: organisms are ‘selfish’, and evolution isn’t cooperative but rather a brutish Darwinian competition that selects individual organisms based on their ability to survive and reproduce. The primordial balancing act performed by our biosphere, if it exists at all, is more or less a lucky accident.
The idea that the Earth itself is like a single evolving ‘organism’ was developed in the mid-1970s by the independent English scientist and inventor James Lovelock and the American biologist Lynn Margulis. They dubbed it the ‘Gaia hypothesis’, asserting that the biosphere is an ‘active adaptive control system able to maintain the Earth in homeostasis’. Sometimes they went pretty far with this line of reasoning: Lovelock even ventured that algal mats have evolved so as to control global temperature, while Australia’s Great Barrier Reef might be a ‘partly finished project for an evaporation lagoon’, whose purpose was to control oceanic salinity.
The notion that the Earth itself is a living system captured the imagination of New Age enthusiasts, who deified Gaia as the Earth Goddess. But it has received rough treatment at the hands of evolutionary biologists like me, and is generally scorned by most scientific Darwinists. Most of them are still negative about Gaia: viewing many Earthly features as biological products might well have been extraordinarily fruitful, generating much good science, but Earth is nothing like an evolved organism. Algal mats and coral reefs are just not ‘adaptations’ that enhance Earth’s ‘fitness’ in the same way that eyes and wings contribute to the fitness of birds. Darwinian natural selection doesn’t work that way.
I’ve got a confession though: I’ve warmed to Gaia over the years. I was an early and vociferous objector to Lovelock and Margulis’s theory, but these days I’ve begun to suspect that they might have had a point. So I’ve spent the past five years trying to ‘Darwinise Gaia’ – to see widespread cooperation as a result of competition occurring at some higher (even planetary) level. I can see a few paths by which a Darwinian might accept the idea that the planet as a whole could boast evolved, biosphere-level adaptations, selected by nature for their stability-promoting functions.
This is not exactly a recanting of views, but it’s certainly a marked departure from how I thought 40 years ago. Darwinising Gaia seems important not just to me personally, but because it would offer a satisfyingly deep theoretical basis for efforts to maintain a habitable planet – and a way to reflect on contemporary environmental crises beyond applying a simple label such as ‘Gaia’s revenge’, with its anthropocentric and theistic implications.
For traditional Darwinian natural selection to work, the entities in question must display some property or ability that can be inherited, and that results in their having more offspring than the competition. For instance, the first creatures with vision, however fuzzy, were presumably better at avoiding predators and finding mates than the sightless members of their population, and had more surviving progeny for that reason. In technical terms, then, selected entities must exist in populations showing heritable variation in fitness, greater fitness resulting in these entities’ differential reproduction.
Even if inherited properties are the result of undirected or ‘random’ mutation, repeating the selection process over generations will incrementally improve on them. This produces complex adaptations such as the vertebrate eye, with its highly sophisticated function. Light-sensitive areas acquired lenses for focusing and means for distinguishing colours step by advantageous step, ultimately producing modern eyes that are clearly for seeing. So even without an overall purpose, evolution creates something that behaves as if it has a goal.
Back in 1979, when Lovelock’s first popular book, Gaia: A New Look at Life on Earth, came out, the wider field of evolutionary biology was becoming a very reductionist discipline. Richard Dawkins’s The Selfish Gene had been published three years earlier, and it promoted a hardcore gene-centrism insisting that we look at genes as the fundamental units of selection – that is, the thing upon which natural selection operates. His claim was that genes were the reproducing entities par excellence, because they are the only things that always replicate and produce enduring lineages. Replication here means making fairly exact one-to-one copies, as genes (and asexual organisms such as bacteria) do. Reproduction, though, is a more inclusive and forgiving term – it’s what we humans and other sexual species do, when we make offspring that resemble both parents, but each only imperfectly. Still, this sloppy process exhibits heritable variation in fitness, and so supports evolution by natural selection.
In recent decades, many theorists have come to understand that there can be reproducing or even replicating entities evolving by natural selection at several levels of the biological hierarchy – not just in the domains of replicating genes and bacteria, or even sexual creatures such as ourselves. They have come to embrace something called multilevel selection theory: the idea that life can be represented as a hierarchy of entities nested together in larger entities, like Russian dolls. As the philosopher of science Peter Godfrey-Smith puts it, ‘genes, cells, social groups and species can all, in principle, enter into change of this kind’.
But to qualify as a thing on which natural selection can operate – a unit of selection – ‘they must be connected by parent-offspring relations; they must have the capacity to reproduce,’ Godfrey-Smith continues. It’s the requirement for reproduction and leaving parent-offspring lineages (lines of descent) we need to focus on here, because they remain essential in traditional formulations. Without reproduction, fitness is undefined, and heritability seems to make no sense. And without lines of descent, at some level, how can we even conceive of natural selection?
I’d hope that Darwin, were he alive today, wouldn’t balk at my non-traditional steps
This is why Lovelock’s Gaia goes many steps too far. Multispecies communities such as Gaia, many symbiotic organisms and most ecosystems don’t in general reproduce themselves. Gaia’s parts (potentially billions of species) do each individually reproduce, but the biosphere as a collective doesn’t reproduce together as a collective. It thus has no simple heredity and produces no single set of parent-offspring lineages, but rather a multiplicity of often incongruent, independently reproducing lines of descent. By standard Darwinian thinking, entities such as Gaia are not units of selection, and can’t exhibit ‘adaptations’. Maybe algal mats do control global temperature and coral reefs do have a beneficial effect on ocean salinity, but this is just good luck. Species within them might ‘coevolve’, using other species as their biotic environment – but each is selfishly on its own. Yes, bees visit flowers, and flowers encourage this attention, because that’s how their pollen is spread, but individuals within each species do this for their individual benefits, having more progeny as a result. Evolution by natural selection at the biosphere level seemed impossible 40 years ago, and still seems problematic now.
I argued this in 1981, and Dawkins made similar arguments a year later, in The Extended Phenotype. There’s no relevant Darwinian population in which biospheres compete, he observed:
The Universe would have to be full of dead planets whose homeostatic regulation systems had failed, with, dotted around, a handful of successful, well-regulated planets, of which the Earth is one.
But even if this were the case, it still wouldn’t be enough, Dawkins notes. We’d also have to offer an account of biospheric reproduction, ‘whereby successful planets spawned copies of their life forms on new planets’.
This view still holds the field. As recently as 2015, while reviewing Lovelock’s A Rough Ride to the Future (2014) for the London Review of Books, Godfrey-Smith could write that:
[F]eedback between different living things is indeed ubiquitous, and some kinds of feedback help life to continue. But those benefits to life as a whole are byproducts – they’re accidental. The interactions between species are consequences of the traits and behaviours that evolutionary processes within those species give rise to, and those processes are driven by reproductive competition within each species … From the fact that life still exists, we can tell that traits too antagonistic to life itself, however beneficial to the organisms that bear them, must not have arisen. If they had, we wouldn’t be around to discuss the matter. But that isn’t what kept those traits at bay.
Godfrey-Smith is alluding to something like the anthropic principle: if Life had not established stabilising feedbacks, we wouldn’t be here – and, since we are, it’s necessary that it did, regardless of how unlikely that actually was. But I want something more than this – a mechanism by which selection at the level of the biosphere would be likely to produce stability. Such a mechanism – a Darwinian way of making beneficial ‘accidents’ into the equivalent of heritable variations that could evolve via natural selection – will be possible, I think. The work is far from complete, and much needs to be aligned or contrasted with emerging work in evolutionary theory. But I’d hope that Darwin, were he alive today, wouldn’t balk at the non-traditional steps I’m about to take.
First, we’d need to accept differential persistence – mere survival – as a legitimate form or mechanism of natural selection. An analogy is this: imagine there are 1,000 radioactive atoms in the process of decaying (losing particles or photons from the nucleus). The few atoms that are left several half-lives later (the time it takes for half the material to decay) are no different from those that decayed at the beginning; they are just luckier. But if there were any ‘mutations’ that gave atoms the ability to resist decay, then those left intact after several half-lives would be more likely to possess such mutations than those that decayed at the beginning. This seems like a sort of natural selection to me; and a first such ‘mutation’ would buy time for a second to occur, so ‘complex adaptations’ might be achieved.
Maybe it’s not possible for radioactive atoms to acquire stabilising mutations. And maybe it’s highly unlikely that whole biospheres, with multiple independent evolutionary lineages, could do so. But the latter is possible, and in biology we should be comfortable with this kind of reasoning. We’re quite content to say that the beneficial mutations that explain why some organisms are selected over others are also highly unlikely – it’s just that natural selection co-opts these improbable events and makes their ultimate success probable over the long run. Here, then, is where something resembling Darwinian thinking could come in.
Put another way, what selection really accomplishes is an increase in the ratio of selec
- In the penultimate heat, Stanford, Georgia, Florida and Kentucky squared off. Florida’s Talia Bates got the Gators off to a good start with a