Stability in Nature – It is said in nature that where many kinds of plants and animals live together there will be a better balance than where there are only a few kinds. This is to say that complexity leads to stability. Ecologists have been saying, quite loudly, and they have been quoted by those concerned about the human impact on our planet. But ecologists are now tending to eat their words. What follows are both the words and the eating of them.
The stability of an argument can best be understood in an extreme example. If only two kinds of animals exist on an island» say foxes and the rabbits they eat, then the future of both kinds looks highly uncertain. If some accident killed off many of the rabbits, this would be very unfortunate for the foxes, most of whom would starve.
The few rabbits survivors of the physical catastrophe would also be in an extremely precarious state because they would be hunted down by the relatively numerous and desperate foxes. But if the natural accident happened to the foxes, then the numbers of the rabbits might get out of hand until more foxes had been bred to eat them up, by which time there might be too many foxes, and so on. The fox-rabbit system would be dangerously unstable.
But if instead of two kinds of animals there were as many as ten different kinds of rodents on the island living with the foxes say several kinds each of rats, mice, and voles. And if there were two or three other kinds of flesh-eater as well, say cats and weasels in addition to the foxes, then in this well-populated island, a catastrophe to any one kind of animal would not matter very much.
If the rabbits on this island suffered a catastrophic loss, the predators would still be safe, being able to feed on the other nine kinds of rodents. The rabbits might also survive their catastrophe because it might not be worth any of the predators* time and effort to specialize in hunting rare rabbits, and the few remaining rabbits might be left alone to pursue their exuberant breeding policy to make good the loss.
Similarly, if the foxes suffered an accident, there would be no wild fluctuation in rodent numbers because the cats and weasels would be there to carry on the hunting. They might even expand their appetites to eat the rodents left by the missing foxes. Life on this hypothetical well-populated island, therefore, should be stable and safe from extinction.
This is the essential part of the complexity-stability theory, a straightforward idea whose only unusual or tricky aspect is that it is the complex that is stable. The theory has deep appeal to naturalists for it fits the intuitive idea of complex nature working well. This feeling was there when we looked at the great plant formations and saw them as entities, with territories set aside to teach.
Then the search for plant societies as real communities of species interacting to preserve the common order revealed the same thoughts. So did the idea of successional communities being but subordinate stages in the building of a climax formation. And the idea is very strongly present in all thoughts of a balance in nature set by predators that are supposed to control the numbers of everything; the spiders that are “good” because they kill “flies” and the wolves that are “bad” because they kill “game.”
But the theory only became important in modem ecology when claims appeared that it had a firm basis in mathematics and satisfyingly erudite mathematics at that. The erudition may have been the snare in which we were caught, for the mathematics never said what ecologists came to think it said. Telephone engineers of the Bell System’s laboratories did the maths.
They were interested in complicated networks of channels down which messages flowed, and the mathematics they devised is called “information theory.” The theory provides a measure of the diversity of channels in a network, called the “Shannon-Wiener information measure” after its authors. Maths also states the relationship between this measure and the capacity of an information channel. If this strikes the casual reader as apparently having little to do with biology, this shows the reader’s good sense.
To get from the Bell System to an ecosystem we first use the Shannon-Wiener measure to describe the diversity of species in a biological community and then we indulge in risky analogy as we compare one system with the other, intuitively. In the first stage in this process, the use of the measure seems to be reasonable and useful in that it speaks to a very real difficulty we have in describing biological systems.
The measure helps with our perennial problem of the common species and the rare, particularly the proper description of commonness and rarity. It is an easy matter to list all the species in a community and to compare the species lists of two communities in the ways of plant sociologists. But what if two communities are made of the same species but these appear in different proportions? Obviously, the communities and the ecosystems that support them will be different.
We say that the two have the same “species richness” but different “species diversity.” Those who love the English language will notice that we have given our own special meaning to “diversity.”
We use the Shannon-Wiener formula to measure diversity in collections of species because it allows us to collapse estimates of species richness and species commonness into a single statement. There is a large ecological literature on when and how to do this and ecology has benefited from the practice.
But it is from here that the errors begin because a measure of species diversity must also be a measure of complexity. And the original information theory gave both a measure of the array of alternative channels (diversity) and of the capacity of a channel for the flow of information, which gave the stability of the flow.
If the measure describes both complexity and stability, which it does for the system of message channels, it is very tempting to think that Shannon-Wiener measures both complexity and stability when applied to biological systems also. And so there we have the snare. We use a measure from another discipline to describe the diversity of our ecosystem and find that it does do so in a general way.
But then we notice that the measure also describes stability in the phenomena of that other discipline, and we are tempted to make the claim that the measure describes stability for our phenomena too. But the phenomena are not the same. We get from Bell System to the ecosystem by analogy only.
Ecologists in the late 1950s suddenly became aware that the telephone men had produced a body of the theory that seemed directly to relate complexity to stability in physical systems. Ecologists were thinking “systems,” and were in fact actively teaching their students that “the ecosystem” was the unit to study.
And here were systems theorists with elegant mathematics purporting to show that complex systems (ecosystems?) should be stable. It was no more than what an ecologist had always expected. Those richly diverse communities that botanists had once called “formations” or “associations,” and which Tansley had said were to be looked upon as parts of “ecosystems,” were able to persist because their complexity gave them stability.
Natural history literature contains many anecdotes that support this view. On the one hand, we have the rain forests of the equatorial basins, biologically rich places with more species than anywhere else on earth. These were communities of immense complexity, and we thought of them as unchanging, timeless ecosystems that had endured for whole epochs in uneventful sameness.
On the other hand, were the arctic tundra, with few species, where the records of fur traders told us of violent oscillations in the numbers of animals, and from which came stories of lemmings taking intermittent marches to the sea. The complex place was stable and the simple place unstable; just as the theory predicted.
More potent still for the success of the theory was its apparent usefulness in describing the difficulties known to farmers. Western agriculture works by clearing the wild complexity away and substituting a single crop. Where there were formerly deciduous forests or prairies, with their complex lists of species, we substituted monoculture; one immensely common plant with a few hangers-on in the form of weeds.
This is creating extreme simplicity where before the system was complex. Information theory predicts that the new ecosystems built by the farmer should be unstable and, lo, the fields of agriculture are afflicted with plagues of weeds and plagues of pests. It seemed like an ecological judgment.
But a closer look at these anecdotes causes disquiet. The instability of arctic animal populations seems clear to have something to do with a highly unstable climate. Indeed, we call upon the vagaries of the arctic weather for our best explanation of why the area is depauperate, saying that species go extinct in the arctic so quickly that a large species list cannot collect.
And we explain the rich species list of the equatorial forests as being since there is so stable a climate in the equatorial lowlands that extinction is rare, letting more and more species accumulate. This introduces a dilemma of priority; does a large species list promote a stable life? Or does stable living in a place of stable climate promote a large species list?
Adding to these doubts was the realization that we were not sure that life in the equatorial forests was stable. We had very little data because very few modem biologists have lived there. Western civilization and its biologists are products of a narrow band of latitude circling the northern hemisphere, halfway up toward the pole. We have more than a passing interest in what goes on to the north because we hunt arctic animals for their fur.
People of our northern outposts have reported what they have seen, and when they have seen something unusual, they have reported it more vehemently. But we have much less news of the rain forests, nor have we had a commercial interest in the systematic collection of small tropical animals. If there have been plagues of mice or monkeys along the Zaire River or in Borneo, we have had no resident scholars there to write to the Times about them.
Now things are beginning to change. Recently a scientist with long experience in the arctic settled in Panama to work and wrote to a scientific journal saying that he had seen as many rodent plagues in four years in Panama as he had during his longer living in the arctic before. I recently flew low over the rainforest in Ecuador and saw scattered trees that had lost all their leaves, perhaps because of a population event in the caterpillars that feed on them.
Many years ago, in Nigeria, I saw the same sort of thing from the ground. One species of tree in the local rain forest was suddenly easy to spot because it was without leaves. A plague of caterpillars had totally defoliated it.
These stories are only anecdotes. But so are the accounts of fluctuating numbers in the north. Neither is a measure of stability; both are merely accounts of fluctuations in the numbers of individual species. The point is that we realize that we are probably going to be able to match descriptions of population events in the depauperate north with descriptions of similar events in equatorial places, where large arrays of species live. We cannot rely on comparisons between latitudes to support complexity- stability theory, quite apart from the difficulty of prime causes introduced by different climates.
The arguments based on agriculture, when examined closely, are even weaker. They say that a very simple system such as the ones the farmer makes should not work at all. A field of monoculture could be likened to my first model of an island inhabited only by rabbits and foxes, with the crop playing the part of the rabbit and the farmer or his pests playing the part of the fox.
The system should be wildly unstable, which means that agriculture should not work. But Western-style agriculture does work, very well indeed. The crops and the farmers both thrive, as they have done for the ten thousand years during which agricultural systems have become even simpler. It is a triumph of stability.
There are troubles of the eggs-in-one-basket kind for farmers in practicing monoculture, but these are not strictly relevant to the complexity-stability argument. When accidents happen to a monoculture crop, they are likely to be catastrophic to local economies, but this is the consequence of not spreading the economic risk and does not speak to the fate of the crop species itself. It is probably true to say that the fortunes of crop plants have little to do with the properties of the system of simple communities.
If there is also no more instability in the north than in the tropics, which cannot be accounted for by the instability of weather, then there are no general biological observations that can be used to support the theory. It becomes no more than an echo of beliefs in natural organization held by the old naturalists who thought that there were self-organizing powers in plant societies or ecological succession.
The information theory itself is certainly valid. Systems that function through an array of intersecting pathways that provide alternative channels for the flow of information or energy do, indeed, become more stable the more crossroads there are. The colossal error behind the application of the theory to biology is in imagining that animals and plants in a food web act as the necessary crossroads.
Real animals and plants do not conduct themselves as channels for the transfer of that important form of “information” or energy called “food.” They work to stop the food from moving. Every individual of every species in the community is working their hardest to secure food and prevent others from taking it.
The information theory description of a food web sees everyone as a channel at a crossroads throughout which food freely passes, but real individuals are in fact roadblocks through which food gets with difficulty. It is this fact that makes the model not only unreal but absurd.
The ecosystem concept is beautiful, letting us express our understanding of how the doings of every kind of living and physical process in the habitat may affect the fortunes of all. With information theory, however, we stretch the systems analogy too far.
It requires that animals and plants act in ways we know they do not. In particular, the theory relies heavily on the efficiency of predators, expecting them not only to control their prey in a very simplistic manner but to be catholic of taste so that they can turn their formidable mouths to whatever victims happen to be plentiful. But we know that real predators do not work like this.
Most hunting animals are small insects, like wasps and beetles, and these are highly programmed to hunt kinds of prey. They do not switch their attention from target to target as the theory requires. Waging their guerrilla war of hiding and seek through a tropical rain forest a kind of wasp and a kind of caterpillar are as alone as the foxes and rabbits of my imaginary depauperate island. Their fortunes are not given stability by the presence of neighbors. They persist only by the logic of run and scatter, search and destroy. They would do the same in whatever community they lived in.
For the herbivores, the hunters who eat plants, the reality is the same. Each specializes in its own kind of plant, or its own few kinds so that the system of interchangeable channels required by information theory does not exist. Again, this is most true for the small insect hunters, which are often totally dependent on a single plant species for their livelihood. But most of the complexity of species in the tropics is made up of insects and the plants they hunt.
In real communities, the animals and plants live much of their lives in isolation from their neighbors of other kinds. It is peaceful coexistence, as the exclusion principle tells us, not the constant death in a skirmish that the information theory model requires. Recently this biological theme has been taken up by mathematic individuals are in fact roadblocks through which food gets with difficulty. It is this fact that makes the model not only unreal but absurd.
The ecosystem concept is beautiful, letting us express our understanding of how the doings of every kind of living and physical process in the habitat may affect the fortunes of all. With information theory, however, we stretch the systems analogy too far. It requires that animals and plants act in ways we know they do not.
In particular, the theory relies heavily on the efficiency of predators, expecting them not only to control their prey in a very simplistic manner but to be catholic of taste so that they can turn their formidable mouths to whatever victims happen to be plentiful. But we know that real predators do not work like this.
Most hunting animals are small insects, like wasps and beetles, and these are highly programmed to hunt kinds of prey. They do not switch their attention from target to target as the theory requires. Waging their guerrilla war of hiding and seek through a tropical rain forest a kind of wasp and a kind of caterpillar are as alone as the foxes and rabbits of my imaginary depauperate island. Their fortunes are not given stability by the presence of neighbors. They persist only by the logic of run and scatter, search and destroy. They would do the same in whatever community they lived in.
For the herbivores, the hunters who eat plants, the reality is the same. Each specializes in its own kind of plant, or its own few kinds so that the system of interchangeable channels required by information theory does not exist. Again, this is most true for the small insect hunters, which are often totally dependent on a single plant species for their livelihood.
But most of the complexity of species in the tropics is made up of insects and the plants they hunt. In real communities, the animals and plants live much of their lives in isolation from their neighbors of other kinds. It is peaceful coexistence, as the exclusion principle tells us, not the constant death in a skirmish that the information theory model requires.
Recently this biological theme has been taken up by mathematical ecologists. Hitherto we had been applying to ecosystems the mathematics appropriate to telephone networks, or simplified physical systems provided with free-flowing feedback loops. It has led us to great error. But now the first systems models are being made on assumptions that the units in the systems behave as we know animals behave, where the feedback between one event and another is resisted or delayed.
For these models, there is no simple relationship between the complexity of species lists and stability in the lives of populations. Indeed, a common result is quite the reverse. In some of these models, if a complex “community” is perturbed, the result is not stability but a reinforcement of the stress, a domino effect of increasing instability the more species there are. There is a resonance, with the original fluctuation being amplified as the shock travels through a complex community.
The claim that complex communities are more stable than simple communities, therefore, it is invalid. It is an echo of the wishful thinking of naturalists, amplified by mathematics they did not understand. Hence, it has done mischief by distracting people from real problems. And it has, for instance, been invoked in the controversy over the Alaska pipeline.
In the claim that the arctic ecosystem is “fragile” (it is simple don’t you see). But this is nonsense. The animals and plants of the arctic spend their whole lives and evolutionary experience struggling against adversities far mightier than any pipeline or road. Fluctuating numbers are normal conditions in many of their lives, and all of them will outlast oil-hungry people.
The Alaskan pipeline is a disaster to the American heritage, both for the aesthetic damage it does to the last wilderness and for the encouragement, it gives to the continued misuse of fuel reserves. I wish very deeply it could have been stopped. But the argument that it is damaging a fragile ecosystem is false. Far more serious for the future is the trans-Amazonian highway because this brings the shock of human activities to a rich variety of tropical species so unaccustomed to shock that many will disappear forever with the coming of the road.
Many species in an ecosystem do not, of themselves, lead to population stability. The stability of the climate, on the other hand, leads to the collection of many species. This seems to be the essential truth of the matter.
But then what does cause the balance we see about us in nature? Many different things conspire to preserve the continuity of life, which is what we mean by balance, but central to them all is the fact that every species is equipped with a strategy for life that lets it persist. A climax tree of the forest has the strategy of holding ground, long life, and growing like a baby in the shade of its mother.
It takes generations, a hurricane, or a plague to displace climax trees, yet both hurricanes and plagues are rare. And among the true climax trees are patches where holes are being filled by succession, but change is slow even in these places. So generations of people see the same forests, even though they will not last forever.
Weed plants come and go, suffering numerous upheavals, but their opportunist strategies let them plant a new generation in some new spot as fast as the old is deposed from the parental site. The coming and going of the weeds are always with us, and we see in this continuity part of that general balance.
Herbivores hunt down their plant prey and move on, but the piece of land they clear will immediately be taken by another plant because that land is unfailingly supplied by the energy of the sun. Plants and their persecutors go on with their endless game of musical chairs, and we see the result as balance.
Insect predators and the other hosts of small hunters pursue their games of search and kill, hide and seek, with their scattered, mobile quarry. This too tells us of persistence, we say “balance.” The larger predators maintain themselves on the old and sick; they are long-lived and must survive the winter; they must, therefore, be rare and will suffer deaths of privation if there come to be too many of them.
This too contributes to the general balance and comes closer to the tooth-and-claw model of balance set by quarrels about limited resources. So too do the activities of the big hunters when they kill the young of their prey, suppressing the population of their victims and thus limiting their own eventual supply of old and sick.
The birds and many other vertebrate animals also have complicated patterns of behavior, which are necessary for them to rear their babies and to survive hazards, and all these have population effects. The territorial habit, whether giving an advantage through food, the union of parents, or sex, always carries with it the possibility of setting an upper limit to numbers.
So, do all the hierarchical structures of social animals. None of these things has evolved to promote “balance” by restricting breeding, but all may tend to have that effect. These habits mean that many of the more conspicuous activities the naturalist sees are nearly the same every year. This gives us a feeling of a general balance in nature that would not have been so strong if we had fastened our minds on plant-eating insects or ichneumon wasps or spiders.
It remains true that the natural balance involves great destruction every year because all species are breeding as hard as they can. But this natural destruction mostly falls on eggs or the young. Beetles drill nearly every acorn. Dandelion seeds floating under their parachutes mostly fall on stony ground. The hunting wasps get growing caterpillars. Yearling animals are the ones who fail in hard winters.
And, when times are very hard through too much crowding, as can happen in nature and always happens in laboratory cages replete with food, it is eggs, embryos, and young that are starved, even as the old die untimely deaths. It is very hard to raise babies in the real world. It is the unmade or the unfinished animal and plant that succumbs. In a sense, nature favors abortion rather than later decimation by tooth and claw.
