Ancient ants fossilized in amber.
What is evolution? It is certainly not an engineer that contrives systems perfectly designed to face all of the problems that it will face. Instead, it is a haphazard process by which entities adapt in an ad hoc manner to whatever problems are facing it at the moment. Its outcomes are not the products of a well-devised schematic, but Rube-Goldberg like contraptions that have managed to deal with their circumstances and survive by whatever slim margin.
These contraptions, unlike those products, are constituted simply by what has had survival-value, not by that which optimizes its chances in a future. Due to the Rube-Goldberg nature of evolved systems, they are vulnerable to events that are able to cause the adaptions they used to survive to function in a manner unrelated to that they were chosen for. A grim example of this is the circular-milling behavior of army ants. Seemingly apocalyptic for the afflicted ants, the behavior is a direct consequence of the ad hoc process by which evolution proceeds. The phenomenal means by which foraging army ants navigate has proved to be a great adaptation for their horde-like style of foraging, but can malfunction in such a spectacular manner.
In The Origins of Order, Stuart A. Kauffman writes of the outcomes of the evolutionary process as Rube Goldberg machines, holistic wholes that are the result of ad hoc problem-solving. Evolution in this perspective is an opportunist that incorporates the few fitness-improving mutations into historically-continent phylogenetic that continue to add-on more and more improvements. There are a myriad possible mutations that can happen to an organism, a very few would be fitness-improving. As those mutations are incorporated into the morphology of any given organism and as mutations cause the tree of phylogeny to split, organisms become more and more historical contingencies.
The radical nature of the historical contingency of all species on Earth can be understood by thinking about an extreme example: RNA, a self-replicating nucleic acid that is vitally important for the functioning of genes. It is one of the most, if not the most, molecules for the existence of life Earth. Without its emergence, one might be led to think that the entire evolutionary history of life would not have happened, but that is a mistaken inference. After all are we so sure that RNA is a necessary feature for the existence of self-replicating organic systems? Could not another self-ordering molecule have emerged in those early days, leading to an altogether alien line of phylogeny from which another great evolutionary epic, one entirely unthinkable to us, would have proceeded? That the entire phylogenetic line of this world's own evolutionary epic is dependent on RNA is not evidence that RNA is a sine qua non for the existence of organisms. Instead, it is merely evidence for how certain adaptions, once they emerge, are locked-into the phylogenetic line.
Let's turn to ants now. As Edward O. Wilson has noted, the ant is one of the evolutionary winners of natural history that has discovered a eusocial way of living around a centralized community that has enabled ants to exploit the gains to cooperation to ways that would only be beaten by Homo sapiens. As Wilson once quipped: “Karl Marx was right, socialism works, it is just that he had the wrong species.” That the ants are one of Earth's evolutionary champions is attested to by the fossil record, which has show that the family Formicidae has existed at least 92 million years, and by the prevalence of ants across just about every landmass on earth Man may very well have been the second animal to have conquered the earth.
The evolution of the ant, though, follows the Rube-Goldberg phylogenetic line above, and that brings great dangers that can lead to the failure of the organism once one of its traits malfunction, so to speak. Such a malfunction can be seen in the circular-milling behavior that can be shown by isolated groups of army ants, colloquially known as a “death spiral.” To find their way, these ants follow the pheromone trails left behind by other ants. Under normal situations, this method of navigation is extremely effective, seen perhaps most vividly in Eciton burchelli's massive raiding swarms. These sorties find prey, overwhelm it with sheer numbers, and bring it back to a temporary nest for final consumption. The tight coordination of movement, ant after another, made possible by the phonemes makes possible a massive swarm, one of nature's most awesome spectacles, from which the other species on the jungle floor can only flee from.
A short explanation of the circular-milling behavior.
However, when the pheromone trails left behind by army ants become entangled, this will result in the ants milling around in a circle until the animals die of exhaustion. That circle expands as more army ants are attracted by the pheromones, only to join the doomed group. This phenomena can even be replicated on a small-scale by putting army ants in a small enclosed space. How this behavior, seemingly so self-destructive for the overall fitness of the ants and their nest, can even be possible after so long an evolutionary history is a puzzle. Yes, the death of foraging army ants is not necessarily fitness-decreasing for the ants given that their caste does not reproduce, but when the milling patterns have the potential of entrapping so many ants, the loss of so many foragers is certainly harmful to the prosperity of the overall colony. So the puzzle remains despite the ant's centralized manner of reproduction; however, it is a puzzle that is easily solved once we remember the historically contingent nature of evolution and how the phylogeny of ants can be seen as the cumulative accumulation of useful accidents.
The circular-milling behavior of army ants is but one manifestation of what can go wrong when evolved adaptations are faced with problems that they cannot cope with. The phylogenetic history of evolution is one of the opportunistic incorporation of accidental mutations that have proved advantageous for the continuation of self-replicating objects, whether what is being considered is RNA perhaps in the primordial soup or the phenomenal navigation of army ants. A direct consequence of this ad hoc phylogeny is that seemingly pathological behavior can emerge.
Normally useful accidents can be turned transformed into self-destructive ones for the whole they've been integrated into. That is simply the nature of evolution, and it should give us pause before we seek to critique spontaneous orders by holding them up to the standard of designed schematics.