Chapter 1

THE IMPROBABILITY OF THE HORSE

Of the more than 4,000 species of mammals that have occupied the earth during the last 10,000 years, the horse is one of fewer than a dozen that have achieved widespread success as domesticated animals.

That low success rate was certainly not for want of trying on our part. The ancient Egyptians attempted to domesticate hyenas, antelope, ibex, and gazelles (figure 1.1). The American Indians kept pet raccoons, bears, and even moose. The Australian aborigines even kept wallabies and kangaroos. Yet none survive as domesticated animals today.

If it were simply a matter of human will, it would be hard to explain why we should have domestic dogs, sheep, goats, cows, pigs, horses, asses, camels, rabbits, and cats -- but not deer, squirrels, foxes, antelope, or even hippos and zebras.

The answer is that it was not a matter of human will. The successful domesticated species were largely "preadapted" to their role through quirks of adaptation and evolution that had nothing whatever to do with human intentions or needs, but that turned out to be vital to their future success in our homes and fields. The horse was no exception. Among the myriad ways of making a living that evolution has cast up, a few -- a very few -- turned out to be compatible with human ways.

The horse, like the other animals that were to enter into domestication, was a generalist, able to survive on a variety of widely available foods. (An animal such as the giant panda, which eats nothing but bamboo leaves, would surely have been a nonstarter.) These generalists were able to exploit their new domesticated niche with a high potential reproduction rate. They had relatively simple courtship patterns, typically harems in which one male readily mates with multiple females. They were social animals, instinctively given to understanding signals of dominance and submission. They were relatively nonterritorial, not given to disruptive intraspecies combat over fixed bits of ground. In other words, the first step toward domestication was one that nature took millions of years before we even arrived on the scene.

The second step also seems to have been more the doing of the animals than of us. It was animals who discovered the mutual compatibility of our species, and it was they who chose to act upon this discovery. Recent archaeological and animal behavior studies strongly support the idea that domestication was not the human invention it was long supposed to have been, but rather a long, slow process of mutual adaptation, of "coevolution," in which those animals that began to hang around the first permanent human settlements gained more than they lost. Some were killed and eaten, but for every cow or sheep or horse killed, many more flourished on the crops they robbed from our fields and the incidental protection they gained from other predators in the proximity of human habitations. Like the starlings, mice and rats, and chimney swifts that invade our homes today for the food and shelter that are a by-product of our domestic habits, those forebears of our domestic stock took the initiative. We followed.

In the process, these semidomesticated but still free-living animals acquired still more of the characteristics that would make full domestication possible. Those individuals that were more curious, less territorial, less aggressive, more dependent, better able to deflect human aggression through submission, were the individuals that had the edge in this new niche.

It was only in the third stage of domestication, when humans began breeding animals in captivity, that human "invention" began to play a predominant role. But in consciously selecting and emphasizing those traits that appealed to our fancy or our needs, we could still only draw upon what nature provided. If the horse had not existed, we most definitely could not have invented it. The species upon which agriculture and indeed civilization have been built were a remarkable gift of evolutionary chance and opportunism.

The Disadvantage of Being a Horse

A number of species that expanded their range throughout the world during the respites that punctuated the Ice Age glaciations of the Pleistocene epoch (15,000 to 2 million years ago) may have acquired "domesticated" traits as a package deal. The twin pressures of climatic upheaval and massive hunting by humans placed specialist feeders at a decided disadvantage; it was the generalists, which could adapt to a wide variety of climates and circumstances, that flourished. Moreover, it was those species that were migratory, curious, adaptable -- as opposed to territorial, suspicious, and conservative -- that thrived on upheaval. Sheep, wolves, cattle, goats, camels, and horses all fit this bill. Thus in at least one sense the coincidence of domestication seems a bit more comprehensible: there were sound evolutionary reasons that made horses ripe for domestication. Even tameability may have been part of this package. The pressures that placed a premium on adaptability favored the retention into adulthood of juvenile characteristics, an evolutionary process known as neoteny. Juveniles are curious and adaptable, traits that were in demand in the Pleistocene; they are also playful, submissive, and dependent, traits that proved valuable to man in domestication.

Yet horses possessed a number of other remarkably convenient characteristics -- convenient from our point of view -- that make the existence of the modern horse seem all the more astonishing. To begin with, its very survival to modern times was practically a fluke. Many things worked against the horse ever making it. The speed, size, and weight-bearing capacity of the modern horse, all vital to its utility to humans, are extraordinarily unusual among mammals. An animal the size of the horse is in fact a prime candidate for extinction, a fact borne out repeatedly in the fossil record of life on earth. Large animals are long-lived as a rule, but they are also slow to reach sexual maturity, require long gestation periods, and rarely bear more than one young at a time. A drought, an insect outbreak that strips vegetation bare, or any other climatic or ecological disturbance can deliver a blow to a population of large-bodied animals that it takes years to recover from -- if it recovers at all. A population of small animals that reproduce quickly and in large litters, on the other hand, can bounce back from repeated calamities.

Large animals face other evolutionary risks, not least of all gravity. Thomas McMahon, a biomechanician at Harvard University who has made an extensive study of the biology of size, notes that when an animal falls the damage it does to itself is directly proportional to its height, or length. (McMahon's argument in a nutshell: The strength of bones is proportional to their cross-sectional area. An animal twice as tall or twice as long as another will have bones that are twice as thick in all of their dimensions; their cross-sectional area will thus be 2 x 2 = 4 times as great [see figure 1.2]. If an animal's body length is proportional to L, the cross-sectional area of its bones will be proportional to length squared, L2. On the other hand, the energy of a falling body is proportional to its mass, and mass typically increases with the cube of an animal's length, or L3, because an animal twice as long as another will, roughly speaking, also be twice as wide and twice as tall, so its total volume will be eight times as great: 2 x 2 x 2 = 8. So if the energy of an impact is proportional to length cubed, and the ability to resist damage is proportional to length squared, the ratio of the two is [L3/L2], which equals just plain L. The old adage is true: the bigger they are, the harder they fall.) As the renowned British zoologist J. B. S. Haldane observed, "You can drop a mouse down a thousand-yard mine shaft; and, on arriving at the bottom, it gets a slight shock and walks away, provided the ground is fairly soft. A rat is killed, a man is broken, a horse splashes."

As we shall see, there were some compensating evolutionary virtues that made large size attractive, but that does not alter the initial fact of the relative rarity, and improbability, of animals the size of horses.

The speed of the horse is a rarity, too; indeed, it has few equals in speed over long distances. The cheetah, the fastest land animal, can hit 100 kilometers per hour for extremely short bursts over a distance of a hundred meters or so; a horse's sustained racing gallop is nearly 70 kilometers per hour.

At the same time, the horse's dietary needs are astonishingly easy to meet for such a large animal. The black rhino, an animal whose body weight is comparable to that of a large draft horse, requires huge quantities of green twigs and legumes. Large carnivores of similar weight, such as polar bears or lions, will eat up to 40 kilograms of meat in a single meal. Yet the horse has adapted to eating the poorest-quality forage, containing the lowest concentration of protein, of any large herbivore. It thrives on grasses that a cow would starve to death on.

And unlike many large grazing animals, horses break down the otherwise indigestible cellulose of stems and leaves in a digestive organ known as the cecum. While ruminants such as sheep and cattle have to rest for hours after eating, the hoofed herbivores that possess a cecal digestive system (horses, tapirs, and rhinos) can eat and run.

It is another remarkable and happy coincidence that the horse possesses a diastema, or gap between the front incisors and the rear grinders (figure 1.3). Control of the horse by man would have proved far more difficult without this anatomical convenience, which allowed the effective placement of bridle and bit.

Horses possess one final anatomical convenience from our point of view: almost unique among hoofed animals, they lack horns or antlers.

The Nature of Evolution

The sheer abundance of horse bones, and especially horse teeth, in the fossil record has made the horse the single most frequently cited paradigm of evolution. There are more than half a million specimens of fossil horses in museum and academic collections in North America alone.

Explaining evolution has never been easy, and educators and museum curators quite naturally seized on the well-documented fossil history of the ever popular horse as Exhibit A. Practically everyone who has visited a science museum or taken an elementary biology course has seen the evolutionary sequence of fossil horses from tiny eohippus (more properly known by its scientific name, Hyracotherium) to modern Equus. Starting as a small, squat, dog-sized, four-toed creature 55 million years ago, the horse step-by-step turned into the tall, fleet, elegant, single-hoofed animal of modernity.

But this simple, linear picture of evolution has led to a couple of unfortunate misunderstandings. One is the idea that evolution really does run in a straight line. This notion, known technically as orthogenesis, was assumed by many early biologists, and is still the popular conception of how evolution works: each species in an animal's fossil family tree gives rise to a (presumably superior) replacement. "The orthogenetic template has...influenced millions of lay people, many of whom visit natural history museums with turn-of-the-century exhibits that convey 100-year-old ideas," says paleontologist Bruce MacFadden, an expert on fossil horses. In fact, paleontologists now know that evolution is full of branches, dead ends, and blind turns.

The recent history of modern equids is no exception. Analysis of the similarity of mitochondrial DNA among the seven modern species of the genus Equus (three zebras, two asses, the horse, and the now-extinct, zebralike quagga of South Africa, which was hunted to its death in the late-nineteenth century) has allowed scientists to reconstruct a branching family tree that is surely far more paradigmatic of evolution than any straight-line model (plate 1). Mitochondrial DNA is the genetic material found in a portion of the cell known as the mitochondria, which is responsible for generating energy and which has the peculiarity of reproducing itself separately from the rest of the cell. Mitochondrial DNA is inherited solely from the mother. Thus, changes in mitochondrial DNA can occur only slowly, by accumulated mutation, not by the recombination of male and female genes through mating in every generation. By comparing this DNA in related species and estimating the mutation rate at 2 percent every million years, biologists have calculated when the various modern Equus species diverged from one another. (The extinct quagga's mitochondrial DNA was extracted from a bit of muscle tissue in a preserved hide at the Museum of Natural History in Mainz, Germany; it appears to have diverged from Equus zebra 3 to 4 million years ago.)

The other, related misperception that the orthogenetic model of evolution has engendered is that evolution has a purpose, or goal. It is commonplace, and perhaps inevitable, for people in love with horses to see this 55-million-year history as a process of "perfecting" the horse. We often read that the horse is "perfectly" adapted to running, for instance, and it is hard for us not to see modern Equus as superior to its forebears.

In fact, the evolutionary explosion of the horse in North America beginning some 20 million years ago gave rise to a multiplicity of other branches, too, with as many as 13 genera existing simultaneously. Some species were larger than their predecessors, but some were smaller (figure 1.4). Some moved toward the "modern" horse diet of grasses, but others specialized in the "primitive" diet of browse (figure 1.5). Some showed a trend toward the one-toed hoof of the modern equid, others did not.

The first step toward understanding how evolutionary forces shaped the modern horse is to understand that the purpose of evolution is not progress or perfection in the long run, but survival in the short run. And that in turn involves a complex interaction of genes and the environment. It is simply wrong to say that the modern horse is "superior" to its extinct predecessors. Many of those predecessors that we so cavalierly dismiss as failures, or as inferior stepping stones on the path to perfection, were in fact brilliant successes that flourished for millions of years -- until an unpredictable change in climate finally did them in. The evolutionary success of the modern horse owes more to its having been a lucky guesser than a pinnacle of progress. That is especially borne out by the fact that abrupt climatic changes at the end of the Ice Age some 15,000 years ago (possibly exacerbated by overhunting by humans) drove the modern horse to extinction in North America and within a hair's breadth of extinction in Europe and Asia as well. Were it not for domestication, Equus caballus would have gone the way of Hyracotherium and all the other ancestral horses that are testimony to the inevitability of extinction.

A new trait, however meritorious it may prove in the long run, cannot become fixed in a population unless it confers some immediate advantage. A trait that increases the odds of an individual's surviving to the age of sexual maturity, successfully finding a mate, and bearing a large number of offspring will be a trait that is preferentially passed on to the genes of the next generation. A trait that is less efficacious in achieving these ends will be preferentially weeded out of the population.

Still, there are general trends that can be traced in the evolution of what paleontologists call clades -- large families of related species. All species that survive today are descendants of species that defied the dead end of extinction. And we also know that evolution can be very conservative. Some traits, selected for at some point in the course of evolution, tend to persist because it is in effect too costly or difficult to weed them out. To change the metaphor, they are trap doors: once evolution has passed through them it is very hard to go back. In each generation, natural selection can only use the raw materials that previous generations of selection have left it. Thus, over long periods of time, certain branches will prove more successful than others; certain trends will emerge. But again, it is always worth remembering that these trends reflect what are in effect the sum total of lucky guesses or accidents. They are the long-term consequences of short-term "decisions," consequences that could not have been anticipated at the time those choices were made. Most extinct species were victims of their own success -- they had the misfortune of being supremely adapted to a niche that did not last.

The Nature of Luck

One general trend that appears early on in the evolution of horses, and which sets them apart from their fellow herbivores in the Eocene epoch (55 million years ago), was a significantly larger brain. In particular horses developed an expanded neocortex -- the part of the brain unique to mammals that is responsible for learning and for correlating multiple sensory inputs.

But if horses were smarter than dinosaurs, it was not because of any inherent tendency toward progress in evolution. There are plenty of phenomenally stupid creatures that make a remarkably fine living these days, and plenty of perfectly smart creatures that ended up in the dustbin of evolution.

Evolutionary biologists look for more narrow selective forces to explain the trends they see in the fossil record. And Hyracotherium's development of a larger brain may initially have been related to a need for increased tactile sensitivity of the lips, important to a grazer or browser's ability to efficiently select desirable forage.

The earliest equids were uniformly browsers; that is, they ate the shoots and leaves of trees and woody plants, which are relatively higher in nutritive value than grasses. At this time in the earth's history, such food was also much more abundant. A remarkable fossil specimen of the equoid Propalaeotherium from the late Eocene epoch was found complete with fossilized stomach contents, which included grape pits. These were forest dwellers well adapted to the lush, moist conditions of this period.

"You are what you eat" may not be a strictly accurate nutritional statement but it certainly has a lot of merit as an evolutionary principle. Much of the evolution of the modern horse -- including its speed, size, and intelligence -- can be explained by diet and changes in diet. As the global climate became drier, savannas and grasslands began to displace forests. After millions of years of relatively slow evolution, the equids showed a sudden burst of diversification beginning about 18 million years ago, just as these sweeping climatic changes began. The browsers began to decline -- though they by no means vanished, holding on for another 9 million years. The key word here is diversification. New niches made many more successful options available. Equids had all been confined within a relatively narrow body type up until this point; body weight of the various equid species ranged only from about 25 to 50 kilograms for the first half of their evolutionary history. The climatic changes of the Miocene epoch beginning about 18 million years ago gave rise to branches that ranged in size from 75 to 500 kilograms. Of 24 ancestral-descendant pairs that paleontologist Bruce MacFadden examined from this period, 19 showed an increase in size, while 5 decreased.

Many of the distinctive and, from our point of view, useful traits of the modern horse first arose in this epoch within those branches of the equid family that would ultimately survive to modern times. The grass-eating branches faced a number of new problems that required new adaptations. Grass is not only tough; the cellulose that contains most of its available nutrients is locked in the plant's cell walls and is basically indigestible to mammals without special tricks (more on this below). To start with, it requires a lot of chewing. While fruits are designed to attract animals as part of a plant's strategy for spreading its seeds and reproducing by offering a tempting (and discardable) appendage, the leaves of a grass are its all. Plants subject to herbivory have therefore evolved an array of defenses, one of which is the incorporation of tough silica particles, known as phytoliths, into their cell walls. And food eaten off the ground is already full of dirt, all of which means herbivores need to have much tougher teeth than browsers.

The grazing horses begin to show rapid adaptations to this problem. The compressive action of the jaws and teeth of the browsing horses is replaced in grazers with a transverse shearing action that acts to grind the food. The depth and the size of the jaw increase to make room for more powerful muscles. The problem of tooth durability is solved by several tricks (figure 1.6). Premolars change to full molars, increasing the surface area available for grinding. Teeth become cement-covered, higher-crowned (hypsodont), and eventually ever-growing (hypselodont).

Interestingly, the initial change toward hypsodonty may have been not so much an adaptation to grasslands as to dirt. At at least some sites in the interior of North America, it has been shown that the early Miocene brought a change to coarser sand in place of finer clay soils. When the grasslands later began to spread in the middle Miocene, those clades that had already evolved toward tougher and bigger teeth were better positioned to exploit this new niche.

This is also the period when the "springing" foot appears in the equids and when the trend toward substantially larger size is first seen, at least in the clade that would lead to the modern horse. The springing foot (plate 2) is basically an arrangement of tendons that allows elastic energy to be stored and reapplied with each stride, making for more efficient locomotion (more on this in chapter 8). And larger size is one way (though as we shall see, only a relatively modest way) to increase top speed.

Size and Speed

The usual explanation for these changes, and for the eventual appearance of the single-toed foot in Pliocene horses (beginning 5 million years ago) is that, as grassland animals, these grazing horses were more exposed to predators and had to be able to flee. The evolution of the diastema may well be related to this fact, too; a long distance between the front of the mouth and the eyes allows an animal to graze and keep an eye out at the same time.

There is no doubt that larger animals are faster, and that the springing hoof allowed for a faster gait. The almost unbelievable discovery of fossil footprints of three Hipparion horses from the middle Pliocene (3.5 million years ago) has provided ample confirmation of the speed and agility of these grasslands-adapted horses (figure 1.7). Although Hipparion still had three toes on each foot, it had already developed the springing foot mechanism; and in spite of its relatively small size (about 1 meter in length), its overall proportions -- leg length relative to body size, for instance -- are quite similar to those of modern horses. The two side toes in Hipparion species, while able to help balance the foot and even add some to the locomotive effort, were already much reduced in size compared to those of their ancestors, and clearly most of the work was done by the large central toe.

The Hipparion footprints, made in soft lava subsequently covered with volcanic ash, were discovered in Tanzania by Mary Leakey in 1979, along with trails of a number of other mammals, including early hominids. A subsequent analysis of the horse footprints makes a convincing case that these Hipparion horses traveled at a good clip utilizing the gait known as the running walk -- the characteristic gait of Tennessee walking horses, Icelandic ponies, and paso finos, in which the length of stride is extended and only one or two feet are in contact with the ground at any given time. Comparison of the fossil footfalls with the footfall patterns of Icelandic ponies suggests that one of the Hipparions was traveling at 15 kilometers per hour. The other two trails appeared to be those of a mother and foal, the latter crisscrossing the path of its mother in much the same fashion as is observed in modern horses. The finding incidentally provides at least some suggestive evidence in support of the contention that the running walk, though associated with only certain breeds these days, is nonetheless an instinctive and natural gait, rather than (as is sometimes argued) one that is artificial and man-taught.

If there was a selective pressure for speed, then getting bigger might seem one of the more direct ways to achieve that end. It seems obvious that bigger things run fasten But in fact that ain't necessarily so. A 700-kilogram horse does run faster than a 26-kilogram greyhound, but not 27 times as fast -- or even 3 times as fast, which is what one would expect if speed were proportional to the length of the legs or the body of an animal. (If mass is proportional to a linear dimension cubed, then an animal 27 times the mass of another would be 5 times its size in any given linear dimension, since 27 = 33.)

In fact, the top racing speeds of greyhounds and horses are not much different -- about 70 kilometers an hour for a horse, 60 kilometers an hour for a greyhound. Obviously a horse is faster than an ant, and there is indeed some correlation between speed and size that has been established both by observation and by calculation (based on fundamental mechanical principles of how muscles and limbs work). But it is a relatively modest correlation, which suggests that top speed increases only by approximately length to the three-eighths power, or mass to the one-eighth power (figure 1.8). Thus, to double in speed, an animal would need to be about 6 times as long or 250 times as heavy. By this yardstick, a 500-kilogram Equus would have a potential top speed only about 30 percent faster than that of a 50-kilogram Hyracotherium, a rather modest gain for a lot of evolution.

Size and Diet

There may have been a more important selective force driving up the size of grazing equids, of which speed was merely a byproduct. In fact, one school of thought suggests that speed and the need to escape from predators have been greatly exaggerated in the interpretation of equid evolution. Zebras, for instance, rarely try to flee from predators at all, instead forming a tight group defended by an aggressive herd stallion. Size itself is of course one defensive mechanism -- large animals are hard to attack. Rhinos, for instance, are generally safe from predation by virtue of their sheer bulk, as are elephants.

A more important consideration in explaining why horses got bigger may be the nature of the horse's diet and digestive system. As we have already observed, mammals are fundamentally incapable of breaking down cellulose, the complex sugar found in the cell walls of the structural parts of plants, such as stems and leaves. Many herbivores avoid the problem altogether by concentrating on the reproductive products of plants -- fruits, berries, and seeds. The earliest equids probably did the same. But all modern hoofed animals (except for pigs) depend more or less on a diet heavy in cellulose.

These animals solve the problem by forming a symbiotic relationship with gut bacteria that break down the cellulose into a form that the animals' own enzymes are then able to digest. It is a lengthy process, requiring that the host animal provide a fermentation chamber where large amounts of plant matter can be stored while the bacteria do their job.

There are two basic schemes for this process found among the hoofed mammals, or ungulates (figure 1.9). One is rumination, the method that sheep, cattle, goats, deer, camels, and hippos have evolved. Ruminants have four stomach chambers; the first two, the rumen and reticulum, are where fermentation takes place. Only when the plant matter has been broken down to a sufficiently small size through bacterial action, assisted by regurgitation and further chewing ("cud-chewing"), can it pass through a sievelike passage that separates this forestomach from the omasum and abomasum -- the stomach proper.

Horses, rhinos, and tapirs have a different fermentation apparatus. This consists of an organ known as the cecum, a large, dead-end alley at the junction of the small and large intestines. Careful study of the bacterial, physical, and biochemical properties of ruminant and cecal digestion has found that their basic mechanism of action is indistinguishable. Rumination, however, makes more efficient use of the food matter taken in. A horse extracts only 70 percent as much energy from a given amount of food as does a cow. This appears to be mainly because the ruminant digestive system holds the food in its "fermenting tank" for a longer time -- it takes 70 to 90 hours for food to pass through a cow, versus 48 hours in a horse. In ruminants, the small orifice between the reticulum and the omasum acts like a valve, shutting off the flow until every last bit of cellulose has been broken down.

For this reason, animal scientists long believed that ruminants were more efficient than cecal digesters in making use of plant food. And in one sense that is true. But animal behavior studies have consistently shown one extremely curious fact: wild equids in competition with ruminants invariably choose the very worst, lowest-protein, highest-fiber roughage. Equus burchelli, the plains zebra, for instance, were observed to eat the same plants as wildebeest, but they consistently ate the stems while the wildebeest ate the leaves. Wild asses and the wild Przewalski's horse live in regions containing only very poor-quality vegetation that is apparently unable to support any of the native ruminants. As researcher Christine Janis has noted, this was precisely the diet that the rumen was supposedly evolved to deal with, yet here were horses thriving where ruminants feared to tread. Experiments with domestic cattle and sheep have also confirmed that a diet containing more than a certain level of fiber cannot support a ruminant.

The reason for this apparent contradiction turns out to be quite simple. Ruminants are indeed more efficient than horses in extracting usable energy from a given weight of food. But there is a limit to how much food a ruminant can move through its system in a given period of time. As the quality of the forage declines, a ruminant's digestive system actually slows down further, giving the food more time in the rumen to be broken down. A ruminant's intake is limited by how much food it can stuff into its rumen; after that, it has to sit back and wait for fermentation to run its course.

A horse, on the other hand, can -- and does -- respond to a poor-quality diet by eating more. It may only be 70 percent as efficient in extracting usable energy from its food, but it can push a lot more food through its digestive system in a 24-hour period than a cow can. Per unit of time, if not per unit weight of food, a horse can get more energy out of a low-quality diet than can a cow of the same weight. In fact, as Janis writes, "this digestive strategy enables horses to exist on a diet on which ruminants of similar body size simply cannot maintain themselves."

What does all this have to do with body size? Like all animals, equids have chosen an ecological niche that allows them to avoid competition from other species. Their niche is the poorest-quality vegetation. For animals that start down this road, an interesting law of energy consumption dictates that the pressure to increase size becomes irresistible. Small animals need a lot more energy compared to their body weight than do large animals (figure 1.10). Thus, hummingbirds spend every second of the day feeding while lions spend hours snoozing. This is because (to a first approximation) the metabolic energy a warm-blooded animal expends just in staying alive is proportional to the heat it loses through its exposed surface area. Surface area increases with increasing size in proportion to L2. Body weight, though, increases much faster, in proportion to L3. An animal that weighs eight times as much as another has only four times as much exposed surface area. Another way to think about it is this: in a small object, all of the interior is close to a surface; in a large one, there's lots of stuff inside that's packed away far from the surface. Thus, it is easier for a larger animal to retain its body heat than a smaller one. Each cell in a sheep's body needs to run at a higher rate, and thus needs more energy, than each cell in a horse's body.

So, having adopted a strategy that depends on eating the lowest-quality, most energy-poor stuff, equids continue to gain a further competitive advantage over ruminants by getting bigger. This general principle relating body size to dietary strategy among herbivores seems to hold up well -- those species of African antelope that are predominantly grazers (such as the topi) are large, those that browse are medium-sized (Grant's gazelle), and those that eat fruit and only the leaves and young shoots of plants are the smallest (Thomson's gazelle). Once body size drops below about 5 kilograms, calculations suggest that the energy requirements of the diet are so high that it doesn't pay to ruminate at all, since nearly all of the animal's food intake would have to come from high-quality foods such as fruits and seeds. This is very close to the weight of the smallest ruminants, such as the dik-dik.

During the Eocene, the tiny, forest-dwelling Hyracotherium would have been able to find a diet with abundant high-quality forage and very little fibrous cellulose. In the tropical climate of this epoch, young plant shoots would grow all year round. But as the climate cooled and growth became seasonal, a fibrous diet would have been all that was available at certain times of the year, making fermentation a necessity. Cecal fermenters are generally at a competitive disadvantage against ruminants in the medium-fiber diet range; the horse's strategy was thus to avoid competition by choosing a diet too fibrous for ruminants to cope with at all (figure 1.11). And that made getting big a necessity. (The other possible strategy is to continue consuming a medium-fiber diet and also to get very big. Calculations suggest that once an animal reaches 1,800 kilograms, rumination provides no further advantage over cecal digestion even in the consumption of a medium-fiber diet. The black rhino, a browser, in fact weighs almost exactly 1,800 kilograms.)

Side Effects of Bigness

The equids' position at the high end of the fiber diet scale explains their considerable success as well as their considerable size. But it also may explain why, as the climate cooled and became more arid and as grasslands replaced forests, the overall diversity of equids began to narrow. There are but 6 species of living equids worldwide today, compared to some 200 species of ruminants. The number of genera of equids began to drop rapidly beginning about 13 million years ago, from a peak of 13 genera to the single extant genus.

Other consequences followed from this selection for size in response to dietary and energetic factors. Speed, as we have already seen, was one. Another was longevity, which is also closely correlated with size. It is a remarkable fact that virtually all mammals live an average of 1.5 billion heartbeats. Since small animals, as we have already seen, have faster metabolisms, their lifespans are correspondingly shorter. (Biomechanician Thomas McMahon says that this size-independent limit of 1.5 billion heartbeats may simply be a coincidence, and that other factors, such as the death and replacement of individual cells in the body, may be more proximate limiting factors in an organism's lifespan. One theory of aging holds that death is the result of accumulated genetic errors in cells as they divide and reproduce. Both cell replacement rate and heart rate might plausibly be related to an animal's size in the same proportion, thus providing the seeming correlation between heartbeats and lifespan. In either case, lifespan is roughly proportional to an animal's mass raised to the one-fourth power. The calculated lifespan for humans by this yardstick would be 33 years, which is about what it was before the Industrial Revolution.)

MacFadden has estimated the lifespan of fossil equids by measuring the height of unworn crown left on the teeth and making a reasonable guess at the wear rate per year, and his figures support the theoretically predicted trend of increasing longevity with increasing body size. Hyracotherium would have lived about 4 years; the Protohippus of 12 million years ago would have lived about 12 to 15 years; the first members of the genus Equus some 4 million years ago would have had a lifespan of 20 or more years.

Another consequence of the increasing size of equids was the evolution of mechanisms to cope with bigness. As we have seen, large animals suffer many disadvantages, not least of which is gravity. Getting up and down is hard work; a large herbivore subject to predation would not want to lie down too much, given the time it would take for it to get to its feet in a moment of danger. But standing is hard work, too, when you're big. The fossils of Pliocene equids begin to show another of the modern horse's remarkable coping mechanisms, the so-called passive stay apparatus (plate 2). Unique to equids, this arrangement of tendons and bones allows both the forelimbs and hindlimbs to lock in place while the horse is standing, obviating the need to expend muscular energy in resisting gravity. In the forelimbs, the tendons have a groove that locks into the humerus; in the hindlimbs, the patella locks into a crest on the femur. A rudimentary form of this apparatus appears first in Dinohippus about 5 million years ago and is fully developed in the first Equus. A careful measurement of energy expenditure in two domestic horses found that they actually burned about 10 percent less energy standing up than lying down. By contrast, cattle and sheep use about 10 percent more energy standing up than lying down.

The bigness of the horse is an oddity among mammals, but almost more than any other big mammals the horse has evolved means to compensate for the many disadvantages that come with size. The important lesson of all this is that its size and swiftness are most likely accidental by-products of the its ecological niche -- at the low end of the herbivore diet.

Paleoecology

With surprising accuracy, it is also possible to predict an animal's social structure from its habitat and diet. Small animals that are selective feeders in resource-rich forests tend to spread out and defend individual territories. Males who hope to find mates under these circumstances are forced to follow suit, which means that the mating system tends to be monogamous.

When females live less far apart, a single male may be able to defend a territory that includes two females, but still he is forced to go where the females are, and to defend that territory against other males. That in turn tends to select for sexual dimorphism -- differences between the sexes, such as males with larger horns or larger body size.

Larger grassland animals that rely on patchier and scarcer food sources, by contrast, tend to form into herds. This gives males the opportunity to hold harems, and so these animals tend to be polygamous.

These principles, which work well when applied to living ungulates, allow for some interesting speculation about the evolution of social organization in horses. A quarry site in Colorado known as Castillo Pocket yielded up a huge find of Hyracotherium tapirinum fossils -- a large enough sample (24 individuals) to get some meaningful statistics about the social structure of these small, primitive, forest-dwelling equids. They fit the pattern well. The ratio of males to females was between 1:1.5 and 1:2. There was a distinct sexual dimorphism; the males' skulls were on average 15 percent larger than the females', and the canine teeth of the males were 40 percent larger.

The body-size dimorphism begins to disappear in the middle Miocene grazers, and vanishes entirely after about 15 million years ago. The failure of equids ever to develop horns or antlers may be explained in part by their relatively early specialization in fibrous, low-quality forage, which made it necessary for them to cover vast ranges in search of food -- areas far too large to effectively patrol and defend as a territory. On the other hand, competition between males for control of harems does occur in modern equids, even if there is no territorial combat involved; this has led to some sexual selection for canine teeth, which stallions use in fighting and in threat displays with one another. The ratio of the male canine to molar size has increased steadily over the last 20 million years, while the ratio of the female canine to molar has declined.

The horse, like all living things, is a product of its evolutionary history. However clever it may have been for man to recognize the potential of the horse in harness or under saddle, it is to dumb luck -- and mainly to a several-million-year-old choice of diet -- that we owe the gift of the modern horse.

Copyright © 1997 by Stephen Budiansky