The crucial question is, how much influence do genes have? Much of the cat's personality depends on other factors. For example, whether a cat will tolerate people depends on whether it has had contact-and the right kind of contact-with people during the first eight weeks of its life. Cats that do receive such contact nevertheless vary greatly in how friendly they are to people in general, and even toward their owner. How much of this variation, not fully explained by the basic process of socialization, is inherited? Is each cat's ability to tolerate other cats due simply to whether it grows up with other cats, or are some born to be more adaptable than others?

Decoding the inheritance of personality is nowhere near as simple as the inheritance of the color or length of a cat's coat. We can track most of those visible differences among cats to twenty or so well-defined genes that operate in a highly predictable manner. If a cat's parents both have black coats, then the cat will also be black; this is not affected by whether it is born in a hedge or in a kitchen.

Genes and environment can interact in complex ways, however. Even coat colors can be affected by the environment: for example, the darker "points" on a Siamese cat's face, paws, and ears come from a temperature-sensitive mutation that prevents the hairs from taking up their usual color at normal body temperature. As newborns, these cats are whitish all over because their mother's womb is uniformly warm. As they grow and the extremities of the body become cooler, the hair there grows darker, producing the characteristic "pointed" coat. Finally, as the cat enters old age and the circulation of blood in its skin deteriorates slightly, it gradually turns brown all over.

The relationship between genetics and environment is evident in personality as well. Cat personality is influenced by hundreds of genes and a lifetime of experience, interacting together to produce the cats we see today.

To search for evidence that personality can be inherited, we might start with pedigree cats. Unlike dogs, which have been bred for different functions for many centuries, pedigree cats have been bred mainly for their looks. Deliberate selection is probably not to blame for any consistent differences in behavior between different cat breeds; we cannot expect to find differences as great as those between, say, a border collie and a Labrador retriever. However, because all pedigree cats are raised by breeders, and, at least within each country, in much the same way, any consistent behavioral differences between them are likely due to genetics.

The breeding of pedigree or "show" cats is regulated by standards laid down by the individual "breed clubs," and the best cats from each breed compete in cat shows run by organizations such as the Cat Fanciers' a.s.sociation and the International Cat a.s.sociation in the United States, and the Governing Council of the Cat Fancy in the UK. Well-known breeds or groups of breeds include the Persian or "Exotic" breeds, stocky cats with long hair and flat faces; the "Foreign" breeds, fine-boned, long-limbed cats such as the Siamese, Burmese, and Abyssinian; and the Domestic breeds, which as their name implies were originally derived from ordinary domestic cats local to the British Isles. Some individual breeds can be defined by a single mutation, such as the short, wavy coat of the Cornish Rex, the downy hair of the Sphynx, and the short tail of the Manx.

Many of the newer breeds of cat are simply color variations on existing breeds. For example, the Havana Brown is genetically indistinguishable from the Siamese, except that it lacks the mutation that causes most of the Siamese's coat to stay cream-colored. Some of the longer-established breeds claim ancient ancestry-for example, the Siamese breed is apparently described in the "Cat-Book Poems" written in the ancient Siamese city of Ayutthaya sometime between 1350 and 1750-but their DNA shows that the breeds have become separate ent.i.ties only during the past 150 years or so.1 This recent evidence separates the breeds into roughly six groups, each seemingly derived from-or possibly allowed to interbreed with-local street cats. The DNA of the Siamese, Havana Brown, Singapura, Burmese, Korat, and Birman shows not only that they are closely related, but that they are also genetically similar to the street cats of Southeast Asia from which they were undoubtedly derived. The Bobtail, a traditional j.a.panese breed, is genetically close to random-bred cats in Korea, China, and Singapore (and, presumably those of j.a.pan, which were not included in the study). The Turkish Van cat, as its name implies, is related to non-pedigree cats from Turkey, as well as Italy, Israel, and Egypt. The Siberian and Norwegian Forest cats are derived from longhaired northern European random-bred cats, while the superficially similar Maine c.o.o.n finds its closest non-pedigree relatives in New York state. Most of the stockier breeds-the American and British Shorthairs, the Chartreux, the Russian Blue, and surprisingly the Persian and Exotic breeds, are all closely related and are presumably derived from Western European stock. The modern Persian, even if some of its distant ancestors did come from the Middle East, seems to have lost most traces of its origins, possibly due to recent breeding to produce the flat (brachycephalic) face preferred by its devotees.

The various breed clubs usually describe typical personalities for their cats. For example, the Governing Council of the Cat Fancy (GCCF) describes the Ocicat, an American breed derived from Abyssinian, Siamese, and American Shorthair lines, as follows: Many owners remark on the almost doglike tendencies of the Breed, in that they are devoted to people, are easily trained and respond well to the voice, but retain their independence as a proper cat should, and are very intelligent. Because of their adaptability they are a joy to be with, they are not demanding in any way and seem to take life in their stride. Ocicats are reasonably vocal and do not like being left alone for long periods, but do make ideal companions for households with other pets, and are confident with children.2 Although such formal recognition permeates the world of cat enthusiasts, scientists have devoted little attention to investigating whether cat breeds have distinctive personalities. The line-breeding necessary to develop cats that breed "true"-that is, where the offspring look the same as their parents-has led to some behavioral abnormalities that have a genetic basis (see box below, "Fabric Eating in Pedigree Oriental Cats"). Because these are essentially pathologies, and isolated within a single breed or group of breeds, scientists do not cla.s.sify these abnormalities as aspects of "personality." Turning to more universal cat behavior, Siamese and other Oriental cats are remarkably vocal; many develop so many variations on the meow that they seem to "talk" to their owners. Longhaired cats, especially Persians, have a reputation for being lethargic and not terribly fond of close contact with people, perhaps because these cats overheat easily. Beyond such self-evident differences, we have little hard information on precisely how breeds differ in personality, and how those differences arise. Most of the information we have is based on surveys of experts-veterinarians or cat-show judges, for example-who tend to see most cats when they are away from their normal territories, and therefore may not always get a complete picture of their behavior.

Fabric Eating in Pedigree Oriental Cats Siamese, Burmese, and other Oriental cat breeds are susceptible to developing an unusual form of pica, the eating of non-nutritive substances. For reasons we do not yet understand, some house cats develop the habit of chewing unusual items, such as elastic bands and rubber gloves, but a significant proportion of pedigree Oriental cats not only chew but also eat fabrics. Their fabric of choice is usually wool, closely followed by cotton; synthetic fabrics such as nylon and polyester are less popular. Most of these cats start by chewing woolen items: many then progress to swallowing the chewed-off fabric chunks, or move on to other materials. In these cases, the cats appear to confuse fabrics with food. I have seen a Siamese cat dragging an old sock up to its food bowl, and then alternately taking one mouthful of one, and one mouthful of the other.

A Siamese cat eating a piece of cloth We also have yet to understand their predilection for wool over other fabrics. One theory held that these cats might have a craving for the natural lanolin in wool, but when I tested this directly, this idea did not stand up.

Because wool eating is largely restricted to a small number of closely related breeds, it must have a genetic basis. However, it does not seem to be inherited directly. When I surveyed the owners of seventy-five kittens produced by seven mothers, three of whom were fabric eaters and four of which were not, one-third of the kittens had become fabric eaters themselves-but many of these had "normal" mothers (their fathers' habits were unknown). Neither simple genetic factors, nor imitation of the mother's behavior, could explain why some had developed this problem while others had not.

Many of the fabric-eating cats did also show other types of abnormal behavior, such as biting their owners and excessive scratching. These also occur in non-pedigree cats and are often a sign of anxiety and stress. Among Oriental cats, fabric eating often starts within a few weeks of the cat being rehomed, when the cat may be feeling stressed by the change in its environment. Onset can also occur at around one year of age even without a move, when the cat is becoming s.e.xually mature and starting to come into conflict with other cats either within the household or outside (even though they were valuable pedigree animals, few of the cats in my study were totally confined indoors).

Fabric eating may therefore start as a soothing oral behavior that these cats adopt when they feel especially stressed, rather like thumb sucking in human infants. Why they choose fabrics, and why chewing often turns into ingestion, is still unclear.

One small-scale study conducted in Norway confirmed that Siamese and Persian cats do indeed behave in characteristic ways in their owners' homes.3 Although the cats' personalities were recorded by the owners themselves (which in itself could have introduced some biases) rather than through direct observation, the Siamese were reported to be more contact-seeking, more vocal, and more playful and active than standard house cats. One in ten Siamese was regularly aggressive toward people, compared to one in twenty house cats and one in sixty Persians. Persians were generally less active than other cats, and apparently more tolerant of unfamiliar people and cats-although their apparent laziness might have simply made them disinclined to run away.

It's highly implausible that every single variation in cat personality will be traced to a different gene. Rather, breed characteristics seem to emerge during kittenhood as general tendencies, such as when making the choice whether to explore or to move away from novel objects or situations. In turn, these tendencies profoundly affect what each kitten learns, and thus how its behavior develops: tactics it learns are useful in one particular circ.u.mstance may become general strategies, used in many situations. The underlying processes have scarcely been investigated in pedigree cats, but in one study, researchers found that the ability of Norwegian Forest kittens to remember novel situations develops more slowly than those of other pedigree cats (Oriental breeds and Abyssinians), whose brains may develop somewhat more quickly than those of ordinary house cats.4 Such slowing-down and speeding-up of the rates at which different parts of the brain grow might have long-term effects on cat personality. Many of the self-evident differences in behavior between dog breeds result from changes in the speed at which different areas of the brain develop: for example, Siberian Huskies display a full range of wolf-type behavior, while breeds with "baby faces," such as bulldogs, signal to one another in a similar way to wolf cubs of just a few weeks old.5 However, scientists have not yet doc.u.mented any such link for cats.

Differences among breeds provide useful insight into whether cat behavior might be influenced by genetics, and pedigree breeds are useful in this regard because each cat's parentage is doc.u.mented. Popular males can sire many kittens, yet rarely even see any of them, so their influence on their offspring must be genetic. In the Norwegian study, playfulness, fearfulness, and confidence in encounters with unfamiliar people were all distinctively different between the offspring of different fathers, although some other traits, such as aggression to cats or people, were not. Because this study was small-scale and carried out in only one country, its details may not apply everywhere; still, the principle that some aspects of a cat's behavior are influenced by its father's genes seems likely to stand.6 Non-pedigree cats also vary greatly in their "personalities," spurring on the myth that a cat's temperament and its coat color are inextricably linked.7 The British refer to tortoisesh.e.l.l cats as "naughty torties"; likewise, blotched tabbies are "real homebodies," mackerel tabbies are "independent," and white patches on a cat's coat have a "calming effect" on the animal. It seems part of human nature to link outward appearance and inner character, and to continue to see those links even when evidence is to the contrary. Some scientists have speculated that the specific biochemistry that generates different coat colors also somehow affects the way a cat's brain works, showing a genetic effect referred to as pleitropy, but little evidence has been found to support this idea in cats.8 Links between coat color and personality do occasionally occur among pedigree cats, and these do provide an opportunity for proper investigation because the family trees are available. The relatively restricted gene pool for each color within each breed does result in certain temperaments accidentally becoming a.s.sociated with particular coat colors. At any one time, only a limited number of high-quality tomcats within each breed are available to produce the desired color; as a result, the temperament of the most popular of those tomcats-or at least those aspects affected by genetics-tends to become predominant within that section of the breed. For example, twenty years ago, Scotland's British Shorthair cats with tortoisesh.e.l.l, cream, and especially red (a rare, un-patterned version of orange) coats were relatively difficult to handle; scientists traced this characteristic back to one male with a particularly difficult temperament.9 Likewise, cats with dark "points" on their paws and ears, even if not pedigree Siamese, are likely unusually vocal, because the gene that causes the points to appear is very rare in any cat without at least one Siamese in its recent ancestry.

Coat color and some aspect of personality can also become linked if the gene that controls the color and a gene affecting the way the brain develops happen to occur very close together on the same chromosome. Because genes are grouped together on chromosomes-cats have thirty-eight: eighteen pairs, plus two s.e.x chromosomes-not all combinations are pa.s.sed on randomly from one generation to the next. If two genes occur on different chromosomes, then the chances that a kitten will receive any particular combination of the two are essentially random. However, two genes that occur on the same chromosome tend to be inherited together. This is not inevitable, because matching pairs of chromosomes do occasionally swap sections between each other, by a mechanism known as crossing over; if the swap happens in between the two genes in question, they are then inherited separately. Such exchanges rarely occur between genes that are sited close together on the same chromosome. For example, the gene that causes a white coat ("dominant white," that is; different from albino) is situated on the same chromosome and close to another gene that causes both the eyes to be blue and the cat to be deaf, a rare example of one gene affecting both appearance and (indirectly) behavior. Blue-eyed, white cats are thus almost invariably deaf.10 In the case of the ginger cats in rural France, the gene that suits those cats to the feral lifestyle might simply be very close to the O(range)-gene (on the X-chromosome), rather than being a direct effect of the cat being orange.

Making a.s.sumptions about a cat's personality based solely on its appearance is often misleading, but cats undoubtedly do behave in individual ways, irrespective of their color. Until about twenty years ago, most scientists considered that only humans could have "personalities," yet now this concept is widely applied to animals-and not just to domestic animals. Even wild animals behave in consistently different ways that reflect different ways of reacting to the world around them: over the past few years, the concept of "personality" has been applied to animals as diverse as lizards, crickets, bees, chimpanzees, and geese. Some individuals may be particularly bold, and therefore the first to exploit a new food source, whereas others are particularly shy and therefore less likely to run headlong into dangerous situations. The success of each strategy is likely to vary depending on what the environment is like, and as that changes, so sometimes bold individuals will do best, other times they will be the ones who perish first. In this way, the genes that influence both types persist in the species.

Some of the most complex effects of personality occur in social situations. Sticklebacks, fish that sometimes swim in shoals, can be cla.s.sed as either bold or shy. When a fish has a choice of shoals consisting entirely of bold or shy individuals, it will choose to join the bold shoal, irrespective of whether it is bold or shy itself. Bold shoals usually find more food, and a shy fish will find the middle of a shoal of bold fish a good place to hide. However, to keep up with the bold shoal, it must swim faster than usual, so it temporarily starts behaving more like a bold fish. Although we don't yet know much about social effects on cat personality, such observations raise the fascinating possibility that each cat may be able to adjust its personality to fit in with those of the other animals-human, feline, and canine-in the household in which it finds itself.

We have two broad approaches to studying cat personality: watching the cats, or asking their owners. Because owners are likely to be biased, observing the cat's behavior is the only way an impression of its personality can be gained. For this reason, most of my own studies have involved recording cats' behavior. To ensure that outdoor cats would be home, I chose to observe them just before and just after their usual feeding time.11 Since many cats interact most intensely with their owners when expecting food, and since hunger usually affects the way they interact, this arrangement had the additional advantage that all the cats would have been hungry when the observation started, and sated at the end.

While their food was being prepared, the thirty-six cats in the study acted as cats usually do when expecting to be fed: walking around the kitchen with their tails upright, meowing, and rubbing on their owners' legs. After the meal, some went straight outdoors, while others sat and groomed themselves; some interacted with their owners again, while others investigated the unfamiliar human in the room-that is, the person making the observations. So far, so obvious; but the first objective of our study was to find out whether each cat behaved in a characteristic way every time. We repeated these visits once a week for eight weeks, and found that indeed they were indeed fairly consistent-so what we had measured was probably a reflection of the cat's personality, or at least its "personal style."

A bold cat studies a scientist From their behavior before they were fed, we separated the cats into various types. Some always rubbed around their owner's legs, purring all the while; others never did. Some walked around the kitchen much more than others, and some continually tried to attract their owner's attention by meowing; their owners didn't seem to find this particularly endearing-they stroked the quiet ones more often. Only about half of the cats took any of these tendencies to extremes; the others split their time between rubbing and meowing, and were moderately active-unsurprising given that all these traits are typical of cats.

After they were fed, some of the younger cats went straight outdoors-possibly more of a habit than a personality trait. Most, however, stayed in the kitchen for a few minutes. Several of the cats that had been more active than the others before the meal continued to interact vigorously with their owners, walking around with their tails up and meowing. Those that had paid most attention to the unfamiliar person before the meal continued to do so.

We rounded out our observations by talking to the cats' owners. Some of the differences between cats we observed seemed to reflect their personalities, but we did watch them only in one (convenient) situation. Would we have seen other sides to each cat's character if we'd also watched it when it was out exploring, socializing with (or avoiding) other cats, or curled up while its owner watched television? Several studies have examined differences in cats' reactions to people by asking owners or other caregivers to report on their cats' behavior. Inevitably, these surveys provide little information on how the cat behaves when it's alone, and owners with only one cat won't necessarily know how well it gets along with other cats.

Three aspects or dimensions of cat personality have emerged from such studies, despite the studies' limitations.12 The first is whether or not the cat gets along well with other cats in the same household or other social group; some cats seem to be more outgoing toward other cats than others-at least, ones that they know well. Second is how sociable the cat is toward people in the household; some cats seem to value close contact with their owners more than others. Third, and possibly most fundamental, is how bold and active or steady and cautious the cat is in general. An individual cat may possess any one of the eight possible combinations of these three basic traits: for example, one cat could be shy and retiring, but affectionate toward her owner and the other cats in the house; another might be highly active and equally affectionate toward his owner, but might keep his distance from another cat in his household. While these traits are defined by their extremes, in real life most cats are intermediate on one, two, or even all three. Insofar as their owners are concerned, there may be no such thing as an "average cat"-but many do in fact approximate to that description.

The bold/shy dimension is perhaps the most important of all, because it affects not only how the cat behaves on a minute-by-minute basis, but also how much and what each cat learns. In some situations, a bold cat learns more from a new experience than a cat that holds back. However, if the bold cat behaves overconfidently and is hurt as a consequence-for example, if it struts up to a belligerent tomcat-then not only may it get injured, but it may also learn less from its experience than a more circ.u.mspect cat that simply stands by and watches the encounter.

Whether a cat gets on with other cats, or is especially affectionate toward people, seems strongly influenced by its experiences as a kitten and during its adolescence. At least, we have yet to detect a strong and lasting genetic influence among crossbred cats. Scientists once thought that some cats carried "friendly" genes and others "unfriendly," but when they investigated this in more depth, they found that the differences were in fact due to genes that affected how bold the cats were. Bold and shy cats learn differently how to interact with cats and people. This is not to say that bold cats are necessarily friendlier than shy cats, or vice versa, although they are likely to express their affection in slightly different ways.

The notion that bold and shy cats learn in subtly different ways emerged from a cla.s.sic experiment in which the offspring of two tomcats, one with a reputation for producing "friendly" kittens and the other "unfriendly," were raised in groups that were socialized in slightly different ways, some given minimal handling, while others were handled daily.13 At one year old, the offspring were compared by placing them in an arena with an unfamiliar object: a cardboard box they'd never seen before. The offspring of the "friendly" father explored the box most rapidly and most thoroughly, and the offspring of the "unfriendly" father tended to hold back. The genetic difference between the two fathers thus influenced something more fundamental than just how friendly their kittens had become; it affected how their kittens reacted to anything they hadn't encountered before.

Similarly, how the kittens interacted with people during their socialization period was affected by their boldness. The kittens with the bold father approached people spontaneously, and as a result learned quickly how to interact with them. The shy kittens took longer to achieve the same level of confidence with people. However, given enough handling, these shy kittens could turn out just as friendly as the kittens with the bold father-though as expected, they tended to show their friendliness in a less "pushy" way than the kittens with the bold father. The shy father's kittens became fearful of people only if they were not handled every day: even at a year old these kittens, like their father, moved away from people, hissing and flattening their bodies to the ground. The amount of exposure to people they had received was, for the most timid, significantly less than the average kitten born in a typical home experiences, so the study did not precisely replicate normal conditions. However, it does provides a valuable insight into how vulnerable the offspring of shy tomcats may be to interruptions in their socialization.

Kittens born in someone's home likely receive enough handling to end up at least reasonably friendly to people, regardless of whether they are genetically bold, shy, or somewhere in between. The ways they show their affection may differ, however, and this seems to interact in a complex way with their early experience. In 2002, my team researched this interaction in a study on twenty-nine cats from nine litters born in regular homes.14 The amount of handling the litters received in their second month varied from twenty minutes to more than two hours per day. When these kittens were eight weeks old, just before they were homed, we tried picking them up, one at a time. Those that had received the least handling were definitely the most inclined to jump down; we could hold those that had received the most handling for several minutes at a time. The amount of handling they had received seemed to have had more of an effect on their behavior than any genetic effects; all the litters had different mothers, and although we had no idea who their fathers were, the homes they'd been born in were far enough apart to make it unlikely that any two had the same father.

When we repeated our test two months later, when the great majority of the kittens had been moved to new homes, we found exactly the opposite: the kittens that had received the most handling during their second month of life were now the most restless, and those that had received the least were now the quietest.

This apparent contradiction probably demonstrates that not only does socialization to people in general start during the second month, so too does attachment to specific people. Very few of the cats appeared distressed when they were picked up at eight weeks old, so all must have received more than enough socialization to create a generally friendly kitten. Nevertheless, those handled the least were still not entirely confident when picked up by a stranger. Two months later, these kittens had gone through the process of learning about a new set of people-their new owners-and showed through their behavior that they were perfectly content to be picked up by anyone. The kittens that had received a great deal of handling in their original home may consequently have become extremely attached to their original owners, and therefore found the transfer to their new home particularly unsettling. Despite the two months of acclimatization that had pa.s.sed, when we returned to test them again they may still have been anxious in these new homes.

The handling they'd received seemed to set these kittens' personalities off in different directions, but all effects of the differences in handling during kittenhood gradually disappeared as these cats matured. At one year of age, they varied in how much they liked being picked up by a person they hadn't seen for eight months, but this bore no relation to how much handling they'd received as kittens, or indeed to their genetics; cats that had originally been littermates were no longer similar to one another. When we tested them again at two and three years old, we found that they had changed little from the way they were at twelve months; so, by the end of their first year, each cat had already developed its own distinctive way of reacting to people.

Somehow, the way cats react to being picked up by unfamiliar people changes as they grow through their first year. This change is probably influenced by their new owner's lifestyle and how it interacts with their developing personality. However, once they are about one year old, their reactions vary much less. For example, some cats presumably become accustomed to being in a busy house and are unphased by strangers; others prefer the company of their owners, and hide when visitors stop by. The way cats arrive at these states of equanimity is affected by the amount of handling they receive before they leave their mothers, and almost certainly also by their genetics, but the end result appears to be roughly the same whatever the route they have traveled.

Most cats are extraordinarily sensitive to human body language, much more so than they usually receive credit for. This sensitivity enables them to adapt their behavior to the people they meet. People who dislike cats often complain that they are the first person in the room a cat makes a beeline for, so I decided to test this theory by staging encounters between cats and people who either liked cats or found them repulsive.15 The people-all men, since we could not find any women who admitted to hating cats-were instructed to sit on a couch and not to move when a cat came into the room, even if it tried to sit on their laps. However, we could not prevent the cat-haters from looking away from the cat, which they usually did within ten seconds of first seeing it. The cats, for their part, seemed to sense the disposition of the people they were meeting within a few seconds of entering the room. They rarely approached the cat-phobics, preferring to sit near the door and look away from them. It was unclear how the cats were detecting the difference between the two types of men: perhaps they could sense that the cat-phobics were tenser, or smelled different, or glanced nervously at the cats. Nevertheless, the cats' reactions show that they can be keenly perceptive when encountering someone for the first time. However, one of our eight cats, while apparently equally perceptive, behaved contrary to the other seven, singling out the cat-phobics for the most attention, jumping on their laps and purring loudly, much to their disgust. Cats such as this one presumably make a lasting impression on cat-phobics, rather than the majority of cats that sensibly avoid them.

Young cats especially seem to be much more adaptable animals than their detractors-and even some of their supporters-might have us believe. During the first year of their lives, they effectively tailor their personalities as much as they can to suit the particular household, or indeed other environment, in which they find themselves. Investigations on how they do this are still in their infancy, and the process is likely to be both drawn-out and contain features that are intrinsically private; still, intriguing links are emerging between cat behavior and their owners' personality. For example, owners who have intensely emotional relationships with their cats tend to have cats that are very happy to be picked up and cuddled.16 This could be the cat simply adapting to the owner's demands, although studies have suggested that cats tend to resist being picked up when they're not ready for it. It's possible, therefore, that people who obtain an adult cat and want one that loves to be picked up deliberately select one with that kind of personality, rather than the cat changing its ways to suit its owner's demands. However, young cats and kittens are almost certainly more adaptable than older cats are.

A typical cat's relationship with its owner is apparently not rigidly determined by whether the cat is genetically bold or shy, even though both of these personality types persist in the general cat population. More important than its personality is the amount of handling the cat receives when it is a kitten, which alters the way it behaves for the first few months of its life; after this time, most cats successfully adapt their behavior to the demands of their new owners. Some kittens do not receive much handling, perhaps because they have been born in a cat shelter where the staff is overstretched, or because their mother, being shy herself, has hidden them away. If such kittens also have an inherited genetic tendency toward shyness, inherited from their mother or indeed their father, they may be at risk of never developing a fully affectionate relationship with their new owners. In all the studies that trace the development of kittens' personalities, some cats have disappeared from their owner's homes. While we have never been able to explain this fully, we have suspected that some were somewhat unsuited to being pets and had chosen to become feral.

Continued gentle interaction with people can buffer the effects of genetic shyness, so that the young cat learns to come out of its sh.e.l.l. If a cat that is genetically disposed toward shyness does not receive this handling, then the innate shyness may persist into adulthood. These cats, if allowed to breed, will still carry the "shy" genes they inherited from their parents; so if they mate with another "shy" individual, these genes will persist throughout the cat population.

We know little about how a cat's genetics affects its sociability toward other cats. The little existing research has focused instead on how a kitten's early experience of other cats alters its behavior as an adult. Kittens hand-raised on their own behave abnormally toward other cats, and kittens hand-raised alongside a littermate less so-nevertheless, all display a bizarre combination of fascination and fear on encountering another cat.17 The presence of the mother, or in her absence another friendly adult cat, is apparently necessary for kittens to develop normal social behavior.

Cats raised in the normal way by their mothers also differ from one another in how friendly they are to other cats, albeit not exhibiting the extremes found in hand-raised cats. Much of this variation may also stem from different experiences during kittenhood: kittens born into extended families normally find it easier to learn social skills than those born to solitary mothers. For example, in parts of New Zealand, some feral cats live around farmyards, subsisting on a diet of rodents and sc.r.a.ps provided by the farmer, much like farm cats in the UK or the United States. However, because of less compet.i.tion from native predators than elsewhere, other cats live in the bush nearby, feeding exclusively on prey they have hunted for themselves and adopting a way of life that must be much like that of undomesticated Felis lybica.18 Males roam between the two populations, keeping them genetically mixed, but females seem to adopt their mother's lifestyle: those born on the farms stay there and share their mothers' territories, while those in the bush strike out on their own. Cats thus appear to inherit a form of social "culture" from their mothers that may have little to do with their genetics.

Even cats that have lived their whole lives in groups can differ markedly in how they react to other cats, hinting at genetic as well as cultural influences. In one study of two small indoor colonies (seven females each), researchers found that the individual cats varied consistently in their calmness while interacting with the others, and varied also in the extent to which they chose to be close to or keep away from them. These aspects of their personalities were distinct from how sociable these cats were toward people, or how generally active and inquisitive they were.19 Each cat appeared to have worked out its own way of interacting with the cats around it, which was not simply a reflection of how "bold" it was.

Each cat's boldness would had affected the way it had initially approached other cats when it was first introduced to them, since "boldness" is a trait that underlies all first encounters with novel situations. However, in repeated encounters with the same individuals, each had developed a new and eventually stable personality trait, sociability to cats, which was unrelated to how bold they were. We can only guess at the reasons behind these emerging differences, since we know little of how such traits are formed. For example, if the outcome of just the first few encounters fixes each cat's strategy for dealing with cats permanently, then its personality might be profoundly affected by whether it happened to be smaller and weaker, or bigger and stronger, than the cats it met early on. However, if this aspect of personality develops over many encounters, then the subsequent differences between each cat could conceivably be driven by genetic factors, distinct from those that affect whether a cat is bold or shy.

Indeed, the domestic cat's capacity to live amicably with other cats should be genetically variable. Its wildcat ancestors seem incapable of extending their affiliations beyond temporary bonds between mother and offspring, but many domestic cats seem capable of forming affectionate bonds with adult cats. It seems highly unlikely that this change has already evolved as far as it can. Some cats, at least anecdotally, are unusually outgoing toward other cats, while others prefer to be virtually solitary. Dissimilarities in early life experience certainly caused some of these differences, but surely genetics also plays a part. We should still see variation between today's cats in how easy they find it to forge bonds with others. How shy or bold they are may affect this variation, although as with affection for people, boldness or shyness may have more influence on how the cat approaches another cat than on whether or not they become friends or enemies.

On the face of it, we would not expect much genetic variation underlying the domestic cat's ability to hunt. Not only is the cat descended from a specialist predator, but the primary function of the cat in human society has been, until recently, to kill rodent pests. Moreover, until the appearance of nutritionally balanced cat food some forty years ago, cats that were incompetent predators would not have bred as successfully as those that were. Scientific studies have confirmed that, given enough exposure to real prey, all cats have the potential to become competent predators by the time they are six months old. The way each cat hunts-whether it moves around constantly or sits and waits for hours where it knows that prey will appear-varies a great deal from individual to individual. So too does the type of prey in which each cat specializes: some cats seem incapable of catching a bird, and others catch more birds than rodents. We could regard these variations as aspects of "personality." However, these differences probably stem from each cat's experiences as it refines its hunting skills, since it will likely repeat tactics that have led to a meal. We have no evidence that some cats are born bird-hunters, and others innate mouse-catchers.

Surprisingly, research has revealed considerable differences among kittens in their prey-catching competence, particularly during their third month of life. By this stage in their development, all kittens are capable of performing their basic repertoire of predatory behavior-stalking, pouncing, biting, and raking with the claws-which they have practiced for the past several weeks through play with inanimate objects around the nest, as well as in mock-predation on their littermates. Despite all this practice, kittens between two and three months old vary greatly in effectively putting these actions together, in a.s.sessing what they can likely catch and what they should avoid, and in selecting the appropriate tactics for the prey in question-for example, not chasing after birds that are already flying away.20 Three months later, however, they are all equally competent; somehow, the laggards catch up. Scientists have found no developmental reasons for the differences between kittens during their third month, so it is possible that these variations at least are genetically influenced. Genetics undoubtedly affects the rate at which kittens develop in general-for example, the age at which their eyes open-so this hypothesis is reasonable.

As every cat owner knows, cats differ not just on the outside. Inner and outer qualities are both affected by genes and the environment in which the cat grows up, but to different extents and in different ways. Cats' personalities develop according to a highly complex interplay between genetics and what the cat experiences during the first year or so of its life. These experiences can have extremely powerful effects that can all but obliterate any trace of genetics. Yet among cats that receive what is, for their species, a "normal" upbringing within a human household, signs of genetic effects on personality are apparent.

Cats also vary in terms of how quickly they learn to hunt. Although experience again plays a large part in determining this aspect of a cat's personality, it also seems likely that genetic factors are also at work. Cats are not only able to adapt as individuals to the circ.u.mstances in which they find themselves; they are also members of a species that contains a significant amount of genetic variation that affects their behavior, giving them the potential to evolve further as our demands upon them change. Today, the most significant challenge facing cats is a growing reputation as destroyers of wildlife, but even their most vociferous critics have to admit that not all cats are to blame. If hunting ability is linked to personality, and personality has a genetic basis, then it may become possible to predict which cats are likely to cause the least offense.

CHAPTER 10.

Cats and Wildlife Few topics get wildlife enthusiasts as riled as predation by domestic cats. In the wrong environment, cats can undoubtedly cause substantial damage to other species, especially where there is little compet.i.tion in the shape of wild predators. Although many pet cats do undeniably go out hunting, it has proved remarkably difficult to pin down the impact of that hunting-whether it does have a significant effect on wildlife numbers. Indeed, when the balance of wildlife changes in a particular place, cats become convenient scapegoats. Well-fed pet cats should not need to hunt to supplement their diet, and in that sense any damage they do is unnecessary. Moreover, their habit of bringing their kill home rather than eating it on the spot does make it easy for their detractors to point the finger at them-it can seem as though they are killing for "sport"-while deaths due to wilder predators go unnoticed.

More insidiously, general anti-cat sentiments can creep into scientific literature addressing wildlife conservation. A group of scientists in Australia has recently called for "restrictions on the maximum number of cats allowed per household, mandatory sterilisation and registration of pet cats, curfews, requiring pet cats roaming outdoors to wear collar-mounted predation-deterrents or compulsory confinement of cats to their owners' premises," even though it is far from clear that any of these restrictions would lead to a recovery in the local wildlife.1 In 1997, the UK's Mammal Society produced an estimate of 275 million animals killed in Britain each year by pet cats. These figures derived from forms completed by their youth wing, "Mammalaction": data from the 696 cats surveyed were extrapolated up to the 9 million cats then in the UK. However, when the full a.n.a.lysis was finally published in 2003, it became clear that very few cats-less than 9 percent-had been included that did not hunt at all, even though most other studies had concluded that only about half of pet cats ever bring any prey home; fewer in urban areas, slightly more in the countryside.2, 3 The reason for this bias appeared to be in the design of the questionnaire, which encouraged Mammalaction members to submit their results only if their cat had brought in some prey during the five months of the survey.

Despite these shortcomings, the figure of 275 million is still widely quoted by many influential organizations, including the Royal Society for the Protection of Birds, the British Trust for Ornithology, and the Bat Conservation Trust. When the figures were first announced in 2001, British wildlife TV presenter Chris Packham, a self-confessed "cat-hater," appeared on BBC radio describing cats as "sly, greedy, insidious murderers," and calling for them to be "shot"; more recently, he a.s.serted that all cats should be given "ASBOs" (Anti-Social Behavior Orders, court-imposed restrictions on people likely to "cause hara.s.sment, alarm or distress to an individual householder or a neighborhood").4 Similarly, when the University of Georgia's Kitty Cams revealed that a small minority of cats in Athens, Georgia, were killing a couple of lizards each week, the Los Angeles Times's Paul Whitfield wrote, "So they're slaughtering wildlife, and you can't trust them. Seems like grounds for government action. . . . Present owners can keep their cats. But as the tabbies die off, so does cat ownership."5 A January 2013 New York Times report of an estimate of predation by cats across the United States made by scientists from the Smithsonian Inst.i.tute generated more email responses than any other story that day. The headline of another report on the same study read, "Domestic Cats Are Destroying the Planet."6 Looking at this situation more objectively, the impact cats actually have on wildlife varies enormously from one type of environment to another. The most dramatic effects undoubtedly occur on small, oceanic islands onto which cats have been introduced. Many of these islands contain unique fauna that have evolved due to their isolation from the mainland. Others are refuges for seabirds that raise their young in safety there, undisturbed by predators. When cats appear, they can cause havoc. Occasionally, these cats have been pets, perhaps the most notorious being the lighthouse-keeper's cat that killed the last specimen of the Stephens Island wren in 1894.7 Usually, however, the cats have either escaped from visiting ships, or have been deliberately introduced to suppress pests such as rabbits, rats, and mice-usually accidental introductions in themselves. Lacking compet.i.tion from other mammalian predators, cats can thrive in such environments, making easy prey of the local wildlife that has previously been unexposed to such hunters. With such an abundance of food available, cats breed prolifically, producing large populations of ferals. Perhaps counterintuitively, these feral cats sometimes do most harm on islands where they are not the only introduced animal: some researchers have suggested that an abundance of house mice, for example, provides the cats with a stable diet, thereby enabling them to increase in numbers to the point where they can exterminate the more vulnerable local wildlife.

We must keep the effects of feral cats in perspective, even when considering the island-cat situation. Island species account for 83 percent of all doc.u.mented extinctions of mammals: isolated from many of the diseases, parasites, and predators that plague their relatives on the mainland, these species are intrinsically vulnerable. Yet scientists have been able to implicate feral cats in only about 15 percent of such extinctions, and even in these, other introduced predators have to take their share of the responsibility. Foxes, cane toads, mongooses, and especially rats can be equally, if not more devastating. Black ("ship") rats probably do more damage than any other introduced predator, and because cats are effective hunters of this species, their presence can sometimes even be beneficial.

For example, on Stewart Island off the coast of New Zealand, feral cats have existed for more than 200 years side by side with an endangered flightless parrot, the kakapo. These cats feed mainly on introduced species of rats (black and brown), which have been held responsible for the extinction of several other species of birds in the same region. Removal of the cats here might lead to an increase in the rat population, and could therefore potentially lead to the extinction of the kakapo.8 However, we cannot deny that eradicating cats from islands has in some cases led to dramatic recoveries in the populations of threatened vertebrate species: examples include iguanas on Long Cay in the West Indies, deer mice on Coronados Island in the Gulf of California, and a rare bird, the saddleback, on Little Barrier Island in New Zealand.

A feral cat stalking a kakapo On the mainland, feral cats can undoubtedly be effective predators in some locations, but their impact is much more difficult to quantify. Nevertheless, the sheer numbers of stray and feral cats in the world, somewhere between 25 and 80 million in the United States and around 12 million in Australia, suggests that they must have a major impact.9 Feral domestic cats are "alien" predators throughout much of their range: indeed, they have been in the United States for less than 500 years. In most places where they have been introduced, cats seem able to compete quite effectively with local, "native" predators, even though the latter should be better adapted to local conditions.

Feral cats do have three advantages over other predators. First, their numbers are constantly added to by cats straying from the pet population, or emigrating from farms where cats are still kept for pest control. Second, because they are generally less fearful of humans than many wild carnivores, they can better take advantage of food accidentally provided by people, such as at garbage dumps, to sustain themselves when prey is hard to find. Third, because they resemble pet cats in everything but behavior, they attract the sympathy of many people, some of whom devote their lives to providing them with food and even veterinary care.

The most scientific attention to this problem, and perhaps the most public outcry, has occurred in Australia and New Zealand, where cats seem to be relatively recent introductions.10 In both countries, many small marsupials and flightless birds have undoubtedly gone extinct, but the main culprit may be loss of habitat, not predation. Even where predators have been a major factor, the responsibility is often shared among cats, rats, introduced red foxes, and (in Australia) dingoes. According to Christopher R. d.i.c.kman of the Inst.i.tute of Wildlife Research in Sydney, "the effects of cats on prey communities remain speculative."11 In some situations, cats can be a major cause of decline; in others, they may be protective. Predation from cats appears to have made a major contribution to the decline of some threatened native Australian species, such as the eastern barred bandicoot in Victoria and the rufous hare-wallaby in the Northern Territory. On the other hand, in a study of remnant patches of forest in suburban Sydney, the presence of cats protected tree-nesting birds, apparently because they themselves were hunting rats and other animals that would normally have raided the nests.12 Cats can also suppress the numbers of introduced mammals, such as mice and rabbits, that compete for food with the native wildlife.

Despite the equivocal evidence, several Australian munic.i.p.alities have pressed ahead with measures to reduce the impact of cats on wildlife. These include confinement of cats to owners' premises at all times, prohibiting cat ownership in new suburbs, nighttime curfews, and impounding free-roaming cats in declared conservation areas-even though only the last of these would control the activities of the feral cats that are probably causing the most damage.

Researchers have yet to evaluate the effectiveness of such control measures comprehensively. However, a recent survey of four areas of the City of Armadale, Western Australia, suggests that cats may not be the primary culprits after all. One area in this study was a no-cat zone, where cat ownership was strictly prohibited; the second was a curfew zone, in which pet cats had to be belled during the day and kept indoors at night; and in the other two areas, cats went unrestricted. The main prey species in the area were brushtail possums, Southern brown bandicoots, and the mardo, a small predatory marsupial a little bigger than a mouse, which was predicted to be the most vulnerable of the three to cat predation. In fact, researchers found more mardos in the unregulated areas than in the curfew or no-cat areas, and saw little difference in the numbers of the other two prey species across any of the sites. What variation there was could be best accounted for by the amount of vegetation available: in other words, habitat degradation, and not cats, may have been the major factor limiting the numbers of small marsupials. The draconian control measures against pet cats had, at least in this one location, produced no benefit to wildlife.13 How much long-term damage to wildlife do pet cats cause? Estimates of what proportion of pet cats ever kill anything vary considerably, but figures of between 30 percent and 60 percent seem reasonable, even when those cats kept indoors without access to prey are excluded. We have little reliable information on how many animals those cats that do hunt actually catch, because such events are rarely observed. What is usually recorded is not how many animals are caught, but how many are brought back dead to their owners-and then a "correction factor" is used to calculate the number killed, to account for those prey items that are eaten where they were killed, or simply discarded. The number brought home per cat is often quite low: for example, 4.4 animals per cat, per year in a recent UK study.14 The proportion of prey brought home has been calculated only twice, coming out at around 30 percent (although one of the two studies examined only eleven cats). Recently, a new study in the United States has provided a much more detailed picture, both metaphorically and literally, as the cats were fitted with Kitty Cams, lightweight video camcorders that provided a view of everywhere they went for a week or more. These cats took home around a quarter of their prey, they ate another quarter, and left the remaining half uneaten at the capture site. What may make this study somewhat atypical is that the main prey taken was a lizard, the Carolina anole, which many cats find unappetizing. In places where the main prey consists of mammals-such as the more palatable woodmouse commonly taken in the UK-both the proportion brought home and the proportion eaten might be higher.

Once we take these "correction factors" into account, and the figures scaled up to the whole of an area's cat population, the total number of prey taken can at first sight seem alarming. The Mammal Society's figures of 275 million per year might be overstating the case, but between 100 and 150 million may be a reasonable estimate for the entire UK. The recent Smithsonian study produced an estimate of between 430 million and 1.1 billion birds killed annually by pet cats in the mainland United States.15 Furthermore, individual anecdotes, taken in isolation, do at first glance suggest that cats might be capable of bringing about local exterminations. For example, biologist Rebecca Hughes from the UK's Reading University reported that, in the cold winter of 200910, "One cat that lived beside a woodland brought in a blue t.i.t [chickadee] every single day for a fortnight."16 Whether such levels of predation make any significant difference to wildlife populations in the long term is far less easy to a.s.sess. To take the blue t.i.t as an example, the UK has an estimated 3.5 million breeding pairs, each producing seven or eight young per year-about 25 million more young birds than would be required to keep the population constant. So, some 25 million blue t.i.ts must die during the course of most years. Some do not make it out of the nest and some are victims of predation, but many starve during cold winters because their metabolism runs so quickly that they are barely able to store enough food to keep themselves alive overnight. Some of the birds brought home by the cat referred to above may well have died of natural causes during the night, and were then retrieved by the cat in the morning. In fact, the numbers of blue t.i.ts in UK gardens have increased by a quarter over the past fifty years, so pet cats are unlikely to be having any major effect on their numbers year on year. Most of the cat's favorite prey species are equally profligate breeders-in built-up environments in the UK, these include woodmice (fifteen to twenty young per pair per year), brown rats (fifteen to twenty-five), and robins (ten to fifteen).

Rather than contributing to the decline in wildlife populations, cats may simply be killing (or collecting) animals that are in any case not destined to survive for much longer-those that are ill or malnourished. Such animals are by their very nature the easiest to catch. One study that examined the condition of birds brought home by cats tended to confirm this idea: these birds were generally underweight and in poor condition.17 Recent evidence shows that garden birds may be evolving strategies to cope with cats. In rural areas, the main predators of European songbirds are usually the sparrowhawk and the kestrel; in urban gardens, the main predator is often the cat. By comparison with their country cousins, common birds living in urban gardens such as sparrows, robins, chickadees and finches were found to wriggle less, were more likely to "play dead," were less aggressive, and made fewer shrieks and alarm calls, all by comparison with their country cousins. The longer an area had been urbanized, the greater the difference, which suggests that the urban birds had not simply learned about avoiding cats, but had actually evolved a new set of defense mechanisms over the hundred or so generations since large-scale urbanization began in the mid-nineteenth century.18 What seems to irritate wildlife enthusiasts the most is not that pet cats hunt at all, but that they should not have to hunt, since the majority are well fed by their owners. Cats are therefore often portrayed as committing "murder," as opposed to other predators, which kill legitimately, to survive. Because they are fed, pet cats can exist at a much higher density than they ever could if they had to catch every meal for themselves. Thus, even occasional hunting could have a substantial impact, simply because there are so many cats.

Cats have not yet lost their desire to hunt; too few generations separate them from the rodent controllers of the nineteenth and early twentieth centuries. They will go out hunting even if well fed-in the past, when they were mousers, a single catch did not provide enough calories to allow them to relax between meals-but hunger does affect how intensely they hunt. Feral cats, even those that obtain most of their food from sc.r.a.ps, spend on average twice as much time hunting as pet cats do. Mother cats with kittens to feed will hunt almost continuously if they are not themselves being fed by someone.

Feral cats foraging in garbage By contrast, pet cats rarely hunt "seriously," often watching potential prey without bothering to stalk it. A hungry cat will pounce several times until the prey either escapes or is caught; a well-fed pet will pounce halfheartedly and then give up, probably explaining why pet cats, when they do kill birds, usually succeed only when they target individuals already weakened by hunger or disease. Furthermore, pet cats rarely consume their prey, often bringing it home as if to consume it there, but then abandoning it.

The quality of a cat's diet also affects its desire to hunt. In one recent study conducted in Chile, cats fed on household sc.r.a.ps were four times more likely to kill and eat a mouse than cats fed on modern pet food. In another, cats eating low-quality cat food would break off to kill and chase a rat, but cats eating fresh salmon ignored the same opportunity to hunt.19 These and other similar observations suggest that cats fed on sc.r.a.ps or nutritionally unbalanced cat food are strongly motivated to hunt, in particular by an impulse that they must supplement their diet to maintain their health. The domestic cat, in common with all its wild felid cousins, has highly specialized nutritional requirements that can be met from only one of two sources: either modern, nutritionally balanced commercial cat food, or prey (see box on page 71, "Cats Are the True Carnivores"). Sc.r.a.ps and low-quality cat food tend to be high in carbohydrates. If eaten day after day, carbohydrates seem to give cats a craving for protein-rich food, which in their world means flesh. Commercial pet foods are of much better quality than they were half a century ago, so most modern pet cats have likely received a nutritionally balanced diet every day since they were weaned, and so are unlikely to become prolific hunters. Cats that have been neglected or strayed at some time in their lives may have been driven to hunt through nutritional necessity: once they have acquired the habit, it may be difficult to lose, so such cats may require extra precautions to prevent them from hunting unnecessarily (see box on page 251, "How Can We Prevent Cats from Hunting?") Of course, we all want to minimize the damage cats do to wildlife. The best approach to this problem will vary depending on what kind of cat we are dealing with, and particularly how closely the cat in question is a.s.sociated with people. Pets and ferals require different solutions. On oceanic islands, cats are invariably introduced "aliens," largely or completely unsocialized to people, and occupying a niche that would otherwise have remained vacant, since medium-sized land mammals cannot reach such places without man's help. Whether the cats were introduced deliberately, or are the descendants of escaped ships' cats or settlers' pets does not matter; either way, they are now essentially wild animals. Eradication of these cats, by humane means, is usually the only means whereby each island's unique ecosystem can be restored, although this should not be carried out in isolation: other "alien" species, such as rats and mice, may themselves decimate the local fauna once they are liberated from the pressure of being preyed on by cats. Perhaps surprisingly, given the publicity that the damage that cats can do receives, only about 100 such eradications have taken place so far, with thousands more islands still affected by the presence of feral cats.22 Ultimately, however, widespread humane eradication may be the only way to fully restore fragile island ecosystems.

How Can We Prevent Cats from Hunting?

Surveys show that the large majority of pet cats catch very few birds or mammals. If your cat is part of this majority, then you need not take any countermeasures, unless you happen to live