As we all know, clients almost always want to know why their dogs are having problems. Many hasten to blame themselves (often citing certain TV shows and their failure to be good pack leaders). Many have focused on a single event from which they are sure the problem stemmed. Many attribute the problems to presumed “abuse” before they got the dog; this happens mostly with dogs adopted after puppyhood so that there’s a gap in the story crying out for filling in. Some have a more realistic view: They grasp the impacts of lack of socialization, for example.
Some trainers insist that it doesn’t really matter if we know the source of a problem. They say that the behavior is what it is, and we must just train the dog in front of us. I believe there’s a lot of truth to this position.
Nonetheless, some clients will persist in wanting—in needing—to understand what happened. Understanding their dog’s story better can help them empathize appropriately and pick battles wisely. For consultants, apprehending what has likely gone wrong can help us to better choose interventions and predict the likelihood of success. Those in shelter and rescue work can make wiser decisions about adoptability and placement with some knowledge of factors that may seriously affect future behavior. Understanding the variety of ways in which genetics and environment influence behavior is useful.
Genetics and epigenetics
Behavior results from both genetics and environment. In some cases, specific behavioral traits have been mapped to specific gene loci. In other cases, we can observe clear evidence of behavioral traits in certain related lines, but have not identified any specific gene or genes responsible for the traits. The study of heritability provides evidence of the inheritance of various behavioral traits. One way of describing the impact of genes on behavior is to imagine that, given a particular genotype, each trait (e.g., boldness, shyness, curiosity, sociality, predatoriness, etc.) has a potential range of expression. A given animal may be more or less bold, depending on subsequent learning. But it may never be as bold as a littermate whose genes dictate a different, bolder range.
Most DNA is functionally identical from one dog to the next. The dog and wolf genomes show even more similarity than the human and chimpanzee genomes, which are over 99 percent identical. Between dog breeds, or individual dogs, that similarity is even greater. Although dogs show genetic plasticity unrivaled in the animal kingdom, the differences between one dog’s DNA and another dog’s DNA are very tiny. Yet these small differences can produce functionally significant behavioral differences. Dog breeds don’t differ just in appearance; they also differ in behavioral tendencies. We can make some behavioral predictions based on breed, though actual familiarity with a given dog’s close family members will greatly improve predictability. Not all Border collies can do a good job herding sheep, but very few Labradors can do even a lousy job herding sheep. We’ve repeatedly seen that herding breed dogs are more likely to react to movement, terriers are more likely to fixate on squirrels, bulldogs are more likely to fight than flee when faced with social conflict, some gun dogs are more likely to invade dog and human personal space while failing to recognize the body language of discomfort, and so on.
Likewise, traits like boldness, social fearfulness, neophobia, bite style and strength, ease of socialization, sound sensitivity, and certain other traits that are highly pertinent to many of the behavior problems we repeatedly work with have a genetic basis. Some of these can track by breed.. Clarence Pfaffenberger looked at Labrador puppies bred for guide dog work, and concluded that the pups were able become well-adjusted with just one weekly socialization session during their socialization period, which is much less than most professionals recommend. My observations tally with this: Typical Labradors and golden retrievers can become successfully socialized to humans in general by meeting relatively few during the critical socialization period. By contrast, the independent or rural working or herding breeds tend to need much more carefully managed, broad, and systematic socialization to a wide variety of people to achieve the same level of social comfort. (This doesn’t give carte blanche to be sloppy about socializing any dog, but it means the risks from not socializing an Australian shepherd correctly to home life might be much greater than slapdash exposures for a Cavalier King Charles spaniel.)
Phenotypic traits (those actually exhibited by the organism) can result from a single gene; an example of this is the merle gene, causing patches of dilute and non-dilute color on the coat. A single gene, SILV, controls whether merle patterning is or is not present. Most traits are likely influenced by multiple genes working together. Epilepsy in most dog breeds is an example; scientists have failed to locate single genes associated with epilepsy. (A couple of breeds have specific monogenic epilepsy.)
We know relatively little about genes that affect behavior, but as a general matter, it’s likely that most behavioral traits are polygenic—meaning it’s not likely that research will discover one gene “for” aggression, herding, or anxiety, rather there will be a set of genes that all contribute to whether a dog has this trait.
Epigenetics refers to heritable changes in organisms caused by modification of gene expression, rather than alteration of the genetic code itself. Researchers have shown that how a mother rat nurtures her young after birth directly impacts the anxiety levels of her pups even as adults, and that this results partly from her behavior affecting which of the pup’s genes are expressed. It is assumed that epigenetic processes affect development of physical and behavioral traits in dogs, and provide a crucial link between the nature of the genes themselves and the nurture of the environment. For our purposes, the key points about epigenetic processes are that (a) they can cause different outcomes in genetically similar organisms, (b) epigenetic changes to genes can then, themselves, be inherited; and (c) epigenetic changes from stress are well-known to impact future behavior.
A good, basic glossary of numerous genetic terms can be found at www.genome.gov/glossary/. There are both audio and text versions of the glossary.
Once conception has taken place, each puppy in the litter has received an assortment of genes, half from the dam and half from the sire. (It’s not really quite 50/50 in males, because the Y chromosome does not have as many gene loci as the X chromosome.) After conception, puppies have the genes they’re always going to have. As each individual grows and develops into adulthood, each trait matures within a limited range of values. Where in that range the organism ends up depends on environmental influences following conception. Environmental influences range from the uterine environment all the way to learning before maturation is complete. These environmental influences affect whether a child ends up being 175 cm tall or 180 cm tall, as a simple example. Likewise, they can control whether a puppy ends up terrified of the world or merely neophobic and cautious, fun-loving or recklessly sensation-seeking.
Stress on the fetus while it is still in the womb is the first environmental factor that can affect development of a trait. Adverse effects on both physical and behavioral traits have been widely documented in many mammalian species. There is no reason not to assume the same problems will occur in dogs. Uterine stress occurs as a result of environmental impacts on the mother. These could include malnutrition, toxic chemical exposure, unsafe living conditions, social stress, and many other factors that have been documented in numerous species such as mice, foxes, and marmosets, though little work has been done in dogs.
Neonatal stress will also have an effect. The same stressors that can affect puppies indirectly via stress to the dam during pregnancy may affect puppies directly after birth, while nursing dam malnutrition can continue to affect puppy nutrition. Additionally, the dam’s behavior may cause adverse effects on the newborn puppies, particularly if the dam is unskilled or exhibits abnormal behavior toward her puppies.
As the puppies mature and their eyes and ears become fully operational, the channels for perceiving the world and its potential stressors increase. The puppy socialization critical learning period is underway, and learning at this time creates powerful impacts on the pups’ expectations of the world. Unkind treatment by dogs and people may start to teach puppies about who and what is safe or dangerous. Likewise, if the dam is defensive about the presence of people or other dogs, the puppies will learn socially that people or other dogs are dangerous or threatening. There are various models or proposed mechanisms for this kind of learning, for example “stress contagion” and “vicarious learning.” These early impressions will influence the puppies’ attitudes toward key members of their social groups forever.
Once the critical primary socialization period has passed, most of temperament formation is probably done. At this point, it’s hard to change the basic “presets” of stable temperament features. As we—or our clients—care for a dog, we can confirm or fail to confirm what we might call the dog’s biases toward the world and its denizens, but our opportunity to move the settings is limited and smaller every day. In my experience, the major exception to this is in how we respond to new behaviors that may emerge around puberty. (I suspect this is what’s referred to as a “second fear period,” but that’s a topic for another article.) A previously unreactive dog who lets out an adult-sounding bark at an unusual-looking stranger at 8 months may well be showing genetically-rooted fear or defensiveness. Since this is the first time he has exhibited this particular behavior, the handler’s response will constitute 100 percent of the dog’s learning history with respect to it. (Of course, the dog will have already experienced responses to bold, timid, noisy, or quiet behavior, which should affect that learning history somewhat.) This means the handler’s response can affect future behavioral choices powerfully at this age, and handlers and trainers must be aware of this final period of unusual developmental opportunity or risk.
A final tricky aspect of the development of personality in the dog is that, while environmental influences become much less potent after puberty, the expression of personality is still maturing and changing until the dog reaches social adulthood—at about 2 to 3 years of age. (Giant breeds mature later in this range, while small and medium breeds, particularly working, terrier, herding, and primitive/Spitz breeds tend to mature earlier in the range. The common rule of thumb in working with pit bulls—that if gameness (the desire to fight uninhibitedly with other dogs) has not manifested by the age of 3 years, it probably is not going to show up—is an example of this.) Mild cautiousness appearing at 8 months may not seem concerning, but at this age, a dog is just starting to display age-appropriate independence. That caution can develop into suspicion at 12 months and forward threat behaviors at 15 months, taking by surprise an owner still operating under their initial impressions of a social, attached, and charming puppy. Thus, even after the ingredients are all added to the bowl, changes are still occurring like a yeast dough rising, unseen, under a kitchen towel. Because future changes can be invisible at the time they are triggered, it can be difficult to get a handler to take these sensitive periods and mechanisms of change seriously while the puppy is still puppylike.
One blindingly obvious conclusion we must draw from this review is that puppies are not a blank slate by the time we start interacting with them and affecting their behavior. Even on the day of birth, genetic, epigenetic, and environmental factors have already formed a good deal of the puppy’s adult temperament. This is why “nervous Pointer” puppies cross-fostered to confident dams at birth, or wolf pups hand-raised just like dogs by the experimenters, still grow up to be nervous Pointers and wolves, respectively, even though they may be on the less fearful or more social ends of their potential ranges.
Well-meaning rescuers may find a litter of puppies with a starving, stray, or feral dam, and while perhaps understanding that the dam may never become adoptable, they are full of hope for the puppies. If the dam dislikes people (or, to a lesser extent, other dogs), is in compromised health, and is stressed, those puppies are already facing a steep uphill battle—even if we know nothing about the dam’s genetics prior to her falling onto hard times. In particular, when I hear of a skittish or aggressive dam who comes into rescue and gives birth to a large litter before anyone realized she was pregnant, my heart sinks. The puppies likely were pretty seriously malnourished in utero, and undoubtedly experienced a load of maternal stress that has not stopped just because the dam is now in a safe place. In some cases, drastic as it may seem, removing these puppies from the dam early (cross-fostering is ideal) may be their best chance to grow into acceptable and comfortable pet dogs. As someone who would prefer to see happy, healthy, behaviorally sound dogs adopted rather than the seriously behaviorally disturbed dogs now being placed too often, I generally do not favor saving and placing these litters that are so highly compromised from birth. I recognize, however, that rescue personnel are unlikely to be so cold-hearted and that having puppies to place is a boost for most rescue operations. In that case, I hope at least that the pups can be set up for greatest success by humanely removing them from their people-aggressive/fearful mothers early, socializing them intelligently (including lots of exposure to happy, people-friendly adult dogs when people visit), and doing whatever else is possible to minimize the impacts of what has already occurred.
Breeders can help by breeding dogs with wonderful temperaments. If compromise on temperament is deemed necessary in exchange for some other genetic advantage (e.g., avoidance of a disease risk), better the less relaxed and friendly parent be the sire, whom the puppies never need to meet and learn from. Breeders can choose to provide excellent nutrition, health care, and social interactions to the dam before, during, and after pregnancy. If the dam is showing any suspicion toward dog or human visitors, the breeder can make sure the puppies receive exposure, from early on, to strange dogs and people while the dam has been removed and cannot model defensive or suspicious behavior. Breeders can also make a point of following intelligent puppy socialization guidelines as the puppies become mobile and their senses develop. This means leaving the premises, riding in cars, and seeing people in safe and comfortable situations some number of times before leaving for their new homes. Positive novel experiences at this age will pack a wallop in helping the puppies become adaptable and happy later. But likewise, unpleasant or frightening novel experiences at this age can do a disproportional amount of damage, and breeders need access to education about appropriate handling.
The topic of puppy socialization receives much attention in dog blogs, articles, position statements, and books, so I won’t rehash that information here. The main point to note is that the last parvovirus vaccine in the puppy series is usually given at about 16 weeks—just after the primary puppy critical developmental period ends. Experts agree that improper socialization is far more risky over time than parvo exposure prior to 16 weeks, but many veterinarians focus on the immediate medical risk at the expense of long-term behavioral health. Continuing education of pet owners, veterinarians, and trainers should help reduce misguided advice to keep puppies at home prior to that last parvo vaccination.
Finally, this information can help trainers and behavior consultants working with dogs who have passed the critical periods and now present a relatively settled temperament configuration. It allows us to help owners understand how much change is reasonable to expect. If we have information that allows us to guess reasonably accurately whether the problem is deeply genetic, or may result from actual organic brain changes early on, or is later learned, we can better estimate how amenable the behavior is to change. It’s a lot easier to change concern about men with hats in an otherwise social, resilient dog who had a few bad “men-with-hats” experiences at 4 to 6 months than with a nervous, unresilient, generally socially fearful dog who lived in a barn until she was 6 months old. Owners often have unrealistic goals (“I want to be able to go to the dog park so he can have dog friends like he used to”), and being able to articulate why this is inconsistent with normal adult dog behavior and development can help them accept expectation resetting. If we are called upon by a prospective adopter to assess a dog for adoption, by a service dog handler for a working career, or by a rescue for adoptability, we can better tailor our assessments and recommendations if we know how deeply set and likely changeable or persistent the existing behavioral traits will be.
We can also tailor our recommendations more accurately. For example, if a dog fearful of visitors to the home has a history suggesting deep genetic, epigenetic, and environmental causes for fear, we can choose to focus behavior modification on teaching the dog to relax in a back bedroom when people visit, while if a dog showing the same behavior has more favorable genetics and early learning, it makes more sense to counter-condition him to visitors and work toward him being able to enjoy the company.
There are no crystal balls in this work, and occasionally a puppy with the odds stacked against it turns out to be truly “fine,” while a puppy with every advantage grows into a miserable adult. The more we understand about how these behaviors come about, the better we can advise and coach our clients.
One caution: The more we know about behavioral development, the more discouraged we can get. But behavior can and does change, and we must not let despair displace realism. Clients need to set expectations appropriately, but do not need to hear that their dog is doomed. Sharing some of this information may help ground a client who has been persisting in unrealistic beliefs, and other clients may simply find it interesting to discuss. But ultimately, change comes from addressing the dog in front of us, so it’s important for neither client nor consultant to bog down in the (often vague or unknown) history leading to that dog sitting in front of us.
A complete account of the information summarized in this article could easily fill some number of textbooks. Necessarily, much simplification and even oversimplification has occurred. It is not necessarily easy to keep up on developments in this area because the topic involves synthesizing information from numerous highly technical fields. I hope that this article has raised awareness of the importance of understanding the role of temperament and behavior development in working with dogs, particularly for veterinarians, dog behavior consultants, working dog trainers, and rescue/shelter personnel. I also hope that scientists continue to investigate these issues in dogs as well as in the more usual research species, since these topics greatly affect canine welfare. I would love to see more information about this become available and accessible to non-scientists. No doubt we will see new information, refinements, and even changes to the information presented here as research continues.
Greta Kaplan, CDBC, is a dog behavior consultant in the Portland, Oregon area. She lives with three dogs, two of them Border collies and the third a relaxed household companion. Greta has dabbled in many dog sports and earned numerous flyball titles. As a former lawyer, she is very fond of a good argument and of citations to authority.