Belyaev and the Farm Fox Experiment

Many a child has seen wild animals at the zoo, in films, or even in cartoons and wished they could own their very own cuddly baby lion. However, adults know that owning their very own Simba could spell drastic consequences—and families instead give their love to docile dogs and cats that make far better companions. However, recent research into the domestication of wild foxes shows that even wild animals can be bred over many generations to become docile and mutual human companions.
The interplay between behavioral genes and physiology and morphological development is significant. Because of the way a particular gene is located on a chromosome, genes are often replicated and inherited in tandem—for example, the gene for docility may be hypothetically located next to the gene for a small skull, and thus they may be replicated and transferred to offspring together.
This phenomenon is especially apparent on the experimental farm in Novosibirsk, Siberia, where Russian geneticist Dmitri Belyaev has bred foxes for over 40 years. Essentially, he has turned over 700 foxes into a group of docile, human loving creatures. With these changes in behavior have also come many physical manifestations, including changes in skull shape, reproductive style, responses to sounds, fur coloration, and ear shape. The changes in these foxes also mimic the differences between dogs and wolves, suggesting that an important link lies between the wild hunters and our best friends.

Belyaev’s Hypothesis

Throughout the course of his work on domestication, Belyaev noticed that many domestic animals had similar morphological and physiological changes. There were variations in body size, hair turned wavy or curly, tails were shorter, ears became floppy, and animals lost their seasonal rhythm of reproduction. Belyaev believed that these changes were the result of selection for domestication. A balance of hormones and neurotransmitters regulates an animal’s behavior. Genes in turn regulate this balance. In theory, selecting for an animal’s behavior would give rise to changes in the animal’s physical and behavioral development since these pathways are finely interconnected.
However, in order for Belyaev’s hypothesis to be plausible, there were two stipulations. The first was that domestication had to put an animal under strong selective pressure. In addition, genes had to be responsible in part for an animal’s tamability. In fact, Belyaev’s institute has studied both of these aspects. They discovered that genes are about 35% responsible for the variations in foxes’ defensive response. Also, when the group tried to domesticate other animals such as rats and river otters, a very small number of them contributed genetically to the next generation. This made it clear that the process of domestication must place wild animals under severe selective pressure and extreme stress.

The Experiment

Belyaev began his experiment by taking 30 male foxes and 100 vixens from a commercial fur farm, bypassing the initial steps of domestication since these foxes were already tamer than ones from the wild. The foxes were placed in cages and were allowed timed brief contact with humans and were never trained. From there, foxes were strictly selected for reproduction solely based on tameness. Once a litter was born they would be caged with their mother until they reached about two months of age and were caged only with their littermates, and finally at three months each pup is put into its own cage.
A strict series of tests was developed to evaluate each fox for tameness. Starting at one month old and repeated for six months, the experimenter would offer food from his hand while attempting to stroke and handle the pup. This was tested both when the pup was caged and while moving freely with other pups in a larger enclosure. Once the foxes reach sexual maturity, at around seven months, they are assigned to one of following three classes:

Aggressive foxes demonstrated the following behavior

while tamer, docile foxes behaved as follows

An even higher-scoring category emerged in the sixth generation, foxes in Class IE known as the “domesticated elite” demonstrate the following behavior:

• Eager to establish human contact
• Whimper to attract attention
• Sniff and lick experimenters like dogs

By the tenth generation, 18% of the pups were elite; by the 20th, 35% were elite; today 70-80% are considered elite.

Today, 40 years and 45,000 foxes later, Belyaev’s experiment has achieved a population of 100 foxes that are:

• Docile
• Eager to please
• Competitive for human attention
• Unmistakably domesticated

Physical Changes

The physical changes in the farm foxes strongly mirrored those found in current domesticated species. However, perhaps the most influential physical change was not an immediately observable trait, but rather a shift in the timing of development. Farm foxes were observed to reach critical sensory development stages earlier than their non-domesticated cousins. However, domesticated foxes reached a critical behavioral landmark much later. As this landmark influenced the pup’s fear response, this combination of developmental mutations effectively lengthened the time a fox pup was able to socialize without fear – that is, the period during which a fox pup could become accustomed to humans. Similar disruptions in developmental timing have been observed at a genetic level in much lower organisms; for example, in C. elegans, mutations in certain “heterochronic” genes influence the duration of larval stages, either accelerating or hindering the development of larval cells into adult cells (Griffiths, 2008). As such genes have been found in vertebrate genomes as well, it seems possible that a mutation in one or more heterochronic genes could be at least partially responsible for the foxes’ lengthened socialization period.

The other observable traits, if they were “linked genes,” may have simply tagged along with these genetic changes in developmental timing by virtue of their physical proximity to the selected-for genes. However, certain traits – such as piebaldness – have been shown to be linked to changes in the development of the nervous system, and as such could result from selection for neurological traits (such as tameness). Melanocytes – the cells responsible for the pigment of our skin – arise early in development from cells of the neural crest. As the cells of the neural crest develop, they fan around the nascent head to form the face, as well as spreading about the body to form the tissues of the nervous and endocrine systems. Some of these traveling cells eventually develop into melanocytes and lodge themselves in the skin, where they produce pigment (Leroi, 2003). As the neural crest is what ultimately develops into the nervous and endocrine systems, a mutation causing tameness could also act to disable melanocytes, preventing them from producing pigment and causing the characteristic white stripe on a domesticated animal’s face.

While the alternative explanations for these changes seem valid on the surface, the authors do a good job of refuting them. In particular, the notion that inbreeding could cause these traits is effectively discredited, as the inbreeding coefficients (a measure of how inbred a population of breeding animals is) are so low that the probability of a trait owing its presence to inbreeding is only 2 to 7 percent. They also argue that the dominant traits they observed could not have been “hidden” in the heterozygosity of parent foxes. While this is true, such traits can actually be hidden due to variable expressivity. In other words, genetic variation at other loci (areas of the genome) could suppress or alter the effects of a gene. This is true of some cases of human polydactyly, in which the trait seems to “skip” a generation despite it being a dominant trait (Griffiths, 2008). While the authors make a valid point using the dominance argument, they seem to oversimplify it.

Such variable expressivity, if it exists in the farm foxes, is arguably a consequence of the genetic system in question being a quantitative trait, controlled by complicated systems known as polygenes. However, the authors address this possibility as well, using the commonalities between the fox mutations and mutations found in other domesticated species as evidence that polygene disruption is not at the heart of the foxes’ physiological changes. Their argument, once again, seems fairly sound. According to the authors, disruptions in the delicate control mechanisms governing the development of the neural and endocrine systems produce the stereotypical changes associated with domestication. Such disruptions are due to selectively breeding for what is essentially a behavioral, not a quantitative, trait. Behavioral traits, unlike quantitative traits, are emergent properties: they arise from complex interactions between the developing nervous and endocrine systems, the rest of the body, and the environment itself. Thus, selecting for “tameness” would produce systematic changes that result in the “domesticated” set of traits. Since mammals (and indeed, many vertebrates) share many things in common between these organ systems, it makes sense that breeding for this one trait would cause similar changes in all of them.

Selection and Development

Beleyaev found a clear link between behavior and physical development. The selection of foxes for docility triggered profound changes in their physical development. For example, the docile foxes opened their eyes at an earlier age (an ontogenetic process) and had floppier ears, which are more characteristic of newborn pups. Additionally, the novel coat colors that were seen in the elite population could also be an indication of the change in timing of embryonic development. The foxes acquired coats with large areas of depigmentation similar to the coats of many breeds of dogs, as seen below.

Another interesting change was that the skull morphology became “feminized”—that is, the skulls of the males became smaller and more similar in proportion to those of females. Finally, and most importantly, the reproductive cycle of female foxes was altered. Domesticated foxes became sexually mature on average a month earlier than wild foxes do, and give birth to slightly larger litters. The mating season also lengthened, and some females even bred out of season, while still others bred twice a year. These changes are extremely significant, because out-of-season mating has never been previously observed in wild foxes.

Lesson Learned

Belyaev and colleagues learned that physical changes can occur earlier in domestication than previously thought because of the stresses of captivity. Some of the most noticeable changes seem to be pedomorphosis, in which juvenile characteristics are retained in the adult form of an organism. Yet many unanswered questions persist. It remains to be seen how much further in the domestication process the foxes can go. They are not quite dogs, but researchers believe that their temperaments point in that direction. The foxes, however, will only realize their potential with closer human contact. With the downturn of the Russian economy, the fox population is dwindling without the financial means to support it. To support its continued research, the Institute of Cytology and Genetics has turned to commercial ventures, such as selling the foxes for pelts or for pets.

Video example of successful domesticated pet fox from the Russian Institute:

Trut, L. (1999). Early Canid Domestication: The Farm-Fox Experiment American Scientist, 87 (2) DOI: 10.1511/1999.2.160