By Sam Basso | SamTheDogTrainer.com
OVERVIEW: Why do littermates develop different temperaments and abilities? Genetics, puppy raising, cloning research, and working-dog selection explain why DNA is only part of the outcome.
Two sisters from the same litter can grow into remarkably different dogs. One may enjoy unfamiliar people and new places while the other is hesitant. One may have the qualities for demanding working-dog training while the other has no interest in that work. Even when they share a home, they do not necessarily become the same kind of adult.
Why? The puppies did not receive identical genetic combinations, and they did not live identical lives. Their bodies and nervous systems developed through individual interactions with their mothers, littermates, people, health, and surroundings. The question is not whether heredity matters. Of course it does. The question is how much we can predict about a finished dog from its inherited material, and what must still happen after conception and birth.
That question has a new, practical test. In October 2026, Warhound announced $200 million in seed financing for a system combining cloning, genetic technology, training, animal care, and equipment to produce U.S. working dogs for military, law-enforcement, and other assignments. These are announced objectives, not demonstrated operating results. Could the technology increase the chances of producing exceptional working dogs? Possibly. Could it repeatedly manufacture the complete capabilities of an elite military dog from an embryo?
I have my doubts.
What Are the Biggest Obstacles Warhound Will Need to Overcome?
In my opinion, these are some of the major obstacles Warhound may face, although their existence certainly does not mean the company cannot succeed. Genetics, cloning, artificial intelligence, and modern reproductive technologies may offer significant advantages. However, based on my review of the scientific literature and the history of working-dog breeding and development programs, I have identified at least thirteen challenges that Warhound will likely need to address.
1. Identifying the genetics of exceptional working temperament. We do not yet know all the genes responsible for the many characteristics that make an outstanding police or military working dog. Nor do we fully understand how those genes interact with one another, influence brain development, or affect other biological systems.
2. Producing the complete package of desirable characteristics. An exceptional working dog needs far more than one outstanding ability. It must possess the appropriate combination of motivation, sensory abilities, environmental responses, persistence, learning, handler cooperation, and other behavioral characteristics in the ideal proportions. As Armin Winkler's work, mentioned later, illustrates, these are not interchangeable qualities, and their interactions matter.
3. Producing the right physical structure and health. The dog must also have appropriate size, conformation, athletic ability, endurance, joint health, and resistance to disease. Selecting or modifying genes for one desirable characteristic may have consequences elsewhere. Even generations of conventional breeding have not eliminated complex inherited conditions such as hip dysplasia.
4. Reproducing exceptional working ability through cloning. Cloning an outstanding dog reproduces nearly the same nuclear genetic material, not necessarily its extraordinary abilities. Those abilities developed through genetics, biological development, learning, training, health, and experience. Published cloning studies provide some encouraging evidence for detection work, but do not establish consistent reproduction of elite special-operations performance.
5. Developing puppies into successful adults. Scott, Fuller, and Pfaffenberger demonstrated why genetic selection cannot substitute for appropriate developmental experiences. Warhound will need a sophisticated puppy-raising, social-development, training, and evaluation system. Even genetically similar puppies will not have identical experiences or necessarily become identical adults.
6. Establishing reliable breeding lines across generations. An exceptional individual is not necessarily an exceptional producer. If Warhound intends to breed from cloned or genetically selected dogs, desirable characteristics must be transmitted reliably to their offspring and subsequent generations. Developing prepotent breeding animals and successful producers of producers is different from cloning an individual champion. Warhound's precise long-term breeding strategy has not been publicly established, so I don’t know more about this than you do.
7. Maintaining genetic diversity. Concentrating desirable characteristics within a limited number of elite donors or breeding lines could also concentrate undesirable inherited variants. A sustainable breeding population must balance selection for performance against risks involving inbreeding, reproductive fitness, health, and genetic diversity.
8. Overcoming reproductive inefficiency and cloning-related welfare risks. Published canine cloning studies report low embryo-to-live-birth efficiency, although methods may improve. Warhound must account for unsuccessful embryo transfers, pregnancy losses, neonatal complications, and the welfare of egg donors and surrogate mothers, not simply the number of surviving puppies.
9. Demonstrating long-term health, fertility, and longevity. Some consequences of breeding decisions or genetic modifications may not become apparent until dogs reach maturity or old age. Others may emerge in subsequent generations. Establishing that puppies appear healthy does not prove that they will remain healthy, fertile, and capable throughout their lives.
10. Meeting the exceptionally demanding requirements of elite military service. Producing a qualified detection dog or conventional police dog is not equivalent to producing a top-tier special-operations dog. Only a subset of existing working dogs possesses the particular combination of characteristics required for these specialized assignments. Warhound must demonstrate improvement against the actual standards for each intended job, not simply an overall working-dog graduation rate.
11. Validating results with sufficient animals, generations, and time. A few successful clones or genetically engineered puppies would demonstrate feasibility, not repeatability. Depending on the claimed improvement, validation could require hundreds or thousands of dogs, multiple genetically distinct donors, successive breeding generations, independent comparisons, and years of follow-up. My estimates, below, are hypothetical; Warhound's actual validation plans and proprietary data are unknown.
12. Providing responsible lifetime care for dogs that do not qualify. Even a highly successful program will produce dogs unsuitable for their intended assignments. Some may find other working roles or good homes, while others may require specialized management or lifelong care. The American shelter system already faces substantial placement challenges, particularly for large dogs. Warhound must demonstrate a viable and adequately funded disposition system rather than assume that shelters or private adopters can absorb its unsuccessful candidates. Is pre-planned euthanasia part of the system so that dogs that are unsuitable for work or as pets been taken into account?
13. Competing economically with established working-dog systems. European breeders and trainers have developed generations of breeding knowledge, selection practices, training infrastructure, and experienced working-dog communities. Warhound must compete with that entire system, not merely with the genetics of individual dogs. The relevant economic measure is the total cost of producing, qualifying, maintaining, retiring, and responsibly caring for each successful working dog, including the costs associated with unsuccessful animals.
The Ultimate Test
Warhound does not necessarily have to solve every scientific mystery surrounding canine genetics to produce useful improvements. Better genetic selection, more effective breeding decisions, and improved developmental programs could produce meaningful results without making working-dog performance completely predictable. But there is an enormous difference between improving the odds of producing a good working dog and reliably manufacturing an exceptional one.
The question is whether Warhound can repeatedly produce healthier, capable, dependable working dogs at a higher rate, across successive cohorts and potentially generations, while maintaining genetic diversity, animal welfare, and economic competitiveness.
That is an ambitious objective. It may prove achievable in some respects and not others. The appropriate response is neither to dismiss the technology nor to accept its promises before the results are demonstrated.
We should judge Warhound by the complete outcomes of its breeding, development, training, and lifelong animal-care system, not merely by its technological achievements or the pedigree of its founding dogs.
Why don’t puppies from the same litter turn out alike?
Full siblings ordinarily inherit different combinations of variants from the same two parents. They share considerable DNA, but they are not genetic duplicates. Conditions during gestation, birth, development, and later life can also differ. Two puppies can receive the same scheduled training and still notice different things, approach people differently, or learn different associations from the same event.
Imagine two sisters who leave the litter for different households. One is exposed to visitors, outings, unfamiliar dogs, and a skilled handler. The other has a quieter life and a different learning history. Perhaps one develops an orthopedic problem that changes the way she moves and responds to handling. We cannot infer that the difference in their adult behavior was caused by their homes merely because the homes were different. Genetic predispositions, physiological conditions, and individual experience could all contribute.
For me, a temperament label is not an explanation. I would want to know what each dog does, under what circumstances, at what age, with which people, and after what sequence of events. Only then could we begin separating plausible inherited tendencies from learned responses, changing physiological states, and remaining unknowns.
What did Scott and Fuller establish about heredity and development?
John Paul Scott and John L. Fuller showed that both inherited differences and developmental experience matter. Their studies of five breeds examined social responses, emotional development, learning, and differences among dogs at different ages. The work, eventually presented in Genetics and the Social Behavior of the Dog (1965), was neither a victory for heredity nor a victory for upbringing.
Breeds differed in measured characteristics, but one breed was not universally superior in every test. Early interactions, particularly during sensitive periods of development, affected later social behavior. A sensitive period does not imply that experience after that period becomes irrelevant, nor that every puppy follows precisely the same timetable. It means that timing can change the consequences of experience.
That distinction matters for breeders. Selecting appropriate parents and creating appropriate developmental conditions are complementary activities. The genetic promise of a puppy and the eventual abilities of an adult are not the same thing.
Where did Pfaffenberger’s guide dogs come from, and what changed?
The original guide dogs at Guide Dogs for the Blind did not all come from generations of selectively bred, successful guide-dog parents. The organization was founded in California in 1942, and its own history states that most of its earliest dogs came from animal shelters. One early German Shepherd came from a Pasadena shelter. The school later recognized that it needed dogs selected more deliberately for the physical and behavioral demands of guiding a person who was blind.
Clarence Pfaffenberger’s 1963 book, The New Knowledge of Dog Behavior, describes a very low early success rate: approximately 9% of dogs started in training became working guides. That is a figure reported by Pfaffenberger about the historical program, not an independently reconstructed modern trial.
Guide Dogs for the Blind established a specialized breeding program in the late 1940s. Instead of depending primarily on the uncertain backgrounds of dogs arriving from shelters, it began producing puppies from selected breeding stock thought suitable for guide work. This was an important genetic and management change. The precise pedigree and previous working status of every foundation breeding animal are not established by the public institutional history, so it would be incorrect to claim that they were all successful guide dogs bred to other successful guide dogs.
But the development of specialized breeding stock was not the end of the story. Pfaffenberger collaborated with Scott and the researchers studying puppy development. The program also developed puppy testing, attention to developmental timing, and family-based puppy raising, including 4-H homes. In those homes, puppies encountered people, ordinary household routines, and environments they would need to negotiate before returning for formal training.
Pfaffenberger reported that by 1958 and 1959, approximately 90% of dogs produced under the revised breeding and developmental system became working guides, despite more demanding standards. That comparison is striking, but breeding, selection, testing, and raising had all changed. It is not a controlled experiment demonstrating that puppy raising alone caused the entire improvement.
The more targeted evidence concerns developmental experience. Pfaffenberger and Scott’s 1959 paper investigated the relationship between delayed socialization and later guide-dog trainability. It supported the importance of the timing of placement and social experience under the conditions they studied. We should not convert the historical findings into a universal deadline or promise of success for every puppy.
The crucial lesson is that the organization moved from a poorly matched supply of dogs to an integrated system of selected breeding, developmental care, and evaluation. Even after breeders improved the genetic starting point, the experience of the puppy remained an essential part of producing a successful guide dog. That lesson is directly relevant to Warhound, but it does not tell us that Warhound will attain a 90% qualification rate for its very different assignments.
There is also a less frequently asked question: what happened to the approximately 10% who did not qualify?The historical figures do not tell us the ultimate disposition of those individual dogs. Modern Guide Dogs for the Blind has a structured career-change process, including reassignment to other suitable work, assessment for pet placement, and adoption through puppy raisers or screened homes. That contemporary system must not be retroactively attributed to the 1950s program. The unanswered question will reappear at greater scale if a new breeding project produces dogs selected for unusually demanding military tasks.
If we clone an exceptional dog, will the clones have its abilities?
Cloning can reproduce nearly the same nuclear genetic material, but it cannot reproduce the donor’s complete life. A cloned puppy does not inherit the donor’s training, memories, relationships, injuries, learned responses, or experience with handlers. Differences involving egg donors, gestation, epigenetic regulation, development, and subsequent environments can also matter.
There is a legitimate positive finding. In a small Korean study, Choi and colleagues (2014) evaluated seven clones derived from an accomplished drug-detection dog. Six qualified in that detector-dog program, an 86% selection rate versus about 30% for the program’s reported conventional comparison. The result shows why cloning successful animals is worth investigating. It does not establish an 86% success rate for all clones, all types of detection work, or an elite multi-purpose military role.
There are also accounts of earlier military evaluations in which cloned dogs reportedly did not reproduce all the donor’s exceptional abilities. Detailed, independently reviewable findings from those claimed evaluations have not been established here. They should not be assigned invented sample sizes, units, reasons for rejection, or measured failure rates. This absence of public records is not proof that the evaluations never occurred. It is a boundary on what the article can substantiate.
The distinction is straightforward: reproducing the donor’s nuclear genome is not equivalent to reproducing the donor’s demonstrated career. An outstanding adult may be an unusually favorable combination of heredity, development, training, opportunity, health, and human relationships. Cloning does not make all those factors happen again.
Prepotency: Can Warhound Produce a Bloodline That Consistently Breeds Exceptional Dogs?
Experienced breeders recognize an important distinction between an exceptional dog and an exceptional producer.
A champion may possess outstanding characteristics but fail to produce offspring of comparable quality. Another dog may not have been the greatest performer itself, yet consistently produce outstanding offspring. Breeders refer to the latter characteristic as prepotency.
A prepotent sire or dam reliably transmits particular desirable characteristics to its offspring. Some breeders speak of a producer of producers, an animal whose offspring become excellent breeding animals themselves. But prepotency for one characteristic does not establish prepotency for every desirable trait.
This is why knowledgeable breeders study more than an individual dog's accomplishments. They examine parents, grandparents, siblings, offspring, and subsequent generations. They want to know whether desirable characteristics occur throughout the family or whether an exceptional individual was an unusual result.
That is consistent with the broader pedigree-based selection principles reflected in Pfaffenberger's work and modern guide-dog breeding programs.
Now consider the challenge facing Warhound. Suppose the company identifies an extraordinary special-operations dog and produces genetically similar clones. The clones may possess nearly identical nuclear genomes. But what happens if those clones are used for breeding? Their puppies will inherit genetic material from both parents, and the combinations will differ. The original dog's exceptional combination of genetic variants will not necessarily be reproduced. Some offspring may inherit more of the desirable characteristics. Others may inherit fewer. Still others may inherit combinations that interact differently. And the behavioral consequences may depend upon development and environmental experience.
If Warhound intends to create a sustainable breeding population rather than continually cloning individual animals, it faces the much more complicated task of making desirable characteristics reliably transmissible across generations.
That is what breeders mean when they talk about creating a line that breeds true. For relatively simple inherited characteristics, genetic testing can help identify animals carrying particular variants and, in some cases, establish predictable inheritance. But a complete working temperament is not a simple inherited characteristic. It includes numerous behavioral tendencies, neural processes, physiological characteristics, and interactions with experiences. Add physical structure, orthopedic health, fertility, disease resistance, endurance, and longevity, and the challenge becomes still more complicated.
Geneticists can improve breeding predictions by collecting detailed information on parents, siblings, and descendants and calculating estimated breeding values. But improving a population's average genetic merit is not the same thing as guaranteeing an exceptional individual in every litter. Nor is making a breeding line more genetically uniform automatically beneficial.
Breeders have long used linebreeding to concentrate desirable characteristics, but concentrating ancestry also increases the possibility of concentrating harmful inherited variants.
A sustainable program must balance the selection of desirable characteristics against the preservation of genetic diversity. The challenge may therefore be greater than identifying the ideal individual working dog. It may be producing a population of breeding dogs that repeatedly produces desirable offspring while maintaining health, fertility, longevity, and enough genetic diversity to continue improving. And those characteristics would have to be demonstrated in successive generations.
The puppies would have to mature, be trained, and undergo evaluation. Some would become breeding candidates. Their offspring would then have to be raised and evaluated. Only through repeated results could we begin to establish the consistency of the breeding system.
Warhound could avoid some of this problem by continually cloning selected donors instead of creating a conventional breeding line. But that creates a different production system, with its own reproductive efficiency, cost, genetic diversity, developmental, and animal-welfare questions.
Producing one exceptional dog is an achievement. Producing a dog that consistently produces exceptional offspring is another. Establishing a breeding population that continues producing exceptional dogs, generation after generation, is a much greater accomplishment. That is another standard against which the long-term success of genetic technology in working-dog breeding should be measured.
What do experiments cloning horses and other mammals tell us?
Experience with cloned horses, livestock, and other mammals confirms that mammals can be cloned. Horses have been cloned to reproduce valuable genetics associated with competition and breeding. Their offspring still require sound development and training; a clone does not inherit the donor’s athletic career.
The comparison with dogs has limits. Athletic horses have complex physical and behavioral requirements, so it would be unfair to describe their performance as simple. But qualification for a particular elite multi-purpose canine assignment can involve a particularly broad combination of detection, environmental adaptability, handler cooperation, physical capability, response to changing conditions, and sometimes apprehension work. A successful example in one species or a narrowly defined performance program cannot establish reliable reproduction of that whole canine package.
A second issue is how the animals are produced. Reproductive efficiency, pregnancy losses, neonatal care, and the welfare of donors and surrogate mothers must be assessed separately from the performance of surviving adult clones.
How efficient and healthy is dog cloning?
Published canine cloning has often been inefficient when measured from transferred reconstructed embryos to live offspring. In Insights from One Thousand Cloned Dogs (2022), Olsson and colleagues reported an average live-offspring yield a little above 2% per transferred reconstructed oocyte across a large research operation, despite considerable technical improvements. That is an embryo-transfer denominator. It is not the mortality rate among puppies born alive, and it should not be compared numerically with ordinary litter survival as if the denominators were identical.
Low embryo-to-live-birth efficiency raises real questions about the number of egg donors, surrogate mothers, procedures, failed transfers, pregnancies, and resources involved in producing each healthy puppy. These costs and welfare consequences are part of the system, even if they never appear in a final photograph of a successful clone.
Does cloning necessarily mean that every surviving adult dog will be sick or die young? No. The evidence does not support that blanket conclusion. The large canine study reported improving postnatal care and encouraging observations among surviving healthy animals, but that is different from comprehensive, independent, lifelong comparative evidence proving equal health, fertility, disease resistance, and lifespan across populations. Studying healthy survivors alone can conceal losses that occurred earlier in production.
In February 2026, the British Small Animal Veterinary Association emphasized the limited success of companion-animal cloning, the welfare of cloned animals and surrogate mothers, and the lack of sufficient long-term health evidence. Its position concerns commercial companion-animal cloning; the scientific welfare issues it identifies are relevant questions for any large-scale canine cloning program, not a direct trial of Warhound.
We should therefore reject two oversimplifications: that cloning inevitably creates unhealthy adults, and that producing a viable puppy proves the technology has no health or welfare cost.
If we cannot breed away hip dysplasia, can we engineer the entire working dog?
Hip dysplasia demonstrates both the value and the limits of genetic selection. Breeders have screened hips and selected breeding dogs for decades. An analysis of Orthopedic Foundation for Animals records from 1970 to 2015 found improvements across many breeds. Yet hip dysplasia remains because it involves multiple genetic influences and developmental and environmental factors. The screening records also represent evaluated dogs, not every dog in the population.
Mikkola and colleagues (2021) investigated 46 genetic markers in nearly 1,600 dogs from ten breeds and validated 21 associated loci across 14 chromosomes. Those are associated regions, not 21 proven causal genes and not a complete blueprint for an ideal hip joint. A region can contain many candidate genes, and an association may not explain the actual biological mechanism.
The challenge is larger when a genetic change could affect other characteristics. Some genes and pathways contribute to several biological processes, an effect known as pleiotropy. Genetic variants can also interact, and associations can reflect linked DNA rather than direct causation. A modification intended to improve one measure cannot be presumed harmless to growth, muscle development, metabolism, fertility, or other traits.
Time complicates evaluation. The Orthopedic Foundation for Animals provides preliminary hip assessments in puppies but requires 24 months of age for final hip certification. That still does not reveal all later orthopedic disease or durability under demanding work. Reproductive outcomes require maturity and breeding records. Disease susceptibility and working longevity require observation across years. Lifetime health and lifespan take a lifetime to evaluate.
Concentrating breeding around a narrow group of extraordinary donors may also reduce population diversity and increase some inherited risks, depending on how a program is designed. Cloning is not the same as inbreeding, but repeated reliance on a restricted genetic base requires deliberate diversity management.
A genetically promising young dog with excellent hips is not automatically a sound, fertile, disease-resistant adult that will work successfully for years.
What do scientists really know about inherited temperament?
Temperament is partly heritable, but it is not one gene or one trait. Researchers measure defined behaviors, sometimes using questionnaires and sometimes structured tests, and their categories do not always map onto the language experienced working-dog evaluators use.
Heritability is a population statistic. It estimates the extent to which differences in a measured characteristic are associated with genetic differences in a particular population and environment. It does not mean that an individual dog’s behavior is a fixed percentage inherited and the rest learned.
In a 2026 study of 4,841 Labrador Retrievers in guide-dog programs, Riser and colleagues estimated genomic heritabilities of roughly 0.08 to 0.21 for 17 measured behaviors. Genomic information modestly improved average breeding-value prediction accuracy from about 0.46 to 0.50. That can help breeders choose among candidates; it is not an individual puppy’s complete temperament forecast.
Other research underscores the limits. Lord and colleagues (2025) examined data for more than 3,000 dogs and did not validate the expected predictive value of previously proposed single genetic variants for behavioral traits. Canine behavioral genetics is advancing, but there is no accepted count of all genes contributing to courage, directability, persistence, or the other qualities required in a special-operations dog.
Think of the familiar question about human intelligence: could a breeder or a laboratory produce a generation of Einsteins? Intelligence has substantial genetic influences in human research, but predicting a historic scientific achievement is not the same thing as estimating a cognitive test score. Education, interests, opportunity, persistence, chance, and numerous developmental factors also matter. The same warning applies to the idea of genetically manufacturing a dog whose exceptional working talents emerge together, on demand.
How do brain development, the body, and pain affect temperament?
Behavior depends on a developing nervous system in continual interaction with the body and the environment. Genes influence neural development, signaling molecules, receptors, sensory and motor systems, and a great many regulatory processes. Experience influences what those systems do. Hormonal, immune, digestive, cardiovascular, and other physiological systems send signals to the brain and receive signals from it.
We know a great deal about neural and endocrine function, but do not possess a comprehensive model that takes an embryo’s DNA and reliably predicts that dog’s precise adult responses to different people, threats, frustrations, and physical discomfort.
Pain is a concrete example. It is not merely a signal from an injury. Nociceptive input is processed and modulated through peripheral nerves, spinal circuits, and multiple brain systems involved in sensation, attention, affect, and action. History, context, physiology, and the ability to avoid contact can alter the dog’s response.
One dog in pain might escape. Another might become quiet or avoid handling. Another may threaten. Another may attempt a physical attack. These observations are not interchangeable, nor do they prove a particular internal emotional state. A threat is not necessarily an attack, and a unilateral attack is not a reciprocal fight. The presence of a violent response does not by itself demonstrate that the animal was experiencing rage.
Camps and colleagues (2012) described twelve clinical cases involving pain-related aggression, eight of them involving hip dysplasia. The small, selected sample cannot tell us how common attacks are among dogs with painful hips. It does show why medical history and the chronology of behavior are relevant to interpretation.
Some dogs continue to perform despite injury or discomfort; others reduce activity or change their responses sooner. Continued performance does not mean an animal feels no pain or is unaffected by it. Pain sensitivity, behavioral response to pain, capacity to continue a task, and welfare are separate dimensions. It would be scientifically indefensible to predict all of those responses from a single cloned genotype.
How many working-dog traits would a genetic program have to get right?
Armin Winkler’s Tell Me About Your Dog articles, published as Dog Terms Parts I and II on Rivanna K9 Services, illustrate why descriptions such as “high drive,” “good nerves,” and “hard dog” are insufficient. Winkler is an experienced working-dog practitioner and instructor with a background in applied animal behavioral science. His articles offer a detailed practical assessment framework, not a validated list of individual genes.
He distinguishes stimulus threshold from the kind of response produced; reaction intensity from persistence and endurance; boldness from courage; and directability, a dog’s willingness to accept guidance, from obedience already produced through training. He also challenges the idea that “fighting drive” is a single natural drive, describing it instead as a combination of motivational and behavioral components that may become relevant in protection training.
His categories should not be equated automatically with the constructs in genetic studies, nor should the presumed internal motivation be treated as proven solely from outward behavior. But the operational question he raises is powerful: before selecting genes for a trait, can we even define and measure that trait consistently across dogs, ages, handlers, and situations?
A dog might be unusually persistent yet difficult to redirect. Another might engage readily with a task but react poorly to unexpected noise. Another might be strong in one training exercise and unable to recover appropriately after a disturbing experience. These characteristics may interact, and the combinations required for detection, guiding, patrol, or elite special operations are not the same.
A 2018 systematic review by Brady and colleagues found important gaps in the reliability and predictive validity of working-dog behavioral tests. Passing an early aptitude test is not the same as proving mature occupational performance. Any genetic program built around these outcomes has to begin with usable definitions, actual measurements, repeated follow-up, and validation in the intended job.
Can a perfect temperament be combined with the perfect body?
Even if a breeding program improved the probability of desirable behavioral traits, the same dog would still need appropriate size, conformation, joint structure, muscle function, sensory capabilities, coordination, health, and endurance. The optimal combination depends on the assignment. A dog used extensively for scent detection may require different capabilities from a dog that performs demanding apprehension work.
These requirements cannot be treated as an independent checklist where every box can be optimized without tradeoffs. Selection may affect correlated traits. Genetic variants may influence more than one biological pathway. An animal may be outstanding in training but have insufficient durability, heat tolerance, or physical recovery for a particular job. Some weaknesses are not visible until maturity or after years of work.
The challenge is therefore not selecting the best hips, the most persistence, the most strength, and the greatest sensitivity to every stimulus. It is selecting an appropriate combination that functions safely and reliably in real assignments, while maintaining health, fertility, genetic diversity, and animal welfare. There is no validated test-tube recipe for that complete package.
Aren’t the best special-operations dogs already selected from a much larger population?
Elite multi-purpose canine selection is a much narrower question than qualifying an ordinary guide dog, detection dog, or general patrol dog. A dog can be very good at one working task and still fail the particular criteria of a highly specialized operational assignment.
European working-dog programs have accumulated breeding, training, and evaluation experience through generations of breeders, handlers, clubs, and working organizations. In some Dutch Shepherd working lines associated with KNPV, performance breeding includes dogs with substantial Belgian Malinois-Dutch Shepherd ancestry rather than closed-register purebred pedigrees. The Working Dutch Shepherd Association of America notes that many dogs carrying KNPV-style Bloodline Registration Numbers have such mixed ancestry. A BRN is a lineage-tracking designation, not proof of purebred registry status.
That fact matters because selecting for function need not be identical to selecting for pedigree purity. It does not prove that every crossbred dog outperforms every purebred one. Nor does it establish that purebred Malinois cannot succeed. Similarly, the claim that German Shepherds are never used is too categorical. An August 2026 Naval Special Warfare multi-purpose canine solicitation expressly allowed Belgian Malinois, Dutch Shepherds, German Shepherds, and their crosses. Procurement eligibility does not tell us the current breed composition of each special-operations unit.
Public evidence does not establish a reliable percentage of all European-bred puppies that eventually qualify for a Tier 1 special-operations assignment. We should not manufacture a figure or substitute ordinary military working-dog acceptance rates. The relevant point is that an elite dog has already survived multiple layers of selection, training, and evaluation. Replicating its DNA is not the same as replicating that entire selection outcome.
Warhound is competing with the cumulative expertise and candidate pool of a working-dog ecosystem, not just with individual European breeding kennels.
Could AI choose the right embryo and produce that package repeatedly?
AI can help search genetic datasets, estimate breeding values, analyze health records, compare cohorts, and potentially improve candidate selection. It may be particularly useful when measurable traits have strong data, consistent definitions, and prospective validation. It can also support developmental tracking as a puppy matures.
But a statistical association is not a complete causal mechanism, and a good prediction within one breed or raising program may not transfer to another. The model must know what the endpoint means. “Passed a puppy test,” “qualified as a detection dog,” and “served reliably for years in a specialized unit” are three different outcomes.
Even with extensive genetic records, no embryo-selection model can know each dog’s future prenatal conditions, illnesses, handlers, training encounters, environmental exposures, or operational events. It also cannot responsibly claim that every genetic modification has been tested for fertility, health, and longevity before the offspring have had time to live and reproduce.
AI could help shift the odds. It has not been validated as a means to engineer an entire elite working dog in advance. The honest test is prospective: select candidates, raise them, follow them, compare their outcomes with matched conventional candidates, and repeat the process across generations and settings.
How many dogs and generations would it take to validate Warhound’s program?
There is no public scientific answer to that question yet, and Warhound has not disclosed the internal data necessary to calculate one. We do not know its true birth efficiency, donor diversity, selection criteria, starting qualification rate, proposed improvement, or future working-dog numbers. Nevertheless, existing dog and livestock studies allow illustrative estimates of the scale and time involved. These are my research-planning scenarios, not Warhound’s projections, predictions of its success, or universally accepted sample-size requirements.
A program can demonstrate that it produced a cloned puppy long before it has established that the dog is healthy throughout life. It may demonstrate higher qualification rates in one cohort before it has shown that the result holds across different genetic donors. And it may demonstrate that one clone performs well without proving that it can build a sustainable, genetically diverse breeding population.
Validation question | Illustrative scale or period | What it would and would not show |
Can the technology repeatedly produce live, developing puppies? | About 50–150 live-born dogs for an exploratory program | Feasibility and common early problems, not rare harms or elite job performance |
Does the result extend beyond one exceptional donor? | Perhaps 20–50 or more genetically distinct donor lines, with several offspring per line | Preliminary cross-donor repeatability, not a definitive number of breeding founders |
Do more dogs qualify for a specifiedworking assignment? | Roughly 500–2,000 or more dogs in a substantial comparative program, depending on the expected effect | Possible qualification advantage, not automatically lifetime superiority |
Does a conventional or gene-edited breeding system improve over generations? | An initial 2–3 breeding generations, with monitored genetic diversity | Whether traits and costs persist in descendants, not lifetime safety |
Are health, working durability, fertility, and longevity acceptable? | Often 10–15 years or more of follow-up | Long-term outcomes, provided all animals and losses are accounted for |
These ranges should not be combined as though they were separate compulsory groups. The same dog can contribute data to several questions. Conversely, a rare elite qualification outcome or an uncommon disease may require far more animals than the table suggests. A cloning-only program does not have to pass through several sexual breeding generations to assess the first clones; multigenerational testing becomes relevant to inheritable genetic changes and sustained population selection.
What do existing working-dog programs and other mammals tell us?
Chen and colleagues (2021) described a Guiding Eyes for the Blind breeding population of 133 active breeding dogs that produced roughly 520 weaned puppies annually. This was an established program combining selection, puppy raising, records, and long-term feedback, not a special-operations breeding operation. It shows the scale of a mature program, not the minimum number of parents Warhound would need.
In 2026, Riser and colleagues analyzed 17 behavioral traits recorded in 4,841 Labrador Retrievers, with genomic data for 1,076. Adding genomic information improved the average accuracy of breeding-value predictions from about 0.46 to 0.50. Even this substantial dataset produced useful but modest improvements; it did not predict each dog’s complete working career.
Cattle studies provide a different clue. Takeda and colleagues (2021) found that 7,000–11,000 Japanese Black cattle could be sufficient as a reference population for useful genomic predictions of certain carcass traits in that population. Carcass traits are not the same as an elite working dog’s multi-dimensional temperament, and cattle results cannot be turned into a required number of Warhound dogs. They show how much measured data complex genetic prediction may require.
Finally, Sinclair and colleagues (2016) examined 13 cloned sheep aged seven to nine years and found reassuring health outcomes in that small cohort. That tells us healthy aging in clones is possible. It does not establish rare-risk rates, working ability, or normal longevity in all canine clones. The FDA’s livestock assessment likewise identified greater reproductive and early-life risks in some cloned animals, while finding that many apparently normal survivors developed comparably to conventional animals. These are species-specific findings, not direct canine guarantees.
How many dogs would a qualification-rate study need?
Consider four hypothetical improvements. The study must follow clearly defined starting cohorts and judge all dogs by the same final working standard. The estimates below use a simple comparison of two independent groups, equal group sizes, 80% statistical power, and a two-sided 5% significance level. They are rounded, approximate planning figures, not a power analysis of Warhound’s undisclosed data.
Conventional pass rate | New program pass rate | Approximate dogs per group | Approximate total across two groups |
50% | 90% | 25 | 50 |
10% | 20% | 200 | 400 |
5% | 10% | 425 | 850 |
2% | 5% | 560 | 1,120 |
The first result is a huge improvement and is comparatively easy to detect. Doubling a rare qualification outcome is much harder to demonstrate. We do not know the actual Tier 1 qualification rate, so these numbers are examples, not estimates of how many puppies currently qualify for special operations. Multiple comparisons, uncertainty about the baseline, attrition before testing, measurement error, and correlated outcomes would usually increase the number needed.
An especially important problem for cloning is that 500 puppies cloned from one donor represent 500 observed dogs but only one source of donor nuclear genetics. Those outcomes are not 500 independent tests of whether different elite donor genotypes produce exceptional offspring. Repeating the result with several unrelated donors, multiple raising teams, and more than one training cohort is essential. Donor replication and total puppy counts answer different questions.
How many breeding animals and generations would be involved?
We cannot derive a defensible exact number of breeding males, females, egg donors, or surrogate mothers from the available information. A program that clones elite donors is fundamentally different from a self-sustaining sexual breeding population. The provisional 20–50 donor genotypes above is a suggestion for testing whether results generalize beyond a few champions, not a validated minimum or a count of active breeding animals. A sustainable population could require a considerably wider founder base, depending on relatedness, breeding contributions, effective population size, and the amount of new genetic material introduced.
A study of a separate guide-dog breeding colony by Cole, Franke, and Leighton (2004) recorded generation intervals averaging about 23 months and examined the accumulation of genetic relationships and inbreeding. That shows why genealogical and genomic diversity records matter, but not how long every Warhound generation must be. Two or three generations would be a reasonable initial research horizon for observing heritable changes; it would not exclude later disease, reproductive problems, or declining performance.
How much reproduction might be required before the puppies are even born?
Olsson and colleagues (2022) reported an average canine cloning efficiency of about 2.2% live offspring per transferred reconstructed embryo in a large cloning program. This is a historical finding from a particular facility, not a measured Warhound success rate. Simple arithmetic illustrates its possible scale:
Hypothetical number of live-born clones | Transferred reconstructed embryos at a 2.2% yield |
25 | About 1,140 |
100 | About 4,550 |
500 | About 22,730 |
An embryo transfer is not a surrogate mother. Multiple reconstructed embryos may be transferred during a procedure, and the number of egg donors, surrogates, retrievals, pregnancies, interventions, and surviving puppies cannot be calculated from this rate alone. These figures must not be presented as Warhound’s future animal use. They explain why reproductive efficiency and the welfare of all animals involved are part of the cost calculation.
How many years before we know whether the result really worked?
Follow-up window after the first births | Questions that might become answerable | What remains uncertain |
First 0–2 years | Birth and neonatal survival, early growth, preliminary health, behavior during raising | Mature job performance and most late-onset outcomes |
About 2–5 years | Adult training and some field qualification, orthopedic status, early service performance | Full working durability, reproductive outcomes across time, later disease |
About 5–10 years | Repeatability in additional cohorts, working careers, some breeding generations | Full longevity and uncommon conditions |
10–15 years or longer | Lifetime health, retirement, mortality, late-onset disease, and descendants | Very rare risks may still be difficult to quantify |
These are planning windows, not predictions about Warhound’s release schedule. They also overlap: a company could test another donor while continuing to follow the first cohort. A dog might qualify in its third year and develop a career-ending medical condition in its sixth. Whether the technology was truly beneficial depends on what outcome was promised.
What if every tested dog looks healthy?
Imagine observing zero cases of a particular serious problem among the dogs followed long enough to be at risk. Under a simplified independent-observation model, the familiar rule of three gives the following approximate upper 95% bounds on the event rate:
Independent dogs fully observed, with zero events | Approximate upper 95% event rate |
100 | 3% |
300 | 1% |
1,000 | 0.3% |
3,000 | 0.1% |
Zero observed cases in 100 dogs does not demonstrate that the true risk is zero. More importantly, the calculation would not be valid if dogs had not reached the age at which the disease appears, cases were missed, or correlated donor lines reduced the number of effectively independent observations. A dog that looks healthy at 18 months tells us little about its eventual lifespan.
So what is a reasonable provisional estimate?
For a credible claim of repeatable improvement in complex elite working-dog production, I would tentatively plan on dozens of distinct donor genotypes, hundreds to a few thousand dogs in appropriate comparisons, more than one independent training and raising cohort, and years of actual service follow-up. For inherited modifications, I would also plan to examine at least two or three breeding generations and continue long-term health and fertility surveillance. None of those figures should be mistaken for an assurance that the experiment will work, or for an estimate of Warhound’s undisclosed plans.
The larger point is not a particular sample-size number. It is the difference between proving that one clone can be produced, showing that an initial set of clones has a higher qualification rate, and demonstrating a sustainable method that repeatedly produces healthier, exceptional dogs through complete working lives. That last claim requires evidence that cannot be compressed into one puppyhood test or one impressive donor.
What happens to the dogs that do not qualify?
Qualification rates tell us how many dogs reach a particular standard. They do not tell us what happens to the rest. Pfaffenberger’s often-quoted 90% figure leaves roughly 10% outside the reported guide-dog outcome, but the records reviewed here do not identify the subsequent placement or care of every unsuccessful candidate. Warhound should not be evaluated on the successful dogs alone.
Does a working-dog washout make a suitable family pet?
Sometimes. A dog might fail an elite assignment because it lacks a particular search capability, does not meet an exacting physical standard, or is better suited to a less demanding job. It could still become an excellent pet, sport dog, or worker in a different field. But this cannot be assumed for every dog. A candidate rejected for difficult handling, unsafe behavior, or significant medical problems may need specialized care or may be unsuitable for ordinary placement. Even an exceptionally capable working dog is not automatically an appropriate family dog.
That matters for the breeds and working-line crosses Warhound may seek to reproduce. Large, powerful dogs with substantial activity and handling requirements are not interchangeable with companion-bred dogs that happen to share a breed name. A placement plan must match the actual individual, including its behavior, health, history, and the prospective owner’s experience. Calling every rejected dog a “career change” does not, by itself, make the placement responsible.
It is NOT unusual for a working dog washout to be euthanized, being unsuitable for other work and unsuitable as a pet.
Can the American pet market absorb unsuccessful candidates?
Some European working-dog communities offer additional outlets among breeders, sport handlers, professional trainers, police organizations, and highly experienced private owners. But that does not mean every European washout is successfully placed, nor does it prove that an equivalent U.S. network already exists at whatever scale Warhound eventually needs.
The United States already has substantial adoption pressure. Shelter Animals Count estimates that approximately 2.8 million dogs entered U.S. shelters and rescues in 2025 and approximately 320,000 dogs were euthanized. In reporting organizations classified as government shelters, the 2025 median time to adoption was 23 days for large dogs and 11 days for small dogs. These are national shelter figures, not an estimate of demand for Warhound candidates. They establish a real capacity concern, not proof that no suitable American homes exist.
There are also existing U.S. breeders supplying Belgian Malinois, German Shepherds, Dutch Shepherds, and crosses for work, sport, and companionship. Any new production system should assess the genuine demand for its rejected dogs rather than assume their genetic ancestry makes them easy to place. Its obligations should not depend on transferring the problem to publicly funded shelters or private rescues.
How many dogs might need another outcome?
We do not know Warhound’s planned annual output, actual working qualification rate, or the proportion of dogs intended for elite missions rather than broader police, military, or detection roles. The following arithmetic uses a hypothetical birth cohort of 500 dogs and one specified pass/fail standard. Dogs that fail one standard may later qualify for another, so the last column means a further decision is necessary, not that all the dogs must be adopted as pets.
Illustrative qualification rate | Meet stated standard | Need another outcome |
90% | 450 | 50 |
70% | 350 | 150 |
50% | 250 | 250 |
20% | 100 | 400 |
Table: Hypothetical outcomes among 500 live-born puppies. These are illustrative scenarios, not Warhound forecasts, observed Tier 1 qualification rates, or estimates of the number of dogs requiring pet adoption.
Even a 90% pass rate leaves 50 dogs for other roles, placement, or continued care. At a 20% pass rate, 400 dogs require another outcome. A complete program needs a plan and budget for both groups from the beginning. It also needs to account for animals lost before the live-birth stage, which are excluded from this table but remain relevant to cloning efficiency and reproductive welfare.
Has military dog disposition already produced welfare failures?
Yes, and that history offers a direct institutional warning. In a March 2018 report on the Army’s Tactical Explosive Detection Dog program from 2011 to 2014, the U.S. Department of Defense Inspector General documented serious deficiencies in adoption screening and disposition records. The report found that some dogs went to homes without complete suitability reviews; it also described a private organization that accepted 13 dogs and subsequently abandoned them to a kennel. These events concern a previous Army program, not Warhound. The lesson is that even an established military system can fail animals when end-of-program placement has not been planned, funded, and independently checked.
Current U.S. military animal disposition law provides for transfer or adoption subject to suitability assessments and recognized placement priorities. Those protections do not, by themselves, prove that a private company’s unqualified puppies would be covered by the same arrangements. Warhound would need its own enforceable and publicly assessable responsibilities, alongside any applicable governmental requirements.
What has Warhound actually said about retirement and adoption?
Warhound’s October 2026 launch announcement says it will work under guidance from the American Humane Society, including certification of training programs and support for retired military working dogs being reunited with former handlers or placed through adoption protocols. This is a relevant stated commitment. But the announcement does not publish a comprehensive policy for puppies rejected before service, dogs with serious behavioral or medical problems, rehoming failures, or animals requiring long-term institutional care. The gap is in public detail; it is not evidence that no internal plan exists.
Guide Dogs for the Blind provides a useful modern model of the kind of infrastructure required. Its career-change process evaluates dogs for other working placements, offers appropriate dogs to puppy raisers, screens prospective adopters, and permits dogs to return when a placement does not work. Even that example should not be treated as a guarantee that dogs bred for specialized protection work can be placed as readily as dogs bred for guide work.
How should the program account for every animal?
The starting denominator should be specified clearly: embryos created or transferred, pregnancies, live births, dogs entering evaluation, dogs reaching each level of work, and dogs needing another disposition. Each dog should then have a tracked outcome, with medical and behavioral reasons for rejection separated wherever possible. A permanent placement should not be counted as successful solely because a transfer occurred.
Cohort or disposition category | What should be recorded |
Embryos, egg donors and surrogate mothers | Embryo transfers, pregnancies, complications, survival, care, retirement |
Puppies and adolescent candidates | Births, deaths, development, health, evaluations, reasons for withdrawal |
Reassigned working dogs | Role, training and qualification, later performance, final disposition |
Privately placed dogs | Suitability review, adopter experience, retention, returns, incidents, support |
Dogs retained for ongoing care | Number, reason, living conditions, oversight, welfare and lifetime costs |
Dogs dying or euthanized | Age, circumstances, medical and behavioral records, independent review |
Qualified dogs and retirees | Service years, injuries, health, retirement care, adoption and long-term outcome |
Table: Minimum outcome accounting for a proposed working-dog breeding and cloning system. Categories may be sequential rather than mutually exclusive; reports should use clearly defined denominators and follow-up dates.
There is a basic economic problem if lifetime responsibility is omitted. The production cost of a puppy is not the cost of a suitable adult working dog. And the cost per working dog can look artificially low if care for rejected dogs, unsuccessful placements, breeding animals, and reproductive failures is shifted to rescues, adopters, or the public. Responsible accounting should include those costs, as well as long-term care for dogs for whom safe and appropriate placement cannot be found.
A responsible program would have documented placement criteria, independent veterinary and behavioral review where indicated, contractual return provisions, post-placement follow-up, reserve funding for long-term care, and transparent reporting of outcomes. These are proposed evaluation criteria, not claims that Warhound currently lacks every one of them.
A successful breeding program should therefore be judged not only by the dogs that reach exceptional performance, but also by what happens to every dog and reproductive animal that does not. Pfaffenberger’s guide-dog qualification figures do not answer that welfare question. Neither will impressive cloning or elite selection statistics unless the rest of the population is accounted for.
Combining Work With Welfare: What exactly would I want Warhound to prove?
Warhound’s announcement describes ambitions involving cloning, enhanced genetics, training, equipment, domestic supply, and lifelong animal care. Its stated goals are not independently established qualification rates. The company might achieve useful progress for one working role without producing the extreme combination of characteristics required for Tier 1 special operations, and the distinction needs to be reported clearly.
For a fair comparison, I believe the company should publish the denominator at every stage: reconstructed embryos transferred; pregnancies; live-born puppies; early and later survival; healthy dogs entering training; ordinary working-dog qualifiers; candidates meeting specialized operational criteria; and dogs that sustain capable field performance over years. Results should be reported across genetically distinct donors, independent cohorts, and relevant training environments, with standardized operational outcomes and comparison groups.
The complete welfare accounting must include breeding animals, egg donors, surrogate mothers, puppies that wash out, dogs reassigned to other work, privately placed dogs, qualified dogs, and retirees. Selection failures remain animals requiring responsible care. A company should record placement suitability, returns, long-term retention, and welfare outcomes, rather than treating transfer from its premises as the end of responsibility. Where genetic modifications are transmitted to descendants, later generations should also be followed.
The cost calculation should include unsuccessful embryos and candidates, reproduction, veterinary care, raising, training, evaluation, equipment, service life, retirement, and obligations to dogs that cannot be suitably rehomed. Warhound’s announced $200 million is financing, not a per-dog production cost. Comparing cost per puppy with cost per healthy, qualified working dog performing the same assignment would be misleading.
The meaningful test is independent, prospective, and long enough to observe the outcomes being claimed. A good first litter, an encouraging qualification statistic, and a verified lifetime advantage are three different achievements.
What does the Warhound experiment tell us about our own dogs?
The most useful conclusion returns to the two sisters from the same litter. Shared parents do not produce identical puppies. Even deliberately replicated nuclear DNA does not produce an identical developmental history, temperament, or adult working career.
Scott and Fuller showed why heredity and development must be studied together. Pfaffenberger’s guide-dog work illustrated that improving a breeding program was not enough; the raising and testing system mattered enormously. Modern genetic selection and cloning can improve what breeders know and may increase the odds of producing suitable animals. But the complexity of brain development, physical health, experience, handler effects, and lifetime performance does not disappear in a laboratory.
If I were evaluating two littermates with different behavior, I would start by documenting the actual differences, their chronology, their health, the environments where they occur, and the relevant experiences of each dog. I would not jump straight from “temperament” to a confident genetic explanation, or from different homes to a confident environmental explanation.
And if I were evaluating Warhound, I would use the same principle on a larger scale: measure the actual dogs and their outcomes, not just the promise of the genetic method.
We do not get the finished dog at conception. We get an individual whose development has only begun.
I’m not yet a believer, but I’m open to being persuaded. If all this is possible, then it might pave the way for better dogs of all types, from working dogs to pets. We shall see. I also recommend them contacting Armin Winkler, who works with such dogs and engaging him as an advisor.
Glossary
Behavioral phenotype: A specific, measured expression of behavior; its meaning depends on how it is operationally defined and assessed.
Cloning (somatic cell nuclear transfer): Creating an animal with nearly the same nuclear genome as a donor by transferring a nucleus into an enucleated egg. This does not reproduce the donor’s memories or experiences.
Embryo-to-live-birth efficiency: The proportion of transferred reconstructed embryos or oocytes that lead to live offspring; not the same as neonatal mortality or adult survival.
Genomic estimated breeding value: A statistical estimate of genetic merit for a defined trait, useful for population selection but not a guarantee for an individual animal.
Heritability: A population- and context-specific estimate of variation associated with genetic differences, not an individual’s fixed genetic percentage.
Locus (plural loci): A region of the genome associated with a characteristic; an association alone does not identify a causal gene.
Pleiotropy: Influence of a gene on more than one biological characteristic.
Sensitive period: A developmental interval in which particular experiences may have especially important effects.
Working-dog qualification rate: The proportion of a defined starting cohort meeting a particular work standard; the denominator and standard must be stated before rates can be compared.
Career-change dog: A dog no longer being prepared for its original working role and evaluated for another role or suitable placement; the term does not guarantee pet suitability.
Placement retention: Whether a dog remains in an appropriate subsequent home or program over a defined follow-up period, rather than merely being transferred on one date.
Washout: An informal term for a dog removed from a particular selection or training program. It does not by itself explain the reason for removal or the dog’s suitability for other work or a home.
Selected Scholarly and Historical Bibliography
1. Brady, K., Cracknell, N., Zulch, H., & Mills, D. S. (2018). A systematic review of the reliability and validity of behavioural tests used to assess behavioural characteristics important in working dogs. Frontiers in Veterinary Science, 5, 103. https://doi.org/10.3389/fvets.2018.00103
2. Camps, T., Amat, M., Mariotti, V. M., Le Brech, S., & Manteca, X. (2012). Pain-related aggression in dogs: 12 clinical cases. Journal of Veterinary Behavior, 7(2), 99–102. https://doi.org/10.1016/j.jveb.2011.08.002
3. Choi, J., et al. (2014). Behavioral analysis of cloned puppies derived from an elite drug-detection dog. Behavior Genetics, 44(1), 68–76. https://doi.org/10.1007/s10519-013-9620-z
4. Lord, K. A., et al. (2025). Genetic testing predicts appearance but not behavior in dogs. Proceedings of the National Academy of Sciences, 122(48), e2421752122. https://doi.org/10.1073/pnas.2421752122
5. Mikkola, L., et al. (2021). An across-breed validation study of 46 genetic markers in canine hip dysplasia. BMC Genomics, 22, 68. https://doi.org/10.1186/s12864-021-07375-x
6. Oberbauer, A. M., Keller, G. G., & Famula, T. R. (2017). Long-term genetic selection reduced prevalence of hip and elbow dysplasia in 60 dog breeds. PLOS ONE, 12(2), e0172918. https://doi.org/10.1371/journal.pone.0172918
7. Olsson, P. O., et al. (2022). Insights from one thousand cloned dogs. Scientific Reports, 12, 11209. https://doi.org/10.1038/s41598-022-15097-7
8. Pfaffenberger, C. J. (1963). The New Knowledge of Dog Behavior. Howell Book House.
9. Pfaffenberger, C. J., & Scott, J. P. (1959). The relationship between delayed socialization and trainability in guide dogs. The Journal of Genetic Psychology, 95, 145–155. https://doi.org/10.1080/00221325.1959.10534251
10. Riser, M. M., et al. (2026). Genomic information increases prediction accuracy of behavior traits of Labrador Retrievers used as guide dogs. Genetics Selection Evolution, 58, 16. https://doi.org/10.1186/s12711-026-01033-0
11. Scott, J. P., & Fuller, J. L. (1965). Genetics and the Social Behavior of the Dog. University of Chicago Press.
12. Zink, C., & Schlehr, M. R. (2020). Working dog structure: Evaluation and relationship to function. Frontiers in Veterinary Science, 7, 559055. https://doi.org/10.3389/fvets.2020.559055
13. Chen, F. L., Zimmermann, M., Hekman, J. P., et al. (2021). Advancing genetic selection and behavioral genomics of working dogs through collaborative science. Frontiers in Veterinary Science, 8, 662429. https://doi.org/10.3389/fvets.2021.662429
14. Cole, J. B., Franke, D. E., & Leighton, E. A. (2004). Population structure of a colony of dog guides. Journal of Animal Science, 82(10), 2906–2912. https://doi.org/10.2527/2004.82102906x
15. Sinclair, K. D., et al. (2016). Healthy ageing of cloned sheep. Nature Communications, 7, 12359. https://doi.org/10.1038/ncomms12359
16. Takeda, M., et al. (2021). Exploring the size of reference population for expected accuracy of genomic prediction using simulated and real data in Japanese Black cattle. BMC Genomics, 22, 799.https://doi.org/10.1186/s12864-021-08121-z
Practitioner, Institutional, and Contemporary Sources
17. Winkler, A. (2023). Dog Terms Part I: Tell Me About Your Dog. Rivanna K9 Services. https://rivannak9services.com/dog-terms-part-i/
18. Winkler, A. (2023). Dog Terms Part II: Tell Me About Your Dog! Rivanna K9 Services. https://rivannak9services.com/dog-terms-part-ii/
19. Guide Dogs for the Blind. Breeding and History. https://www.guidedogs.com/about-guide-dogs-for-the-blind/dog-departments/breeding ; https://www.guidedogs.com/about-guide-dogs-for-the-blind/meet-guide-dogs-for-the-blind/history
20. Working Dutch Shepherd Association of America. How to Find a Breeder (discussion of KNPV-related BRN ancestry). https://www.wdsaa.org/breeder
21. Naval Special Warfare Command. (August 19, 2026). Solicitation H9224026QE076, multi-purpose canines. https://app.govly.com/public/opportunities/16988908
22. British Small Animal Veterinary Association. (February 2026). Cloning of companion animals: Position statement. https://bsava.com/position-statement/cloning-of-companion-animals/
23. Warhound. (October 23, 2026). Warhound launches with $200 million to rebuild America’s military working dog supply, using advanced genetics and defense technology. PR Newswire (company announcement). https://www.prnewswire.com/news-releases/warhound-launches-with-200-million-to-rebuild-americas-military-working-dog-supply-using-advanced-genetics-and-defense-technology-302918435.html
24. U.S. Food and Drug Administration. (2008). Animal Cloning: Risk Management Plan and Risk Assessment. https://www.fda.gov/animal-veterinary/animal-cloning/risk-management-plan
Guide Dogs for the Blind. Career Change Dogs: Career Change Process, Strategic and Community Placements, and Dog Placement Program. https://www.guidedogs.com/puppy-raising-guidebook/beyond-puppy-raising/career-change-dogs
Guide Dogs for the Blind. Dog Adoption: Career Change Dogs. https://www.guidedogs.com/about-guide-dogs-for-the-blind/dog-departments/career-change-dogs/dog-adoption
U.S. Department of Defense Inspector General. (2018). The Army’s Tactical Explosive Detection Dog Disposition Process from 2011 to 2014 (Report No. DODIG-2018-081). https://media.defense.gov/2018/Mar/01/2001884674/-1/-1/1/DODIG-2018-081_FINAL_REDACTED.PDF
United States Code. 10 U.S.C. § 2583, Military animals: transfer and adoption (2026). https://uscode.house.gov/us/usc/119-121/t10/s2583/
No comments:
Post a Comment