We Built the Hardware, Then We Found the Literature.
We built the hardware, then we found the literature, then we formulated the category language, then we developed the education material.
A living evidence review of chewing, agency, feeding mechanics and the modern domestic dog
First published: 8 September 2026 by, Team Petelli
This is a living article. We will update it whenever relevant research appears, including evidence that weakens, contradicts or changes our interpretation. If we get something wrong, we will correct it, and if a better explanation emerges, we will use it. Structural Feeding™ is a developing framework rather than a finished scientific doctrine.
We aren't asking what your dog eats, but asking what your dog gets to do.
BoneZone® did not begin with an academic paper; it began with the ordinary problem of giving a powerful dog a substantial bone or chew and then watching the dog decide where that object would go. It might be dragged across the yard, into dirt, over paving, onto grass, beneath something or against a wall, depending on where the dog could create enough leverage to work it. From there, the engineering question was simple: what changes if the food stays put?
Years of prototypes followed, producing immediate and practical benefits such as control over location, easier supervision, a defined cleaning area, adjustable positioning and stable resistance against which the dog could grip and pull. We eventually called the wider idea Structural Feeding™ because the apparatus was doing something that a bowl generally does not need to do: it was becoming part of the physical feeding event. Only much later did we seriously ask whether anybody had studied the larger question behind that idea, and this is where the story became far more interesting.
We have not found a paper that proves BoneZone® works
We should begin with the evidence that does not exist. We have found no controlled trial showing that BoneZone® improves welfare, no evidence that secured feeding is universally better than loose feeding, no evidence that bowls as hardware are harmful, and no proof that every dog prefers an effortful feeding experience. We will not pretend otherwise.
What we have found is a substantial and rapidly developing body of research around nearly every component of the question. That work examines the functional significance of chewing, food-acquisition behaviour, canine agency, choice, control, challenge and competence, feeding enrichment, owner barriers to providing chew opportunities, the effect of physical food presentation, dental and gastrointestinal risk, hygiene and microbiological risk, and the distinction between nutritional provision and behavioural opportunity. None of those papers was written about BoneZone®, which is precisely why they matter: rather than inventing a welfare problem to justify a product, we appear to have engineered our way independently into a legitimate research question that science was already assembling piece by piece.
Nutrition is not the only welfare domain
One of the most useful discoveries was the Five Domains Model of animal welfare, whose current structure considers Nutrition, Physical Environment, Health, Behavioural Interactions and Mental State as separate but connected domains. [1]. That framework immediately exposes a false choice that appears repeatedly in feeding arguments, because a food can be nutritionally complete while the way it is presented remains a separate behavioural question. There is no contradiction in saying that a bowl can deliver excellent nutrition without automatically recreating every physical component of feeding, nor does the bowl need to do so.
Structural Feeding™ asks a different question: what becomes possible when the presentation of food is treated as another variable? Nutrition tells us what enters the dog, while behavioural science also asks what the dog is able to do.
Then came the chewing literature
In 2025, researchers from the University of Sydney published a major review titled Functional significance and welfare implications of chewing in dogs. Rather than treating chewing as an inconvenient habit or merely something dogs happen to do with toys, the authors examined its place within canine food acquisition, ingestive behaviour, self-care and social interaction. The review considers dismemberment of tissue, access to nourishment, bite force, chewing rate, dental implications, stress, psychological function and telos, meaning the characteristic nature and behavioural purpose of the animal. It concludes that chewing appears to contribute positively to canine physical and psychological welfare, while remaining explicit that any benefit must be weighed against genuine risk. [2].
That balance is almost exactly where our own work has led. We do not need to claim that every dog should eat bones, that hard bones are safe or that every meal should become a physical challenge. The more important finding is also the more defensible one: chewing and food manipulation are legitimate behavioural subjects in their own right.
The whole-food argument was already decades old
The 2025 review did not create the idea that the physical form of food matters. Beginning in 1991, Australian veterinarian Dr Tom Lonsdale developed a public professional record around oral disease, food form, raw meaty bones and whole-carcass feeding, with that work continuing through veterinary proceedings, professional articles and later working-dog publications over several decades. [3] [4]. The significance for this discussion is straightforward: whole-food feeding, and the physical work involved in eating it, were subjects of professional discussion long before BoneZone® existed.
In a 2011 article for working-dog handlers, Lonsdale described dogs ripping, tearing and gnawing at sufficiently large meaty portions, while using their paws to resist and position the food. He also noted the familiar domestic outcome in which food is dragged across the floor or carried elsewhere, recommending outdoor feeding or an easy-to-clean floor. [5]. Read mechanically, that account reveals the same practical problem from which BoneZone® emerged: the dog needs resistance, yet the owner needs some control over where the feeding event takes place.
That is where the work aligns. The biological behaviour and whole-food practice were not new; the unresolved household question was how to provide stable counterforce while improving control over location, movement, working height, supervision, removal and cleanup. BoneZone® developed from that engineering problem before we encountered Lonsdale's archive, and Structural Feeding™ became our language for the broader practice: presenting suitable substantial foods, including appropriately selected raw meaty bones and meaty sections, against a stable external structure so the dog can grip, pull, tear and manipulate them within a managed environment.
Bowls still have a place. This is an additional feeding format, not an instruction to throw them away. The important connection is that BoneZone® gives modern owners a practical way to support part of a much older whole-food practice without surrendering the control that often determines whether they will attempt it at all.
And owners are already limiting those opportunities for practical reasons
The same research group followed with a 2026 cross-sectional survey of how owners provide chewing opportunities. Across 655 responses, reported chewing time varied considerably; older and smaller dogs were associated with lower chewing rates, bone provision was reduced among dogs without outdoor access, and practical concerns such as mess, smell, dental risk and professional advice influenced what owners were willing to provide. [6].
Those findings connect directly with the problem from which BoneZone® emerged. A behavioural opportunity may have genuine value in theory and still disappear from domestic life when it is messy, difficult to supervise or incompatible with the home. Human convenience is therefore not automatically the enemy of animal welfare; sometimes more effective management is precisely what allows the opportunity to exist.
The feeding interface is already being studied
A University of Liverpool survey, memorably titled 'Bowls are boring', asked 1,750 dog owners about canine enrichment feeding. Kongs, chews and activity toys were among the most common approaches, with owners frequently perceiving benefits to mental stimulation while also identifying lack of time and practical management as barriers. Because this was a survey, it cannot establish that enrichment feeding caused the benefits owners reported, and the authors acknowledge that limitation. What it does establish is that researchers already regard the method by which food is delivered as a variable worth examining. [7].
A small experimental study published in 2023 compared a stainless-steel bowl with feeding toys and found that the four participating dogs spent longer eating and interacting with the toys, alongside differences in activity and physiological measures. Four dogs cannot support sweeping conclusions, but the narrower point remains sound: changing the presentation can change the feeding event. [8].
Physical food form changes behaviour in other canids too
A 2025 study conducted at Taronga Western Plains Zoo makes that point particularly difficult to ignore. Researchers observed a pack of nine African wild dogs receiving food in three physical forms: individual pieces, butchered carcasses and whole carcasses. Nutrition remained part of the same broad feeding event, yet the change in presentation altered what the animals did; the pack spent significantly longer feeding from carcasses than from individual pieces, and patterns of social interaction changed as well. [9].
African wild dogs are not domestic dogs, and a carcass is not BoneZone®, so no responsible argument can leap directly from that study to our apparatus. It does, however, support a narrower proposition with considerable force:
Food presentation itself can be behaviourally consequential.
Once feeding is viewed mechanically, that proposition is hardly surprising. A piece of meat lying loose on the ground, the same tissue attached to a carcass, the same material hidden inside a puzzle and food served in a bowl do not create the same physical problem for the animal, even if their nutrient totals could somehow be made identical.
This is where agency becomes important
Although agency can sound abstract, welfare science is giving the idea a more practical form. Littlewood, Heslop and Cobb describe animal agency through voluntary, self-generated and goal-directed behaviour, while identifying choice, control, challenge and competence as important characteristics of positive welfare. Their qualification is equally important: a challenge must suit the individual, because one that exceeds the animal's competence may create frustration rather than enrichment. [10].
That changes how Structural Feeding™ should be understood. The objective is not to make a dog work harder on the assumption that effort must always be good; it is to create a physical opportunity with which the dog can choose to engage. This raises an apparent contradiction that deserves a direct answer: if BoneZone® restrains the food, does restraint reduce agency? In one dimension, it does. The dog cannot carry the complete item elsewhere, bury it or relocate the entire feeding event across the property, but agency is not a single switch that is either completely on, or completely off.
Within the controlled location, the dog can still decide whether to engage, where to bite, how to orient its body, how hard and in which direction to pull, which tissue to target, when to stop, when to return and how to solve the physical problem presented by the food. That leads to a more useful design principle than simply calling BoneZone® "natural":
Control the environment, not every behaviour within it.
Flourishing is not the same as simply removing every challenge
A 2020 conceptual analysis by Benz-Schwarzburg, Monsó and Huber approached dogs through cognition and ethics rather than feeding. The authors argue that human responsibility extends beyond satisfying basic needs and should include opportunities for dogs to develop their capabilities and flourish as the type of beings they are, connecting a meaningful animal life with the exercise of agency. [11]. This does not validate Structural Feeding™, but it helps define the philosophical territory in which our work now sits. An environment containing food, shelter and protection from harm is not automatically equivalent to a rich life that also provides appropriate opportunities to interact, solve, manipulate and choose.
That distinction matters because the engineering problem has always required compromise. Nobody can reasonably place an unrestrained carcass on the floor of a modern home and declare the welfare question solved, but neither should we assume that efficient nutrient delivery makes every other part of the feeding event irrelevant. Between those extremes lies the ground Structural Feeding™ is trying to explore.
What about the research saying dogs prefer easy food?
This evidence belongs in the discussion. A 2024 Scientific Reports study tested contrafreeloading in pet dogs, asking whether they would work for food when identical food was freely available. Thirty-eight dogs could choose between food in an accessible tray and the same food in a snuffle mat that required effort; as a group, they showed no preference for the effortful option, although they were willing to use it. [12].
Rather than weakening our position, that result prevents us from making a bad claim. The evidence does not support saying that dogs naturally want to work for all of their food, and opportunity should never become compulsion. It also leaves an experimental distinction that deserves attention, because searching through a snuffle mat to obtain pieces of food is not necessarily the same behavioural task as using the mouth and body to work the physical food itself by gripping, pulling, tearing and chewing. Those questions require separate tests.
Opportunity should not become compulsion.
And not every enrichment idea produces measurable benefits
Another useful study, published in 2023, gave companion dogs toys over a short period but found no statistically significant improvement in the measured outcomes for activity, eating rate or cognitive bias. [13]. That result belongs here because, if Structural Feeding™ eventually produces a measurable welfare benefit, the finding should come from testing the actual mechanism rather than attaching the word enrichment to a product, and assuming the science arrives automatically.
Then there is the ground
The ground appears simple until it is examined as part of the feeding environment. An Australian study published in 2022 collected 1,581 canine faecal samples across 190 urban parks and found that 44.2% of the sampled parks were contaminated with at least one canine soil-transmitted helminth species. [14]. That does not mean a dog will become infected by eating from the ground, nor does it prove that BoneZone® prevents parasite exposure. We make neither claim. The finding supports something much narrower: ground that looks clean is not necessarily biologically clean.
Raw animal products add another set of microbiological considerations. Multiple studies have associated raw feeding with increased faecal carriage of antimicrobial-resistant bacteria, while veterinary organisations appropriately emphasise careful handling and hygiene. [15] [16] [17]. There can be no universal answer when somebody asks, "Why not just throw it on the ground?" because location, surface, residue, children, other animals, indoor access, supervision and cleanup all affect the situation.
Structural Feeding™ neither sterilises food nor makes biological risk disappear. What it provides is more practical: a defined feeding location that the owner can manage deliberately. Hygiene management, not a hygiene guarantee.
Bones are not risk-free either
The risks associated with bones must remain just as visible as any possible benefit. Studies involving small groups of Beagles have shown that chewing bovine bones can rapidly reduce supragingival dental calculus, while other veterinary evidence documents gingival injury, retained fragments, fractured teeth and clinically significant bone foreign bodies. [18] [2]. AAHA warns against very hard chews, including bones, because of the risk of tooth fracture, and WSAVA highlights microbiological and nutritional concerns connected with raw meat-based feeding. [19] [17].
None of that destroys the behavioural significance of chewing, although it should destroy lazy marketing. The responsible conclusion is neither that bones are safe nor that every challenging chew should be removed in an attempt to eliminate all chewing risk. It is that item selection, supervision, sizing, individual assessment and residual risk all matter. Structural Feeding™ was never designed to make chewing risk-free; it was designed to give the owner more control over the environment in which chewing occurs.
The strangest thing is the experiment we still cannot find
After following each of these research threads, one gap remains conspicuous: we have not found a controlled canine study that asks the simplest question raised by BoneZone®.
What changes when the same substantial food is loose versus externally stabilised?
A useful experiment would keep the dog, food and environment the same while changing one variable: external resistance. It could then measure how stabilisation affects paw use, pulling, feeding duration, bite location, body position, disengagement, residue spread and cleanup, while testing whether restraint frustrates some dogs, benefits others or produces different effects at different heights. Most importantly, where genuine choice is available, it could ask what the dog chooses.
The result might show that we are partly right, entirely wrong or missing something more interesting than our current assumptions, and that is exactly why we want the experiment done.
So what have we actually discovered?
Science has not endorsed BoneZone®, but we have found something more valuable than a borrowed endorsement. We built hardware to solve a physical problem before we possessed language for the wider problem space, then discovered researchers independently examining chewing, food acquisition, feeding enrichment, food presentation, agency, choice, control, challenge, competence, owner barriers, hygiene and risk. Across that literature sits a developing idea that good animal welfare may be defined not only by the harms we prevent, but also by the meaningful opportunities we make available.
Structural Feeding™ is our attempt to connect those ideas around feeding mechanics. The connection may ultimately prove less important than we think, become an obvious extension of contemporary canine-welfare thinking or reveal external stabilisation as a significant variable that nobody had isolated. We do not yet know which answer is correct, but we now know the question is not imaginary, and that is enough reason to continue asking it.
The living evidence commitment
This article will change as the evidence changes. We will add relevant studies, show findings that contradict something written here and revise our conclusions accordingly; if direct Structural Feeding™ research is eventually conducted, whether its results are positive or negative, it belongs here as well. The working evidence register behind this article contains peer-reviewed studies, welfare frameworks, counter-evidence, professional guidance and our current claim boundaries.
We are not asking science to ratify our certainty. We are asking it to attack our uncertainty. If the idea survives that process, it will become far more valuable than anything marketing could manufacture.
References and further reading
- Mellor, D. J. and colleagues (2020). The 2020 Five Domains Model: Including Human-Animal Interactions in Assessments of Animal Welfare. DOI.
- Quinn, R. and colleagues (2025). Functional significance and welfare implications of chewing in dogs (Canis familiaris). Frontiers in Veterinary Science.
- Lonsdale, T. Papers and Articles. Raw Meaty Bones archive, including professional publications from 1991 onward. Archive.
- Lonsdale, T. (1991). Oral Disease in Cats and Dogs. University of Sydney Post Graduate Committee in Veterinary Science, Control and Therapy Series No. 3128. Publication.
- Lonsdale, T. (2011). Raw Meaty Bones: Essential Food for Working K9s. K-9 Cop Magazine, September/October 2011. PDF.
- Quinn, R. and colleagues (2026). Owner-reported influences on dogs' opportunities to chew. Applied Animal Behaviour Science.
- Heys, M., Lloyd, I., & Westgarth, C. (2024). 'Bowls are boring': Investigating enrichment feeding for pet dogs and the perceived benefits and challenges. Veterinary Record.
- Boonhoh, W. and colleagues (2023). Effect of feeding toy and the presence of a dog owner during the feeding time on dog welfare. Veterinary World.
- Jordan, N. R. and colleagues (2025). Effect of Carcass Feeds on Feeding Behavior and Social Interactions in Zoo-Based African Wild Dogs (Lycaon pictus). Zoo Biology.
- Littlewood, K. E., Heslop, M. V., & Cobb, M. L. (2023). The agency domain and behavioral interactions: assessing positive animal welfare using the Five Domains Model. Frontiers in Veterinary Science.
- Benz-Schwarzburg, J., Monsó, S., & Huber, L. (2020). How Dogs Perceive Humans and How Humans Should Treat Their Pet Dogs: Linking Cognition With Ethics. Frontiers in Psychology.
- Rothkoff, L., Feng, L. C., & Byosiere, S.-E. (2024). Domestic pet dogs do not show a preference to contrafreeload, but are willing. Scientific Reports.
- Chan, K., Arellano, C., & Horowitz, A. (2023). Here Puppy, Chew on This: Short-Term Provision of Toys Does Not Improve Welfare in Companion Dogs. Animals.
- Massetti, L. and colleagues (2022). Faecal prevalence, distribution and risk factors associated with canine soil-transmitted helminths contaminating urban parks across Australia. International Journal for Parasitology.
- Groat, E. F. and colleagues (2022). UK dogs eating raw meat diets have higher risk of Salmonella and antimicrobial-resistant Escherichia coli faecal carriage. Journal of Small Animal Practice.
- Mounsey, O. J. and colleagues (2022). Evidence that faecal carriage of resistant Escherichia coli by 16-week-old dogs in the United Kingdom is associated with raw feeding. One Health.
- World Small Animal Veterinary Association. Raw Meat Based Diets for Pets. WSAVA Global Nutrition Toolkit PDF.
- Pinto, C. F. D. and colleagues (2020). Evaluation of teeth injuries in Beagle dogs caused by autoclaved beef bones used as a chewing item to remove dental calculus. PLOS ONE.
- American Animal Hospital Association. (DON'T) chew on this. AAHA.
Petelli supporting material
BoneZone® Structural Feeding™ Canine Enrichment Apparatus & Chewing Station: Product page.
For the full explanation of Structural Feeding™, including the principles, terminology and thinking behind the concept:
and.....
For the deeper research, educational material, feeding mechanics, evidence, articles and developing knowledge base behind the concept:
and....
BoneZone® gives Structural Feeding™ a physical structure to happen against.
More from BoneZone Bone Science
Bones for Dogs: Why Structure, Resistance and Chewing Matter More Than We Thought
A turkey thigh, a poultry frame, a meaty rib section and the dense shaft of a bovine femur may al...
Read Article →
To the Scientific Community
Modern housing changed the setting in which dogs eat, but it did not erase the biological apparat...
Read Article →
BONEZONE® Bone Science Blog Series - Ground Exposure and Controlled Chewing - Pt 6 of 6
Research into food-based enrichment and long-lasting chews suggests that these items can help eng...
Read Article →
Comments