Anyone who lives with dogs is aware that dogs are almost universally attracted to meaty foods and treats. Trainers use these preferences to select different levels of “treat value” for dogs and almost invariably, the treats that are of highest value to a dog are those that have a meaty texture, smell, and (we assume) taste.

It is also true that most dogs are highly attracted to and readily consume high-protein diets that include cooked, extruded, or raw meat of various types.

Are these preferences a vestige of the dog’s predatory past? If so, are such preferences something that dogs are born with, or is there a strong influence of learning and environment on our dogs’ apparent taste for meat?

A recent set of experiments conducted by researchers who study free-ranging dogs in India asked these questions and provide us with some new information.

Street Dog Diet Studies

Free-ranging dogs exist in numerous countries around the world, including Mexico, Italy, Nepal, Japan, many African countries, and India. They survive almost entirely by scavenging and occasionally augment their diet by begging and hunting small animals.

In India, the history of free-ranging dogs is well-documented, extending back to the 9th century BC and representing more than 1,000 generations of dogs.

Indian free-ranging dogs consume a diet that is rich in carbohydrate (biscuits, bread, and rice) and relatively low in protein. The protein that is consumed is in the form of scraps of meat or fish adhering to bones, decomposing meat, and carcass remains.

Domestic dogs are better adapted to scavenging and a diet that is higher in carbohydrate foods than were their wolf-like ancestors because of changes in foraging behavior (increased scavenging/decreased pack hunting) and an enhanced ability to digest starch (increased copies of the gene AMY2B, the gene that codes for pancreatic amylase; for more about this, see “Dogs and Carbs: It’s Complicated“).

However, just because dogs can consume and digest diets that contain a high proportion of carbohydrate (starches), it does not necessarily follow that they prefer such diets or that it is the healthiest or best way to feed them.

Although the study of these free-ranging dogs could shed light on many interesting questions, the two that the Indian researchers attempted to answer were: “Do dogs have a strong preference for meat in their diet?”  and, “If so, is such a preference innate (are puppies are born with this preference) or is it reliant upon or strongly influenced by learning?” 

In the first study, the researchers offered 30 free-ranging dogs a variety of food choices in four separate experiments:

• In the first, dogs chose between bread, bread soaked in water, and bread soaked in chicken broth.

• The dogs selected between bread, bread soaked in gravy, and cooked chicken in the second experiment.

• The third offered the dogs choices between dry dog kibble or bread soaked in varying concentrations of chicken broth.

• The final experiment offered the dogs varying combinations of bread and dog food kibble, soaked with different concentrations of chicken broth.

The purpose of this final set of choices was to separate the factors of meat smell from nutrient (protein) content, because dogs have been previously shown to be capable of self-selecting a diet according to its macronutrient (protein/fat/carbohydrate) content.

Street Dog Study Results

The following preferences were found in the adult, free-ranging dogs:

Meat (smell) beats carbs. The dogs consistently chose bread soaked in chicken broth over dry bread or bread soaked in water, even though chicken broth contains only a small amount of actual protein. When allowed to choose only visually, they selected chicken meat over chicken-soaked bread or dry bread.

Smell beats all. When the dogs were offered kibble (high protein food) or bread (low protein food) soaked with varying concentrations of chicken broth, they consumed all of the foods equally, showing no absolute preference in terms of the quantity that was consumed.

However, the order of selection depended completely upon how much chicken broth was soaking the food, regardless of its nutrient content. In other words, the dogs chose according to smell, not in accordance with the actual amount of meat protein present in the food.

Rule of thumb: Choose the food that smells the most intensely of meat first. The cumulative results of the four experiments support the existence of the above rule of thumb for food choice. This means that the dogs preferred foods that smelled of meat (but that were not necessarily good sources of protein) over those that smelled less meaty, even when the less meaty-smelling foods actually contained more meat ingredients and a higher protein content.

This of course, makes sense, since in nature, a stronger meat smell is highly correlated with high meat and protein content and invariably predicts higher meat quantity. This relationship only becomes skewed when clever experimenters enter the picture and mess with it.

The authors concluded that while domestic dogs have adapted a scavenging lifestyle, they appear to have done so without giving up a strong preference for meat. They suggest that while the domestic dog has indeed evolved to more efficiently digest carbohydrate and exist on a carbohydrate-rich scavenged diet, they continue to be strongly attracted to the smell of meat and preferentially select meaty-smelling foods. (Not surprising at all to most dog owners; but again, good to have science backing up experiences and beliefs).

But wait, they’re not finished. The same researchers then asked, “Are domestic dogs born with a preference for meat, or is it a learned trait?”

The researchers conducted the same series of the experiments described above with the puppies of free-ranging dogs. The puppies were 8 to 10 weeks of age at the time of testing. Using a clever design, they found that:

Puppies do not discriminate. Unlike the adult dogs, puppies near weaning age showed no clear preference for foods that smelled strongly of meat and chose each food selection equally, regardless of how intensely the food smelled of meat.

Dogs use a sniff and snatch strategy. While the adult dogs tended to first smell and inspect all available food choices before choosing and consuming one, puppies did not show this behavior. Rather, they would smell a food, eat it, and then move to the next food, showing little to no preference. The vast majority (89 percent) of choices made by puppies followed this behavior pattern.

The authors speculate that because puppies consume a protein-rich diet in the form of their mother’s milk, there is little selective pressure for an innate selection bias toward the smell of meat.

It is only after weaning, when pups begin to scavenge, that preferentially selecting foods that smell like meat (and are correlated with a high protein content) becomes important. They suggest that, as has been shown in a number of other species, puppies learn their food selection preferences from the mother (i.e., cultural transmission of knowledge) and then as they mature and begin to scavenge, operantly.

Take-Away for People with Pet Dogs

The first study’s results with adult, free-ranging dogs tell us that the dogs in this set of experiments were selecting foods based primarily on smell rather than an ability to discern actual meat content. The adult dogs were operating under the (pretty efficient) rule of “If it smells like meat, eat it.” (We all know and love dogs who do this!)

This strategy is probably strongly selected for in environments where resources are limited, there are few energy- and protein-dense foods available, and competition between dogs is high. This is not really a surprising result – except for the fact that the authors found that the scent of meat was more important than the food’s actual meat (or protein) content.

Newly weaned puppies, in contrast, lack this choice bias and appear to learn to choose “meaty” foods after weaning, either from the food choices of their mother, operantly, or most likely, a combination of the two.

So, what does this tell us about feeding our own dogs? These results suggest that while dogs are predisposed to enjoy the taste of meat ingredients and clearly prefer these foods, puppies do not appear to be born with an attraction to the smell of meat per se; these preferences are influenced by learning early in life.

On a practical level, these data, along with those of earlier studies of taste preferences in dogs, tell us that the foods that are offered to a puppy at a young age should be expected to strongly influence the pup’s food and taste preferences as an adult dog.

An Advocate for “Streeties” in Mumbai 

The photos we used to accompany this article were taken by Aditi Joshi, a street-dog advocate who lives in Mumbai, India, a city that is home to an estimated 250,000 street dogs. Several government agencies and NGOs are involved in vaccination and sterilization programs, as well as treating severely injured and ill street dogs, but they do not address everyday caregiving. There are adoption programs aimed at finding homes for these dogs, but the vast majority of dogs in the city continue to live as street dogs.

By last count, Aditi’s neighborhood is home to 76 “streeties,” as the dogs are affectionately called. Over the years, Aditi has worked to build a culture of stewardship that views the streeties as companions for which the community shares responsibility, thus helping to bridge the gap in the care they receive. She has built a loose network of people in her neighborhood – residents, shop owners, street vendors, domestic workers, garbage collectors, delivery personnel, dog walkers, and doctors – who assist in caring for these resident dogs in varying capacities. Through this network, she coordinates preventative health care (vaccinations and sterilization), provision of shelter and food, grooming, and emergency care in the case of injury or illness, both on site and in her home.

“When cared for, these healthy, settled dogs are the best buffer for minimizing human-animal conflicts and maintaining public health,” Aditi says. “This is also integral to keeping pet dogs in the community healthy, disease-free, and safe.”

Aditi’s biggest challenge is building positive human-animal relationships in a context where many people still fear and misunderstand street animals. Her current goal is to create a sustainable and scalable program to teach people about safely interacting with dogs, and to teach street dogs skills that would smooth their integration into human environments.

To help achieve that goal, Aditi is currently enrolled in the prestigious Academy for Dog Trainers (founded and led by Jean Donaldson) and is working on the development of curriculum for her street-dog education program. If you’d like to see a video about how Aditi trains street dogs, or to help support Aditi’s work for street dogs in Mumbai, see her GoFundMe page.

Cited Studies

1. Bhadra A, Bhattacharjee D, Paul M and Ghadra A. “The meat of the matter: A thumb rule for scavenging dogs.” Ethology, Ecology and Evolution 2016; 28:427-440.
2. Bhadra A and Bhadra A. “Preference for meat is not innate in dogs.” Journal of Ethology 2014; 32:15-22.
3. Hewson-Hughes AK, Hewson-Hughes VL, Colyer A, Miller AT, McGrane SJ, et al. “Geometric analysis of macronutrient selection in breeds of the domestic dog, Canis lupus familiaris.” Behavioral Ecology 2012; 24:293-304.
4. Roberts MT, Bermingham EN, Cave NJ, Young W, McKenzie CM and Thomas DG. “Macronutrient intake of dogs, self-selecting diets varying in composition offered ad libitum.” Journal of Animal Physiology and Nutrition; 102:568-575.

Linda P. Case is the owner of AutumnGold Consulting & Dog Training Center in Mahomet, Illinois. Linda is the author of Dog Food Logic, has a new book, Dog Smart, and writes The Science Dog blog.

The question of how best to feed dogs stimulates great debate and evokes strong emotions among dog folks. (Yes, this an intended understatement.) One of the most contentiously defended viewpoints in recent years is that dogs should not be fed diets that contain digestible carbohydrate (starch).

Two primary arguments are used to defend this position. These are:

A) Dogs are carnivores and have no dietary requirement for carbohydrate.

B) Dogs are unable to efficiently digest starch. Therefore, including starch-providing ingredients in dog foods is unhealthy and provides no nutritional value.

Like many persistent beliefs, there is some truth and some falsehood in both of these blanket claims. Let’s start with the first.

PROPOSED: Dogs are carnivores and have no dietary requirement for carbohydrates.

The first bit is false; the second bit is true. Dogs are classified within the taxonomic order of Carnivora, but like many other species within this order, dogs are omnivorous.

© Tommy Maenhout | Dreamstime.com

The term omnivore simply means that an animal consumes foods that are of animal and plant origin (dogs do this) and can derive essential nutrients from both animal and plant foods (ditto). Based upon this definition, animal nutritionists consider the dog to be an omnivore. By contrast, the domestic cat, along with other felid species, is classified as an obligate carnivore. This classification means that cats cannot derive all of their essential nutrient needs from plant foods and therefore have an obligate need for foods of animal origin in their diet.

The fact that dogs are omnivorous does not signify that they are not predatory (they are), nor that they do not seek out and enjoy eating meat (they do). All that it means is that dogs can consume and derive nutrients from both animal and plant matter.

If we consider the dog’s feeding behavior, it is clear that the majority of dogs enjoy and probably prefer to consume meat in their diet. However, they also scavenge and ingest a wide variety of food types, including starch-containing foods. Nutritionally, just like bears, who also preferably seek out animal source proteins, dogs are omnivores.

Personally, I am baffled as to why “omnivore” has become a fighting word among dog people. This label does not turn the dog into a carrot-munching, Birkenstock-wearing, canine hippie. Rather, it simply describes what the dog eats and is capable of deriving nutrients from: meat and plant matter. That’s all. Time for us all to calm down about this one.

Let’s move on.

So what about carbs? It’s true that dogs have no dietary requirement for carbohydrates; they can derive all the nutrients they need from protein and fat.

That said, cooked starch can provide a highly digestible energy source to dogs when included in their diet. From a nutrition standpoint, dietary carbohydrate spares protein. This means that when a body uses carbohydrate to provide needed energy, dietary protein is conserved from being used for this purpose and continues to be available for use to provide essential amino acids, build and repair body tissues, and support a healthy immune system. Therefore, including at least some digestible carbohydrate in the diet of dogs is generally considered to be beneficial.

The controversy about starch in dog foods revolves more around how much starch is in the food and the source of that starch, rather than its absolute presence or absence. Dogs can thrive on low-carbohydrate diets provided such diets are balanced and contain all of the essential nutrients. Diets formulated in this way are often highly palatable because of their high proportions of protein and fat. These foods are also generally very energy dense (lots of calories packed into a small volume of food), which means that portion control is important to ensure that dogs maintain a healthy weight.

Now for the second persistent statement that is often made about the dog’s nutritional requirements:

PROPOSED: Dogs cannot digest starch.

It should be obvious from the vast majority of dogs who survive on carb-heavy kibbled diets that this is unequivocally false. Dogs efficiently digest cooked starch, just like humans. However, they cannot digest raw starch (and neither can we).

Cooking results in the expansion of the small granules that make up starch, which allows digestive enzymes better access and increases digestibility. This is true for humans as well as for dogs, and this fact explains why we generally do not munch on raw potatoes.

We actually know the exact degree to which cooking increases digestibility of various starches. Ground grains such as rice, oats, or corn are about 60 percent digestible when fed raw to dogs. Cooking these ingredients increases the dog’s ability to digest them to almost 100 percent! This means that if you feed your dog 100 grams of uncooked oats or rice, only 60 grams will make it into his body to nourish him; 40 grams ends up in the large intestine where microbes ferment some of it, and a lot of that 40 grams ends up in your yard, as feces.

Conversely, when cooked, almost the entire 100 grams are digested and absorbed to nourish your dog. Again, not to put too fine a point on this, but the same holds true for humans.

The AMY2B Enzyme

Like humans, dogs have an enhanced ability to digest starch-containing ingredients, a change that has been directly tied to domestication. In 2013, a ground-breaking paper by Erik Axelsson of Uppsala University in Sweden identified a host of genetic changes that occurred as dogs evolved from their wolf ancestors.1 Three of these changes were alterations of key genes that code for enzymes involved in starch digestion, most notably and consistently, one labeled AMY2B.

This gene codes for the production of pancreatic amylase, an enzyme that functions to digest dietary starch. Although variation exists among individual dogs and breeds of differing geographic origin, the increased copies of the AMY2B gene correlate with higher levels of circulating pancreatic amylase in a dog’s blood, which means that higher AMY2B leads to more efficient starch digestion.2, 3, 4

On average, dogs have a sevenfold higher copy number of this gene when compared with present-day wolves. These changes in the dog’s genetic makeup coincide with the expansion of human agricultural practices and increased reliance upon starch-providing plants in both human and dog diets.

Dog Food Diet Selection

It is a fact that domestic dogs are better adapted to scavenging and to a diet that is higher in starch-containing foods than were their wolf-like ancestors. However, just because dogs can consume and digest starch, it does not necessarily follow that a diet that contains a high proportion of digestible carbohydrate is the healthiest way to feed them. One way of determining how much protein, fat, and carbohydrate dogs should have is to ask the dogs directly.

© Yevheniia Sednieva | Dreamstime.com

Historically, nutritionists have viewed diet selection in animals principally from the standpoint of energy balance. The basic assumption was that all animals, including dogs, eat to meet their energy (caloric) needs first. However, in recent years this premise has been challenged.

There is evidence that a wide range of species, including many birds, fish, and mammals, will self-select diets containing consistent proportions of the three major macronutrients -protein, fat, and carbohydrate – and that they regulate and balance their nutrient intake to maximize lifespan and reproductive fitness.

The recognition that macronutrient selection can be a driver for appropriate diet selection has led to several new studies with dogs and cats.

Domestic cats were studied before dogs and were found to consistently select a diet that was high in protein and fat and low in carbohydrate.5 This profile is consistent with that of other obligate carnivores and with the cat’s wild feline cousins. Interestingly, a recent study found that cats preferentially balanced their diets to a set protein:fat ratio, even when offered foods of different flavor preferences and containing animal- or plant-based protein sources.6 Although flavor and smell were important influences, the strongest factor for food selection appeared to be the total amount of protein in the food, rather than its source.

To date, only two controlled studies have been completed with dogs. In both, dogs have also demonstrated a similar talent to their feline friends for self-selecting the macronutrient content of their diets.7,8 The studies were conducted by different research teams and used somewhat different methodologies, but both reported that dogs preferentially selected diets that were low in carbohydrate and high in fat and protein.

When expressed as a percent of energy, dogs gravitated to a general distribution of 30 to 38 percent protein, 59 to 63 percent fat and 3 to 7 percent carbohydrate. Interestingly, wolves self-select diets that are even lower in carbohydrate: only about 1 percent. Initially, the dogs in these studies were attracted to very high fat diets, but over a period of several days reduced the proportion of fat and moderately increased protein.

An important finding of the most recent study was that when dogs were allowed to choose these dietary proportions over a period of 10 days, they substantially over-consume calories. Because of this, on average the dogs gained almost 3.5 pounds in just 10 days of feeding.

It’s Complicated

At this point in time, we know that dogs can better digest starch in their diet compared with their wolf ancestors (and with present-day wolves). This increased capability is at least partially due to an increase in the production of pancreatic amylase.

We also know that, like us, dogs digest cooked starches very efficiently, but cannot utilize raw starch. The inclusion of at least some level of starch in a dog’s diet provides an efficient source of energy (calories).

Finally, most recently, we have learned that when given the choice, dogs preferentially select a diet that is low in starch and high in protein and fat. However, self-selection of this type of diet (if fed without portion control) may lead to overconsumption and weight gain.

Still, none of this information provides evidence for the healthfulness of a diet containing some starch versus a diet that contains very low (or no) starch in terms of dog’s vitality, ability to maintain a healthy body weight and condition, development of chronic health problems, and longevity.

This has not stopped proponents of low-carb or carb-free diets from making such claims, however. The fact that dogs gravitate to a diet that is high in protein and fat and low in starch is not to be confused with evidence that such a diet has been proven to be healthier or is capable of preventing illness. We simply do not know.

What we need is evidence of whether or not dietary carbohydrate is harmful, beneficial, or, well, neither. Dogs are generalists after all. It is quite possible that they, like many animals, are capable of thriving on a wide variety of diet types, including those with some level of starch.

Like I said, it’s complicated.

Cited Studies

1. Axelsson E, Ratnakumar A, Arendt ML, et al. “The genomic signature of dog domestication reveals adaptation to a starch-rich diet.” Nature 2013; 495:360-364
2. Arendt M, Fall, T, Lindblad-Toh K, Axelsson E. “Amylase activity is associated with AMY2B copy numbers in dogs: Implications for dog domestication, diet and diabetes.” Animal Genetics 2014; 45:716-722
3. Arendt M, Cairns KM, Ballard JWO, Savolainen P, Axelsson E. “Diet adaptation in dogs reflects spread of prehistoric agriculture.” Heredity 2016; 117:301-396
4. Reiter T, Jagoda E, Capellini TD. “Dietary variation and evolution of gene copy number among dog breeds.” PLOSone 2016; 11:e01148899
5. Hewson-Hughes AK, Hewson-Hughes VL, Miller AT, et al. “Geometric analysis of macronutrient selection in the adult domestic cat, Felis catus.” Journal of Experimental Biology 2011; 214:1039-1051
6. Hewson-Hughes AK, Colyer A, Simpson SJ, Raubenheimer D. “Balancing macronutrient intake in a mammalian carnivore: disentangling the influences of flavor and nutrition.” Royal Society of Open Science 2016; 3:160081.
7. Hewson-Hughes AK, Hewson-Hughes VL, Colyer A, et al. “Geometric analysis of macronutrient selection in breeds of the domestic dog, Canis lupus familiaris.” Behavioral Ecology 2013; 24:293-304
8. Roberts MT, Bermingham EN, Cave NJ, Young W, McKenzie CM, Thomas DG. “Macronutrient intake of dogs, self-selecting diets varying in composition offered ad libitum.” Journal of Animal Physiology and Nutrition 2018; 102:568-575

Linda P. Case is the owner of AutumnGold Consulting & Dog Training Center in Mahomet, Illinois. Linda is the author of Dog Food Logic, has a new book, Dog Smart, and writes The Science Dog blog.