1. Introduction
Food preference tests are used as an effective tool to estimate some aspects of the feeding behavior in dogs [
1,
2], contributing to new diet formulations to meet animals’ nutritional and palatability requirements. In preference tests, dogs are simultaneously exposed to different diets over a certain amount of time (which is usually shorter in dogs than in other species, because of their voracity) [
3]. Dietary preferences in mammals are intimately linked to the sensorial characteristics of foods, reflecting, in part, their nutrient composition. In this light, it has been described that dogs are able to select and prefer foods even when they liberate small amounts of volatile compounds, enhancing consumption when there is more moisture content and a small particle size [
4,
5]. Similarly, although the dog maintains strong preferences for protein foods [
6], domestication has created adaptive physiological changes that have generated preferences for simple carbohydrates [
7], such as sucrose [
4,
8]. In addition, dietary preferences in dogs are influenced not only by the sensorial or nutritional proprieties of foods, but also by intrinsic variables of animals [
9]. Alliesthesia could affect the intake or preference of dogs for specific nutrients or related flavors depending on their internal status that generates different pleasure sensations [
10]. Thus, variables such as a dog’s breed, age, body condition, weight, or sex may affect their feeding behavior. However, there is little information about how all these variables may affect food choices in domestic dogs. Knowing the effect of these variables allows a better understanding of their food selection and food acceptability. The aim of this study was to understand the relationship between the nutritional compositions of dogs’ diets and their associated preferences, and to analyze how intrinsic variables of these animals may influence their feeding behavior.
4. Discussion
The present study provides information about possible relationships between food compositions and dogs’ intrinsic characteristics on their food preferences. It is already known that diet selection could reflect the internal needs of an animal, helping them to reach homeostasis [
10]. Describing different factors that may influence feeding behavior of dogs can improve diet-specific formulations to increase their palatability and to amend animal health and welfare problems. It was observed from PCA that DM and CF negatively affected dogs’ food preferences, being parts of the third and fourth principal components, respectively. As noted in this study, dogs, as with other mammals, have been reported to prefer moist and semimoist diets to dry diets [
4,
5]. Nevertheless, de Brito et al. [
20] observed that increasing dietary moisture did not affect food preferences of dogs. Nonetheless, it is possible that the difference in moisture level used in that study was not enough to influence preferences. The pet food industry has considerably reduced the water content of their products with dry extruded foods, facilitating their storage and reducing purchase frequency and feed contamination [
21]. However, as we observed that DM is negatively related with dogs’ preferences, it is necessary for the industry to include palatable additives to their formulations such as highly digestible ingredients, natural or artificial umami flavors, and/or an external palatable fat cover, which in some cases may disrupt dogs’ intake regulation, leading to obesity problems [
22].
It is well known in other mammalian species such as pigs that food digestibility and palatability are closely related, and CF presented a negative relationship with dogs’ preferences [
23,
24]. The feeding behavior of domestic dogs (
Canis familiaris) could be described by the knowledge of the feeding patterns of their ancestral species, the wolves (
Canis lupus) [
8], which consist predominantly of the consumption of fresh meat obtained by group hunting with low contents of both DM and CF [
6,
25]. However, we must also consider that dogs can accept and even prefer different kinds of foods, moving away from their ancestors in the preference for protein-rich foods [
26,
27]. Conversely, the absence of relationship between PC1, mainly represented by palatable components (CP, EE, LIP, ME) and dogs’ preferences is intriguing. These nutritional components should be key elements when increasing the preference for one food due to their biological value in dogs. Nevertheless, we have to consider that in this study, the type and quality of the protein or fat sources that may change an animal’s preferences were not analyzed [
28].
The effect of intrinsic variables such as weight, breed, age, and sex on feeding behavior has been previously studied in several mammals. For example, it is described that humans differ in their dietary preferences explained by some of these factors [
29,
30]. However, little information was available in the case of the domestic dog [
6,
31]. In our analysis, breed of dog influenced intake, in which heaviest breeds such as Boxers and Labradors showed higher consumption, even when intakes were corrected by the animals’ body weight. Nevertheless, Beagles were more selective than Boxers as reflected in their higher food preferences across trials. Dog breeds have been developed to perform different activities around human populations, acquiring different motor and cognitive capacities [
32,
33]. It has been observed that brachycephalic breeds, such as the Boxer, have a lower olfactory capacity due to the morphological changes of their skull that directly affect the position of the olfactory lobe [
34], affecting the identification and selection of food. However, it has been observed in a recent experiment that brachycephalic breeds such as Pugs outperformed the German Shepherds in odor discrimination and the Greyhounds in motivation to perform feedings tasks [
35]. Bloodhound breeds such as Beagles are commonly used in food preference tests because of their olfactory abilities and for the ease of working with them as small and docile animals [
36]. The existence of breeds in different environments may also affect food selection because its availability and cost of acquisition may vary. Even animals genetically related that live in similar environments may develop different feeding behaviors which would be given by a component of "tradition" within the subpopulations [
37]. Flavor preferences have been identified due to pre-and postnatal learning in dogs and other species, so this must also be considered to understand their selection of a particular diet [
38,
39].
Results also showed a relationship between breed and body weight in terms of the animals’ food preferences, with the heaviest Boxers presenting higher preferences than did the heaviest Beagles. Animals’ experience as well as aging could be differently developed in dog breeds. The large amounts of food that dogs are able to consume in a short time [
6,
25] may predispose some breeds to obesity if they are given diets with high energy or fat content, as in the case of Labradors and Beagles [
40]. It is clear that body condition is a better measure than body weight to understand how physiological changes produced by overweight could change food perception and food metabolization, especially when animals have different ages and/or sizes. Nevertheless, the data collection of body condition scores were inconsistent through the years of the study in this experimental center, having used dogs’ weight data that could make more sense in intrabreed comparisons. Obesity may deregulate food intake as well as dietary preferences, and obese animals could need more palatable ingredients to activate neural pathways of pleasure leading to a deterioration in the perception of ingredients that differ little in their reward value. In humans of different body mass indexes (BMIs), there are differences in receptors related to the intensity of lipid perception, affecting preferences for fatty foods [
30]. Leptin levels are directly related to body mass, affecting preferences for sweet flavors [
41]. A chronic increase in leptin levels has been shown to reduce sweet taste sensitivity, increasing the consumption of sucrose or other sweet components of diets [
42]. On the other hand, Boxer breeds tend to maintain their body weight throughout adult life, and lower body weights are probably related to young, less experienced animals. Moreover, because body weight is related to age, especially in the case of Beagles, heavier animals may also suffer changes in preferences due to body changes that could directly affect food selection. A deterioration of dental structure, a reduction of sensorial capacities, a decrease in the energy demand, and other metabolic changes could directly affect the discrimination and selection of food in domestic dogs.
Sex would also have a potential effect on dietary behavior of mammals due to differences in energy needs, olfactory abilities, and hormonal changes among individuals [
29]. In dogs, it has been observed that females present a higher preference for diets containing 1% of sucrose and other additives as compared to males [
31,
43]. The absence of a sex effect over preferences in this study could be explained because previous studies in dogs compared males against sterilized females [
31]. The latter tend to accept a wider range of foods, being less selective than males as a result of the suppression of estrogen effects that can increase food intake and appetite [
44], suggesting that more than sex influences the effect of animals’ reproductive status over feeding behavior. Finally, similar to behavior observed in several species, dogs were affected by environmental temperature and presented a lower food intake during the hot season of the year [
45,
46,
47,
48]. This reflects the animal’s lower energy needs to maintain their temperature in warm environments and should be considered when designing food preference trials to eliminate possible biases, especially when environmental temperature cannot be totally controlled as in the present study. In regards to this, the current results could differ in some way from those performed in completely controlled environments, and may not be able to be extrapolated to all other circumstance. In addition, the care given to the dogs may also differ between experiments, affecting animals’ food preferences. For example, different levels of stress in dogs could change their pleasure perception in front of the diets delivered and thus their preference and acceptability as it was described in other species [
49,
50,
51].