Page 3

Payne (1965), using Abrams's data, published estimates of metabolizable energy requirements for maintenance of growing, adult, pregnant, and lactating dogs.

Abrams (1976) recalculated available data related to the determination of BMR in dogs. Estimates were made of energy requirements for maintenance of adult dogs in terms of DE and expressed as joules (J) needed per day or per kilogram of body weight. Constants were determined by regression analysis and an equation for calculating kJ energy requirement was derived. Of significance is that estimates for males exceeded those for females.

Abrams (1976) reported greater metabolizable energy requirements for dogs weighing less than 20 kg than those calculated by the method of the 1974 NRC report ^{(132 × BW)}, but his values did not differ greatly for heavier dogs. More recently, Blaza (1981) measured ME requirements for maintenance of medium and giant breeds of dogs. Great Dane and Newfoundland dogs required 1.5 and 1.3 times, respectively, the calculated ME from the equations of NRC (1974). Requirements of Labrador Retrievers were comparable to requirements predicted from NRC (1974). The Subcommittee on Dog Nutrition decided to seek available data for maintenance based on controlled feeding conditions and to develop a prediction equation using the approach advocated by Thonney et al. (1976).

Limited data were available for individual dogs weighing from 4 to 36 kg. Seven breeds (Beagle, Boxer, Labrador Retriever, Pointer, Poodle, and two types of Dachshund) were included. From visual inspection of the scatterplot it could not be concluded that either breed or sex differences affected the relationship of daily ME required for maintenance of body weight.

The data were fitted to a linear model and an allometric model. The simple linear model (ME = b0 + b1W, where b0 is the intercept, b1 is the slope, and W is weight in kilograms) described the within-species relationship between basal heat production and weight. The allometric model used was ME = b0Wb1 (where b0 is the mass coefficient and b1 is the exponent). The allometric approach to energy data was first used by Kleiber (1932) on data in which the lightest animal (rat) weighed close to zero compared to the heaviest (cattle). This model implies that the relationship will intersect the origin. There is no reason to use the allometric model unless a curved line that intersects the origin fits the data better than a simple straight line. Since large weight differences comparable to those in the data used by Kleiber (1932) do not exist within most species (including dogs), it is unlikely that the best model is one that requires the relationship to intersect the origin.

The results of fitting the two models to the data are shown in Figure 1. The linear equation, ME = 144.4 + 62.2 W, and the allometric equation, ME = 99.56 W0.879, both explained 85.6 percent of the variation with a standard deviation of 260 kcal. The daily ME requirement predicted by these equations is shown in Table 5 (see p. 45) for dogs varying in weight from 1 to 60 kg. Either equation may be used.

The NRC (1974) equation (ME = 132 W0.75) is also shown in Figure 1, but it explains less of the variation than the allometric and linear equations actually derived from the data. It underpredicts the ME requirements of the larger dogs and, thus, supports the work of Blaza (1981). More data are needed, however, to predict accurately the energy needs of dogs greater than 35 kg mature body weight. Therefore, until more data are available, it may be advisable to determine feeding levels based on the 1974 equation predictions.

These values or any others cannot be taken as absolute ME requirements for any individual or breed of dog, since needs vary with age, activity, body condition, insulative characteristics of the hair coat, temperature, acclimatization, external environmental circumstances, and psychological temperament.

Finally, it would not appear to serve much practical purpose to further refine energy requirements of even ''average" dogs, since the ME concentration of foods available to such "average" dogs is frequently either unknown or can only be calculated by methods resulting in no greater precision.

Generally, adult dogs adjust their food intake to energy requirements. Cowgill (1928) found that dogs previously adjusted to an appropriate intake of a particular diet consumed fewer grams, but a similar number of calories, when a higher-energy-density diet was offered. Durrer and Hannon (1962) reported that caloric intake varied inversely with long-term changes in environmental temperature. In July, when the mean temperature was 17°C, Beagles consumed approximately 163 kcal of ME per ^W per day, while Huskies consumed 127. In November, when mean temperatures were - 17°C, the respective daily ME intakes for Beagles and Huskies were 278 and 205 kcal per ^W. Huskies exhibited a marked increase in hair growth during November and December, while little seasonal change in hair growth was seen in Beagles. Dogs of both breeds minimized heat loss during extremely cold weather (less than - 40°C) by curling into a ball and tucking their noses and tails underneath their bodies. While Huskies showed no evidence of shivering and refused to sleep in plywood shelters, Beagles shivered and sought shelter. These data illustrate the marked effect on energy requirements imposed by the environment and the additional influence of differences in breed and behavior.

While weight changes were small, both Beagles and Huskies were heavier in the summer than in the winter.