have a higher DE for animals with substantial gastrointestinal microbial fermentation than for animals that must depend exclusively on endogenous digestive enzymes. As a consequence, DE concentrations in foods are most meaningful if determined during consumption of those foods by the target species in typical amounts per day. Such determinations have seldom been made with nonhuman primates, and it is presently necessary to use DE values for foods coming from studies of other species (usually domestic) that have gastrointestinal anatomy and physiology similar to that of the target primate species.
Apparent metabolizable energy (ME) of a food is equal to food GE minus GE lost in the feces, urine, and combustible gases. Subtraction of the latter quantity is an obviously arbitrary feature of the definition of apparent ME, in that the loss of food GE in combustible gases is a consequence of digestive processes. In most cases, gaseous GE lost is largely in the form of methane from microbial fermentation in the foregut or hindgut. That loss is not accounted for in apparent DE but for some species could account for a high proportion of the food energy that is unavailable for support of metabolic processes. Analogously to the calculation of true DE, true ME is calculated by subtracting metabolic losses of nonfood origin from apparent ME. Apparent ME values are used much more commonly than true ME values.
For some animal species, systems for expressing energy and nutrient requirements are based on ME intake. It is desirable to express requirements based on ME; however, research in primates has not been conducted to allow use of an ME-based system. Given the diversity of primate species and food items fed to these primates, ME values for the majority of food items have not been determined. This lack of data presently hampers development of more refined estimates of nutrient needs.
Research is needed to determine ME values of particular food items for specific primate species. Obviously, not all species can be studied, due to the intensive nature of the research and the limited availability of research animals. A reasonable approach to obtaining critical information on ME would be to conduct experiments with several model primate species, from which estimates could be extrapolated for other similar species. Primate species most important to study might be those 10 model species identified in Chapter 11: (1) macaques, (2) baboons, (3) squirrel monkeys, (4) cebus, (5) howlers, (6) marmosets and tamarins, (7) colobus and langurs, (8) lemurs, (9) chimpanzees, and (10) humans.
Nitrogen-corrected ME, net energy, and other expressions of energy concentrations in foods are presented in NutritionalEnergetics of Domestic Animals (National Research Council, 1981b). The system that has been most widely applied to foods for primates involves calculation of physiologic fuel values (or physiologically available energy, an approximation of apparent ME); the system has been reviewed by Widdowson (1955) and is based on the German studies of Rubner in 1880-1901 and studies of Atwater (Rubner’s student) in 1895-1906 in the United States. Most tables of composition of foods for humans list physiologically available energy values (and conversion factors for carbohydrates, protein, and fat in specific foods) based on digestibility trials conducted by Atwater and others (Merrill and Watt, 1955). The general physiologically available energy conversion factors of 4 kcal·g-1 for carbohydrates and protein, and 9 kcal·g-1 for fat yield reasonable approximations of apparent ME in the typical US human diet but not in specific foods or in high-fiber diets (National Research Council, 1989). For those, specific conversion factors, such as those in US Department of Agriculture Handbook No. 8 (Watt and Merrill, 1963) should be used.
Souci et al. (1994) used the general conversion factors of 4 and 9 kcal·g-1 for protein and fat, respectively, but applied the carbohydrate conversion factor of 4 kcal·g-1 only to available carbohydrate. Available carbohydrate was defined as monosaccharides, disaccharides, oligosaccharides, nonstructural polysaccharides, and the sugar alcohols sorbitol, xylitol, and glycerol. If concentrations of those compounds were unknown, available carbohydrate was defined as 100 - (water + protein + fat + minerals + total dietary fiber + available lactic, citric, and malic acids). The conversion factor used for available organic acids was 3 kcal·g-1. Total dietary fiber included primarily cellulose, hemicellulose, and lignin (or water-soluble + water-insoluble fiber) and was assigned an available energy value of 0 kcal·g-1. Ethanol was assigned a value of 7 kcal·g-1.
The general conversion factors of Souci (1994) assume that there is no energy derived from dietary fiber and ignore interactions among macronutrients that may impact energy availability. A more robust approach to estimate dietary energy has been proposed by Livesey (1999, 2001). This empirical approach accounts for energy derived from all macronutrients and accounts for nutrient-nutrient interactions (Baer et al., 1997).
To conform with the first and second laws of thermodynamics, energy intake by an animal must equal energy used plus energy lost. Thus, GE in ingested food must equal