Commercial feed manufacturers offer a variety of closed-formula diets for nonhuman primates. Although the specific amounts of each ingredient in the formula are not usually revealed, most manufacturers will furnish estimates of typical nutrient content in printed form or on a Web site. The information from the manufacturer can then be compared with the estimated nutrient requirements listed in Table 11-1. However, commercial feed manufacturers routinely alter feed formulations based on the quality and availability of feed ingredients, and customers are typically not notified when these formulation changes occur (Knapka, 1997). Although changes might only involve alterations in the ratios of the ingredients listed, in order to control the variation in the dietary nutrients of interest and perhaps permit use of low cost ingredients, changes might also occur in dietary constituents that are not being measured. For example, dietary ingredient changes can result in alterations of phytoestrogen concentrations, which are not typically reported, but can have a significant effect on reproductive efficiency and tumor rates in laboratory animals. These changes in feed composition can have potential impacts on the health of the animals being fed and the quality of research conducted with experimental animal colonies. Because of the potential variation in nutrient composition and other nonnutrient factors that may have physiologic effects, closed formula diets are not recommended for many research situations. If closed formula diets are used in research, they should be used with extreme caution and the researcher should conduct independent analyses of the diets throughout the experimental period. Researchers and caretakers should maintain detailed knowledge of the composition of diets, and those dietary constituents—nutrients and nonnutritive components—that may be of special interest.

Persons conducting research with primates often use an open-formula diet, publishing the amount and identity of each ingredient. Information on diet composition has utility in the interpretation of research findings, but one should be wary of uncritically adopting diets based on formulas published in the past. The definitions of feed ingredients (and their nutrient compositions) tend to change, and it might be difficult or even impossible to formulate diets as originally specified. For example, an open formula might specify the use of a fishmeal containing 70% protein. Fish-meal containing 70% protein has traditionally been derived from processing of sardines and is no longer widely available. The fishmeal used in most feed mills today is derived either from menhaden (60% protein) or from anchovies (65% protein), and few commercial feed mills have more than one type of fishmeal on hand. Another example is related to the use of wheat in an open-formula diet. The many types of wheat (such as soft white winter, hard red winter, and durum) vary in protein concentration from 10% to 15%. Most feed manufacturing plants will have only one type of wheat, and that makes it difficult to meet specifications that require a particular type of wheat or wheat with a particular protein level. An example of an ingredient specification that is not consistent with current technology is related to the form of vitamin C. Most older published diet formulations specify ascorbic acid, whereas modern formulas use L-ascorbyl-2-polyphosphate, a biologically active vitamin C form that is much more stable.

Because of concern that natural-ingredient diet formulas published in this document would be used without critical consideration of the issues raised above, we have chosen instead to refer the reader to relevant literature. A National Institutes of Health open-formula high-fiber diet that was developed to study the effect of fiber on rhesus monkeys during quarantine has been used as a maintenance ration in a number of colonies (Morin et al., 1978; Knapka et al., 1995). Diets used for longevity studies with rhesus and squirrel monkeys, in which food was restricted, have been published by Ingram et al. (1990). Diets for marmosets (Flurer et al., 1983; Barnard et al., 1988) and diets for lemurs, howlers, colobus, langurs, mangabeys, and drills (Edwards, 1995) also have been described.

A number of investigators have used purified diets in their research, and these diets are referred to in many of the studies cited in this report. Purified-diet formulas for macaques (Macaca spp.) (Kark et al., 1974; Kemnitz et al., 1993; Thornberg et al., 1995), African green monkeys (Cercopithecus aethiops) (Scobey et al., 1992), and squirrel monkeys (Saimiri sciureus) (Rasmussen et al., 1979; Martin et al., 1972) and a liquid diet used for alcohol investigations with baboons (Papio spp.) (Leiber and DeCarli, 1974) have been published. They can be used as a starting point by those wishing to formulate a diet for a specific purpose. The original publications should be studied carefully and formulas modified as appropriate. Adjustment of nutrient levels is particularly important for diets that were used to produce nutrient deficiencies.

EFFECTS OF PROCESSING

Feed processing typically includes grinding of dietary ingredients to produce particles of approximately equal size suitable for mixing and then pelleting or extrusion. Such processing promotes diet homogeneity and reduces the likelihood that primates will select and consume only the ingredients that appeal to them, regardless of their relative nutritional importance. Many primates manipulate their food and generally prefer the physical characteristics of extrusions or pellets to ground meals.

Manufactured diets for nonhuman primates usually are prepared by extrusion. This process involves passing steam-moistened feed through a high-pressure, high-temperature chamber and forcing it through a small opening. The pres-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement