unfamiliar animals nearby for a few days and observing them for the presence of combative behavior. Many descriptions of this procedure are available (for chimpanzees, Fritz 1986, 1989, 1994; for macaques, Coe 1991, Bernstein 1969, Bernstein and others 1974a, b, Crockett and others 1997, Reinhardt 1988, 1989a, 1991a; for capuchins, Cooper and others 1997; and for squirrel monkeys, Vermeer 1997).
Group housing can pose a problem in gaining access to individual subjects for testing or biomedical sampling. At least three solutions have been used: training of individual animals to enter small transfer cages, movement of animals to smaller gang cages and then to transfer cages, and the inclusion of tunnels within group enclosures so that animals can be herded into the tunnels and then moved one at a time into transfer cages (Clarke and others 1988; Knowles and others 1995; Phillippi-Falkenstein and Clarke 1992; Reinhardt 1992a; see also "Restraint and Training" later in this chapter). Those techniques are superior to techniques that require personnel to enter group pens with nets and gloves to capture specific animals. The latter procedures are stressful and dangerous to technician and animal alike. Although the limited use of nets is recommended in Chapter 6 for some small New World monkeys, we do not recommend it in general or especially for macaques, for which training to enter a transfer cage is much preferred, because it reduces the risk to personnel of exposure to bites, Circopithecine herpesvirus, and other zoonoses.
Before a social group is established, the social organization of the species under free-ranging conditions should be examined. For example, young male rhesus monkeys form associations, so pair housing of males might be successful (Reinhardt 1989a, 1994b, 1995). Adults of other species are often intolerant of members of their own sex (Coe and Rosenblum 1984; Crockett and others 1994), especially in the presence of adults of the other sex; for example, because of the natural social affiliations of squirrel monkeys, females are more readily housed together than males (see Saltzman and others 1991). Knowledge of natural sex or age-class affinities can aid in the planning of social units.
Several key elements of the housing area for social units should be addressed in either cage design or SOPs. Provision of refuges might be beneficial to prevent fighting in some species. Provision should be made for easy removal of individual animals if fighting occurs and for ready access to animals for protocol purposes or husbandry routines. The safety of facility personnel should also be a driving force behind SOPs or enclosure design for routine husbandry procedures. For example, the use of a shift or transfer cage or run might be necessary when staff enter an enclosure to perform routine tasks. Flexibility in converting social to individual housing of animals might be desirable to control feeding and to treat or examine an individual animal. The ability to partition a larger cage for short periods greatly facilitates cleaning and maintenance.
Although social housing of primates can enhance reproductive capability and development of species-typical behavior, some animals cannot be success-