In this section, the theory of reproductive states is put to a test. Can the theory explain not only the data from which it was constructed, i.e., the various hormone effects upon specific loci in the nervous system, but also data on a different type of problem for which it was not designed? Such a problem, seemingly amenable to this analysis, is the process of domestication.
Domestication involves both genetic and ontogenetic effects in laboratory muroid rodents. It has been shown that laboratory stocks differ genetically from their wild cousins, even when rearing conditions are equalized, in various types of defense behavior. Laboratory deermice have less neophobia than wild deermice (Price, 1972). Laboratory rats have less neophobia and are easier to handle than wild rats (Barnett and Stoddart, 1969). Laboratory mice are less resistant to handling or to recapture and vocalize 1ess to handling than wild mice (Connor, 1975), and are less likely to show freezing (Smith, 1972, 1978). Also, domesticated rats and mice have much smaller adrenal glands than do their wild cousins (Richter, 1950; Boice, 1973; Christian, 1975). There are also ontogenetic effects as shown by the fact that domesticated animals such as rats (Rasmussen, 1939) and gerbils (Clark and Galef, 1977) can become "wild" if they are raised in an environment where they have an opportunity to learn to run and hide when young. Also, it is known that there is a taming effect of handling upon animals, particularly if they are handled when young (Stone, 1982, Galef, 1970).
The genetic effects of domestication may be explained in terms of the mechanisms of the reproductive states. When wild animals are brought into the laboratory, most of them adopt a reproductive postponement state in response to the stress of capture and the new (neophobic) surroundings. They produce CRF, ACTH, and adrenal corticosteroids and suppress the reproductive system. Thus, for example, wild muroid rodents are less likely than their domesticated counterparts to respond to vaginal-cervical stimulation with ovulation and pregnancy, presumably because their extreme reaction to handling evokes an adrenal-mediated antigonadotrophic effect (Kenney et al, 1977). Among wild rats only a small proportion of females will engage in reproduction in the first generation brought into the laboratory (Boice, 1972, 1973). Similarly, Powell (1973) has noted species differences in the domestication of wild muroid rodents; black rats may breed in the laboratory, but cotton rats will not breed unless given nest boxes into which they can escape, and wood rats will die of stress rather than reproduce. A few animals do reproduce, however, and these, it should be predicted, will be the animals with the lowest response of CRF and ACTH and the least suppression of reproductive function. Thus, within the first generation, there is already a powerful selection for animals with low pituitary-adrenal activation and low adrenal gland weight. In other words, domestication selects immediately and powerfully against animals with a reproductive postponement state.
From the schema presented in figure 1, it is evident that the genetic selection effects of domestication could take place at several links in the motivational system of defense. If the only effect were to weaken the development of the motor pattern of CRF-ACTH secretion or the output glands themselves, then one would expect that no other direct behavioral effects would be observed. If, on the other hand, the genetic selection were to weaken any of the sensory filters for defense or the motivational mechanism for defense, then one would expect to observe a decrease in various types of defense behavior as the lunge-and-bite attack or neophobia, etc. Presumably, the latter occurs in many cases, which helps explain why laboratory stocks have decreased levels of these defense behaviors as well as smaller adrenal glands
In addition to genetic effects, there are ontogenetic effects of domestication. Some of these may be explained in terms of the ontogenetic development of the consociate modulator. Rather than being totally "wired-in" by the genetic code, the recognition of familiar consociates is probably developed during ontogeny by a. process of tactile contact and olfactory familiarity with littermates and parents (Adams, 1978). The taming effect of handling, mentioned above, occur as a result of its similarity to this process.
Another ontogenetic effect of domestication may not be related to reproductive states, however. It would appear that part of the success in taming laboratory animals derives from the practice of raising them in cages where they have no opportunity to run and hide in a tunnel or nest (Clark and Galef, 1977). In this way, the necessary- prerequisites for the development of defense behavior in wild animals can be circumvented in the laboratory, and animals can be raised that are less defensive without regard to genetic factors.
In summary, it is proposed that the processes of domestication act in part upon mechanisms that evolved under natural conditions to mediate shifts from a reproductive readiness state to reproductive postponement and emigration states. In other words, thanks to these mechanisms of reproductive states, there were pre-existing mechanisms that made domestication relatively easy and rapid.