haut de page



















haut de page




mise à jour
27 mars 2003
Rec. Progr. Horm. Res 1972; 28; 707-733
Gonadal hormones and behavior of normal and pseudohermaphroditic nonhuman female primates
RW Goy and JA Resko
Wisconsin regional primate research center


I Estradiol, progesterone and the behavior of primates [...]
II Androgens and the behavior of normal and pseudohermaphroditic female monkeys

The hormonal induction of heterotypical behavior in female primates seems to be achieved in very limited ways and in an unpredictable fashion when testosterone is injected at varions times after birth, and sexual behavior per se seems highly resistant to any masculinizing influence.

For example, in a recent experinient by Dr. W. D. Joslyn, testosterone was injected three times weekly for 8 months into three normal female monkeys, beginning at approximately 5 months of age and ending at 13 rnonths of age. Studies of the behavior of these animals were made beginning with the first injections, of testosterone propionate, and continuing throughout the final month of testosterone propionate treatment, and for several months after termination of testosterone treatment. These females became hyperaggressive and almost all their aggressive attacks were directed against the male members of their peer groups. In addition, there were marked and significant increases in yawning behavior, which is ordinarily related to the display of social dominance by males. These behavioral changes induced by testosterone in young females could be considered a kind of masculinization, but it is essential to point ont that this masculinizing influence did not extend to other kinds of behaviors which are ordinarily sex-related.

The behavior of young rhesus monkeys in social groups is markedly sexually dimorphic in a number of respects. For example, young male monkeys display sham threat, rough and tumble play, chasing play and mounting behavior much more frequently than do females in the same groups. Similarly, young males initiate play signifieantly more often than do females. However, none of these dimorphic behaviors were influenced by testosterone injected for 9 months into the juvenile female monkeys studied by Dr Joslyn. Their averages for frequency of performance of these activities during the last month of injection are seen in Table, and their deviations from the standards set by normal control females studied over many years are not large enough to exceed chance expectancy.

The limited masculinization achieved with these postnatal injections of testosterone propionate for a long period of time contrasts markedly with the changes induced in the same patterns of behavior when genetic female monkeys are exposed to androgens prior to birth and rendered pseudohermaphroditic. The patterns of sham threat, rough and tumble play, play initiation, chasing play, and mounting behavior shown by such pseudohermaphroditic females differ significantly from those of normal control females and illustrate the broad scope of the masculinization achieved with prenatal androgenic exposure. Longitudinal and developmental changes in all of these social behaviors are highly similar. For illustrative purposes, changes in the frequency of Play Initiation with age for normal males, normal females, and pseudoherinaphrodites are shown in Fig. 9.

The limited effects that androgens have on the behavior of normal females in adulthood can be ascertained from the following set of studies. The first of these studies shows that the behavior of adult rhesus males differs markedly from that of females. When male and female subjects are paired for testîng at frequent intervals throughout the menstrual cycle of the female, behaviors bearing no obvious function in reproduction or sexuality, as well as those forming a constant part of the sexual repertoire, differ between the sexes. Data on the frequency of two sexually dimorphic behaviors, "yawning" and mounting, were obtained by Stephen Eisele in a study of seven pairs of rhesus. In this study the pairs were tested for 12 minutes or until the male ejaculated, if ejaculation occurred sooner than 12 minutes. Under these conditions, as shown in Table IV, the fenmales never yawned or mounted, regardless of the stage of the cycle when the tests were given.

Males, in contrast, displayed both behaviors with reasonable consistency, and there was a tendency for the frequency of performance to the higher when tests were given on the female's follicular days than when the female partner was in the luteal phase of her cycle.

In part, the dimorphisin observed in Eiscle's study of heterosexual pairs can be attributed to the insufficiency of endogenous androgens in the female compared with the male, and in part it is due to the psychological influence of the partner. When females are tested with another female rather than with a male, a higher frequency of yawning and mounting typically occurs. In short, in the pair-testing situation, the probability of performance of yawning and mounting is increased by the use of a female as partner regardless of whether the performer is male or female.

A study carried out in our laboratory in collaboration with CH Phoenix showed that even when this psychological factor was taken into account, the behavior of adult females injected with testosterone propionate was only partially masculinized. Of some 28 beavioral characteristics recorded during standardized 10 minute tests of females, tested with female partners, only one behavior, which we call and which is normally characteristic of the adult inale tested under similar circumstances, was influenced. The effects of testosteron propionate administered for 4 weeks and in two different doses on both yawning, and mounting behavior in adult females are summarized inTable V.

The reader can observe in Table V that in female-female pair tests both yawning and mounting are displayed prior to the injection of testosterone propionate. This difference from the results obtained for females when adult males were used as partners. When testosterone propionate was injected, the frequency of mounting did not increase significantly and remained substantially below that for normal males. In contrast, the frequency of yawning behavior not only incresed significantly during injections of testosterone propionate, but it was substantially higher than that for normal males tested with females. Agan, as in the previous example with juvenile females discussed in an earlier section of this article, testosterone propionate in the adult female, although it induced a 20-fold increase in yawning behavior, failed to induce any significant increases in homosexual mounting activity. We interpret these results to mean that testosterone has little or no ability to "masculinize"the sexual behavior of normal female rhesus even though some influence of this hormone on social behaviors can be measured. Althoughthe hamster is more sluggish in its response than other rodents, the results of similar experiments in females from lower mammalian species contrst markedly with our results on primates. In the former species mounting behavior is regulary induced or augmented with injections of testosterone. [...]

The results of our experiments with primates show that the effects of testosterone given before birth are not greatly different from those of similar treatments in nonprimate mammals. However, such a conclusion seems premature, and the manner in which the hormones act to accomplish masculinization may be in fact, quite different in these two mammalian groups. The influence of experience on the development of sexual and social behaviors is small in lower mammals compared with higher primates. For the former, it may be possible to postulate, as Dorner (1970) has done, that the androgens present early in development act upon separate neural circuits for the expression of masculine and feminine components of sexual behavior. This seems all the more plausible since very small brain lesions and very discrete placements of steroïd hormones in the brains of castrate animals have been shown to abolish or restore, respectively, full sexual behaviors in a variety of nonprimate animals. When one considers, however, the important contribution which experience may make to the expression of sexual and social beliaviors in primates it does not seem reasonable to postulate a simple manner or a diserete locus of androgen action. If in higher primates a large part of sexual and sex-related behavior is experientially determined, then it seems reasonable to infer that androgens present before birth ut upon the neural circuits mediating the effects of experience as well as upon the neural circuits for masculine and feminine behavior. Our experiences with the rhesus monkey suggest that one might turn one's attentions instead to the effects of androgen on the development of neural systems underlying drive and reward, i.e., those systenm which permit experience to exert its ultimate influences on behavior.

« It is ironic that testosterone "the male sex hormone," is more closely associated with the yawning rate than with the mounting or intromitting rates » Charles Phoenix
Sexual steroids exert several effects on both central dopaminergic and oxytocinergic systems by acting either at the genomic or membrane level  
credit photo : "Asif A. Ghazanfar and Aristides Arrenberg"
Max Planck Institute for Biological Cybernetics