Determinants of behavioral rhythmicity during artificial menstrual cycles in rhesus monkeys (Macaca mulatta).код для вставкиСкачать
American Journal of Primatology 15:157-170 (1988) Determinants of Behavioral Rhythmicity During Artificial Menstrual Cycles in Rhesus Monkeys (Macaca mulatta) RICHARD P. MICHAEL AND DORIS ZUMPE Department of Psychiatry, Emory University School of Medicine and The Georgia Mental Health Institute, Atlanta Reproductive success in many mammals depends on synchrony between copulation and ovulation, which is insured by the phenomenon of heat in the female. Certain anthropoid primates including rhesus monkeys do not show heat but may copulate throughout the menstrual cycle, especially when pairs are isolated from conspecifics. In social groups, however, mating mostly occurs around midcycle. We wished to test the hypothesis that copulations are more closely linked to ovulation when males have simultaneous access to several females in different cycle phases. Artificial menstrual cycles were therefore induced by giving hormones to ovariectomized female rhesus monkeys observed in small social groups that each consisted of four females and one male. The cycles of two hormone-treated females in each group were either made to synchronize or be offset by 7-day increments so that the estradiol peak of one female occurred 7 days before, and 7 and 14 days after, the estradiol peak of the other. Radioimmunoassay of plasma samples (N = 224) confirmed the timing of the estradiol peaks. Results from eight unique male-female groups (4 males, 8 females, 16 male-female pairs, 718 tests) fully supported the hypothesis. Compared with synchronized cycles, the amplitudes of rhythmic changes in offset cycles were reduced for ejaculations made by males but greatly enhanced for ejaculations received by females. We propose that this socio-hormonalintegration of behavior in the group is highly adaptive and enhances the reproductive success of both males and females. Key words: menstrual cycle, hormones, behavioral rhythms, mate choice INTRODUCTION In mammals, the reproductive success of both males and females depends heavily on synchrony between the timing of ovulation and maximum copulatory activity. From a reproductive and genetic standpoint, copulations occurring during unfertile phases of the ovarian cycle are simply wasted. Mechanisms that prevent this wastage therefore have enormous adaptive significance, and the familiar heat behavior of many female mammals is one such mechanism. Rhesus monkeys, like several other primates including the human, do not Received September 21, 1987; revision accepted January 31, 1988. Address reprint requests to Dr. Richard P. Michael, Department of Psychiatry, Emory University School of Medicine, 1256 Briarcliff Road, N.E., Atlanta, GA 30306. 0 1988 Alan R. Liss, Inc. 158 I Michael and Zumpe show heat but can copulate throughout the entire menstrual cycle. Fertility is confined to a period of a few days near midcycle when plasma estradiol levels peak and are followed 24-48 hours later by ovulation [Norman et al., 1976; Ferin et al., 1977; Knobil, 19831. Earlier field studies reported that rhesus monkeys live in semi-closed social groups containing several adult males and females with their young. It was noted that mating occurred seasonally during the fall, and that there was a monthly periodicity in female mating activity [Carpenter, 1942; Altmann, 1962; Kaufmann, 1965; Southwick et al., 1965; Lindburg, 19711. In laboratory studies, the length of the menstrual cycle was determined [Hartman, 19321, as were the associated hormonal events [Hess & Resko, 1973; Knobil, 1974; Bonsall et al., 19781. Studies in observation cages with male-female pairs established that frequencies of copulatory activity increased a t midcycle and that gonadal hormones determined this increase [Michael et al., 1967, 1968; Michael & Welegalla, 1968; Johnson & Phoenix, 1976; Wallen & Goy, 19771. There was much variability in the behavior. In some pairs, rhythmic changes were well marked, while in others they were virtually absent. As more data accumulated, however, the species-typical pattern gradually emerged [Michael & Zumpe, 1970a; Michael & Bonsall, 19771. The amplitudes of the rhythms of copulatory activity in pair tests were much less marked than those in feral social groups and in large outdoor enclosures [Wilson et al., 19821, although seasonal influences can modulate the amplitude of changes [Herndon et al., 19871. Most copulations occur in the context of temporary but exclusive consortships, during which a male and female remain in close proximity; feeding, grooming, and copulating together. There is reason to suppose that consortships during unfertile periods involve not only the wastage of copulations but also energetic and nutritional costs [Hausfater, 1975; Dittus, 1979; Bercovitch, 19831. In certain macaques and baboons, for example, consorting males spend more time in aggressive interactions, and consorting females spend less time feeding. A mechanism insuring that consortships are restricted to periods of maximum fertility is therefore likely to be highly adaptive for these species. Because of the differential between the sexes in gamete production and investment in offspring, evolutionary theory [Darwin, 1871; Williams, 1966; Trivers, 19721 predicts that male rhesus monkeys will attempt to maximize reproductive success by maintaining a stable, high level of copulatory activity with as many mates as possible throughout the mating season. On the other hand, females would be expected to copulate exclusively during periods of maximum fertility. In pair tests conducted in the absence of other conspecifics, a mix of the opposing tendencies of males toward constancy and of females toward periodicity would result in only moderately rhythmic mating activity. In a social context when other females are present, as long as their menstrual cycles are not synchronized, one would predict the continuously active male to mate with several females in succession as each one approaches midcycle, moving from female to female as the previous partner enters the luteal phase. This would result in high, constant levels of male copulatory activity (total numbers of ejaculations given by males) and high-amplitude rhythms of female copulatory activity (ejaculations received by females). The pattern of change would depend on the precise timing of one female’s menstrual cycle in relation to those of others in the group. This hypothesis cannot be tested under natural conditions because there is no practical or reliable way to regulate the timing and length of menstrual cycles in the wild. We have developed an “artificial” menstrual cycle in ovariectomized females [Michael et al., 19781that enabled us to test these predictions experimentally under rigorously controlled conditions. With the advent of specific radioim- Behavioral Cycles in Rhesus Monkeys I 159 munoassays, we now have detailed information about the plasma levels of the principal ovarian hormones as they change during the menstrual cycles of intact females. It is possible to reproduce these changes accurately with daily subcutaneous injections of a changing mixture of different doses of estradiol benzoate, progesterone, and testosterone propionate [Michael et al., 19821. Ovariectomized females so treated have 28-day cycles that are terminated by 1-2 days of vaginal bleeding indistinguishable from normal menstruation; animals have the typical midcycle change in basal body temperature, and the anterior pituitary gland responds to the fluctuations in circulating plasma hormone levels by producing a pulse of luteinizing hormone that resembles the ovulatory pulse of intact females. The behavioral changes seen in pair tests conducted during these cycles are virtually identical to those observed when the same animals were intact [Michael et al., 19821. We have used these hormonally induced artificial menstrual cycles in small groups of rhesus monkeys, consisting of one male and four ovariectomized females, to examine the behavioral effects of varying the timing of one female's cycle in relation to that of another female in the group. MATERIALS AND METHODS Animals Four mature male (weighing 9.2-11.3 kg) and eight adult female (weighing 4.6-5.9 kg) rhesus monkeys were obtained as adults from the wild. Animals were housed in individual cages in large colony rooms where temperature was maintained between 20-24 "C and artificial lighting gave a 14-hour day between 06:15 and 20:15 hours. Food consisted of Purina Monkey Chow, supplemented by vitamins, fresh fruit, and vegetables, and water was available ad libitum. General maintenance procedures have been previously described [Zumpe & Michael, 19771. All experimental procedures were in accordance with institutional regulations and with the NIH Guide for the Care and Use of Laboratory Animals (NIH Publication Number 85-23, revised 1985). Operative Procedures and Hormone Treatments Six months before the start of behavioral observations, all females were ovariectomized bilaterally through a midline subumbilical incision, and histological examination confirmed that ovariectomy was complete. In each of two groups of four females (females A-D, females E-H), two females (females A and B, females E and F) were given eight artificial menstrual cycles produced by daily subcutaneous injections of a changing mixture of steroids in 0.2 ml oil. Each injection was given a t 08:OO hours and contained estradiol benzoate (pg), testosterone propionate (pg), and progesterone (mg) as shown in Table I. This dose schedule was slightly modified from that published previously [Michael et al., 19821 to replicate more accurately the plasma hormone levels of intact females. Day 1 was always a Saturday so that all estradiol peaks (day 12) occurred on Wednesdays. Cycles were offset by treating one of the two females with daily subcutaneous injections of 5 pg estradiol benzoate for the offset interval (7 or 14 days). The two other females in each group (females C and D, females G and H) received daily subcutaneous injections of 0.2 ml oil throughout. Collection of Plasma Samples and Steroid Assays Once daily during days 10-15 (Monday through Friday) and three times a week a t all other times (Mondays, Wednesdays, and Fridays), 3 ml blood was obtained from the saphenous veins of the untranquilized, hormone-treated females 160 / Michael and Zumpe TABLE I. Doses of Steroids Administered in Daily Subcutaneous Injections of 0.2 ml Oil to Produce Artificial Menstrual Cycles in Ovariectomized Female Rhesus Monkeys Cycle Days 1-7 8-9 10 11 12 13 14 15 16 17 18-19 20-21 22 23-24 25 26 27 28 Estradiol benzoate Testosterone propionate (Fd (FLR) Progesterone (md 5.0 7.5 10.0 17.5 30.0 2.5 2.5 2.5 2.5 5.0 5.0 5.0 5.0 5.0 2.5 2.5 5.0 5.0 14.0 14.0 14.0 17.5 25.0 5.0 7.5 10.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 14.0 14.0 14.0 0.00 0 .oo 0.00 0.00 0.20 0.30 0.30 0.75 1.oo 1.00 1.25 1.50 1.25 0.75 0.75 0.50 0.20 0.00 that previously had been adapted to the procedure. Blood samples from oil-treated females were obtained once a week (Wednesdays). All samples were collected at 16:OO hours, 8 hours after injection. To verify the timing of the midcycle estradiol peaks in hormone-treated females, samples from alternate cycles were analyzed for plasma estradiol levels by radioimmunoassay without chromatography, using a highly specific antiserum [Wright et al., 19781. Inter- and intra-assay coefficients of variation were less than 8.4%, and water blanks read 6.7 ? 3.6 pg (224 samples). Behavioral Testing In the period 1983-1986, 1-hour behavior tests (N = 718) were conducted twice daily 5 days a week with groups, each consisting of one male and four females. Two tests are not included in the total: one because of computer malfunction and one because of injury. Testing took place in a large room (4.9 m wide by 4.9 m deep by 2.1 m high) where animals were observed from behind a partition. Behavioral observations were made by nine experienced observers who were assigned tests in random order and unaware of the compositions of the hormone injections, which were numerically coded. TRS Model 100 computers were used to record all behavioral interactions and the identities of the animals involved, and interactions were timed to the nearest second from the start of the test [Bonsall et al., 19853. A group of four females (females A-D) was tested with each of the four males (males 1-4) in turn during successive tests: with males 1 and 2 (morning and afternoon, respectively) on one test day, and with males 3 and 4 (morning and afternoon, respectively) on the next. Two successive test days therefore provided data from a 1-hour behavior test with each of four unique male-female groups, and each 28-day artificial cycle produced data from ten tests with each of these four male-female groups. At the conclusion of these tests, the entire study was repeated using the same four males and a second group of females Behavioral Cycles in Rhesus Monkeys / 161 (females E-H), thus providing data from a total of 16 male-female pairs with artificial cycles in eight uniquely composed male-female groups. Cycle Offsetting To assess the behaviorial impact of varying the timing of the cycle of one female relative to that of the other hormone-treated female in the group, their estradiol peaks were either synchronized (zero offset) or offset by 7-day increments so that the estradiol peak of one female occurred 7 days before (-7 offset), 7 days after ( + 7 offset), and 14 days after (A14 offset) the estradiol peak of the other. After a period of testing to permit acclimatization and behavioral stabilization, the following sequence provided data for 2 cycles per offset per female: 1) two successive cycles in which the estradiol peak of female A coincided with day 19 of the cycle of female B (-7 offset for female A and + 7 offset for female B); 2) one synchronized cycle (zero offset);3) two successive cycles in which the estradiol peak of female A coincided with day 5 of the cycle of female B ( + 7 offset for female A and -7 offset for female B); 4) one synchonized cycle (zero offset); and 5) two successive cycles in which the estradiol peak of female A coincided with day 26 of the cycle of female B and vice versa ( + 14 offset for female A and - 14 offset for female B). This sequence was repeated with females E and F in the second group of females. Behavioral Definitions Twelve behavioral categories were used: 1) number of ejaculations made by a male-the total number of ejaculations per test made by a male with all females, 2) number of ejaculations received by each female, 3) time to first ejaculationtime in seconds from the start of the test to the first ejaculation received by each female (a default of 3,600 seconds was given when a female received no ejaculations), 4) number of male mounting attempts-number of successful and unsuccessful male mounting attempts received by each female, 5 ) latency to the first male mounting a t t e m p t t i m e in seconds from the start of the test to the first male mounting attempt received by each female, 6) number of male redirected aggressive gestures [Zumpe & Michael, 1970,19791 received by each female-threats and attacks by a male at other females or nothing in particular while in proximity (within touching distance) with the female, 7) number of female sexual invitations-total number of presents, hand-reaches, head-ducks, and head-bobs [Michael & Zumpe, 1970bl made by each female to the male, 8) number of redirected aggressive gestures by each female-threats and attacks by a female a t other females or nothing in particular while in proximity with the male, 9) number of approaches by a male received by each female-number of times a male approached to within touching distance of a female, 10) number of approaches by each female received by a male-number of times a female approached to within touching distance of the male, 11)time spent in proximity-time in seconds when a male and female were within touching distance, 12) time spent allogroomingtime in seconds spent by a male and female grooming each other. Numerical Treatment of Results The behavior of an individual can be considered in terms of all other individuals in the group together or in terms of each individual singly. Thus, a male’s copulations can be considered as the number of ejaculations made with all four females or as ejaculations made with each of the four females individually (male-female pair). To test the hypothesis, a comparison of these two very different behavioral measures was made. For brevity, other behaviors are reported only in 166 I Michael and Zumpe SYNCHRONIZED - 7 OFFSET +7 OFFSET f 14 OFFSET 4 ? h 4 2 4 ru P ? I s Fig. 4., For latencies to the first ejaculation and first mounting attempt received by females, and for proximity times, offsetting the cycles of two females in a group greatly increased the amplitudes of changes in these behavioral measures of consort bonding. Each point is the mean of 30-32 observations. Upward arrows indicate the timing of the estradiol peaks of the females whose data are illustrated and downward arrows indicate those of the other females. Symbols a s in Figure 1. or artificial cycles [Michael et al., 19821. Beyond this, the enhancement of the rhythms in female mating activity in the presence of other females depended on the interaction between the hormonal status of the female herself and that of the other hormone-treated female in the group. Taken together with the male's tendendency toward relatively stable levels of mating activity, this interaction produced well-marked periods of maximum sexual activity in each female, periods that would be further accentuated in a more complex social setting involving several cycling females. Offsetting the cycles affected other measures of behavior that are involved in the formation and maintenance of consortships. Latencies to the first ejaculation and to the first male mounting attempt received by females, as well as numbers of mounting attempts, redirected aggressive gestures, approaches and grooming times, all became more rhythmic during offset cycles. A word about Proximity times which were completely nonrhythmic in synchronized cycles. This observation confirmed an earlier finding that proximity times did not change with hormone treatments of the female [Michael & Zumpe, 19841. The reasons for this somewhat surprising finding are complex but depend on an overriding effect of partner preferences, which create strong social bonds throughout the cycle and tend to mask any effects of changes in sexual motivation on proximity times. In offset cycles, social bonding was overridden by changes in sexual preference, and proximity times became highly rhythmic. Behavioral Cycles in Rhesus Monkeys / 163 - 7 OFFSET SYNCHRONIZED i 14 OFFSET + 7 OFFSET 4 0 J I r d I 7 14 21 28 1 7 14 21 28 I 7 14 21 28 1 7 I 4 21 28 DAYS OF ARTIFICIAL CYCLE Fig. 1. Changes in the numbers of ejaculations per test made by a male with both hormone-treated females in a group of rhesus monkeys (four males, eight male-female groups). In synchronized cycles, ejaculatory activity by males showed moderately rhythmic changes, but behavioral rhythmicity was virtually lost when cycles were offset. Downward arrows indicate the timing of the estradiol peaks of females A and E and upward arrows indicate those of females B and F. Each point is the mean of 14-16 tests. Vertical bars give S.E.M.S. Solid squares indicate vaginal bleeding. Ejaculations Made by Males None of the males ejaculated with oil-treated females, whose data are not considered further. Figure 1shows changes in the numbers of ejaculations made by males with all hormone-treated females during synchronized cycles and during each of the three types of offset cycles. In synchronized cycles, ejaculations showed rhythmic changes, with a peak on days 13-14 and low levels between days 19-28. Differences between cycle phases were significant (ANOVA,cycle segment x male design, F[9,1201 = 2.8, P = 0.005). In offset cycles, ejaculations showed no consistent rhythmic changes, and differences between cycle segments were not significant except for the -7 offset (F[9,1191 = 3.1, P = 0.002), in which the first peak (days 7-10) was higher than levels a t midcycle. Ejaculations Received by Females For each test there is one male but two hormone-treated females, and a + 7 or + 14 offset for one female is inevitably a -7 or -14 offset for the other. The number of ejaculations received by each female can be aligned according to the female’s estradiol peaks, which determine the offset intervals. In Figure 2, each point is the mean of 14-16 observations (2 artificial cycles per female, two females, four males in eight groups), and the data show the effect of cycle offsets on the ejaculations received by each of the hormone-treated females in a group. In synchronized cycles, ejaculations received by both females were only moderately rhythmic, and there were no significant differences between cycle phases (cycle segment x male design, F[9,1201 = 0.6 and 0.2, N.S.).However, during all offset cycles, the amplitudes of the changes in the numbers of ejaculations received by both females were massively increased, and cycle phases differed significantly in each case (F[9,119-1201 = 2.3-7.3, P < 0.02-0.0001). Compared with synchronized cycles, the midcycle peaks in offset cycles were higher when they coincided with the early (-7 offset) or late (k14 offset) progestational phase of the other female, whose copulatory activity was coincidentally lower. The exact pattern of change depended on the phase difference between cycles. 164 I Michael and Zumpe : SYNCHRONIZED +7 OFFSET - 7 OFFSET + 14 OFFSET 1-1 F: SYNCHRONIZED - 7 OFFSET + 7 OFFSET - u-• FEMALES A 0 E FEMALES B n 14 OFFSET 4 I c I I I 7 14 21 28 I 7 14 21 28 I 7 14 21 28 I 7 14 21 28 DAYS OF ARTIFICIAL CYCLE Fig. 2. Changes in the numbers of ejaculations per test received by each of the two hormone-treated females in a group (four females, eight male-female groups). The amplitudes of rhythms were massively increased when cycles were offset. Compared with synchronized cycles, females received more ejaculations during days 13-14 when these coincided with either days 19-20 or days 27-28 of the other female whose mating activity was coincidentally decreased. Each point is the mean of 14-16 tests. Symbols as in Figure 1. A two-way ANOVA comparing the data from females A and E with those from females B and F showed no significant differences between them during synchronized cycles or any of the offset cycles (females x cycle segment design, F[1,297-3001 = 0.0-2.5, N.S.), and also confirmed the analyses described above, namely, that the differences between cycle phases were significant in -7, + 7, and 214 offsets (F[9,297-3001 = 3.8-10.4, P < 0.0001) but were not significant in synchronized cycles (F[9,3001 = 1.4, N.S.). Data from all four females were therefore combined in Figure 3, where each point is the mean of 30-32 observations. Figure 3 also shows the corresponding changes in plasma estradiol levels. The enhancement of the behavioral rhythms that resulted from offsetting the cycles becomes very obvious. ANOVAs using a cycle segment x male x female design showed that there were no significant differences in the combined data between cycle phases in synchronized cycles (F[9,3041 = 1.7, N.S.) but that there were highly significant differences in offset cycles (F[9,301-304] = 5.6-15.1, P < O.OOOl), confirming the previous analyses. Other Measures of Sexual and Social Behavior Many other behavioral measures (for which we have the data) could be given the extended analysis accorded ejaculatory activity. But in the present context we have chosen to illustrate just three additional measures, which are particularly interesting and show dramatically increased rhythmicity during offset cycles (Fig. 4). In view of the rhythms in the numbers of ejaculations received by females during offset cycles, it was not surprising that there were similar rhythms (but opposite in direction) in times to first ejaculation received by females (upper panel). The time to the first male mounting attempt received by females (middle panel) may be regarded as a measure of male sexual motivation, and this time showed changes that were almost identical to times to first ejaculation. Proximity times (lower panel), a measure of consort bonding, only showed rhythms when cycles were offset, and these resembled those in ejaculations received by females. Behavioral Cycles in Rhesus Monkeys I 165 - 7 OCFSET SYNCHRONIZED I 7 14 21 28 I 7 14 21 +7 28 I DAYS OF ARTIFICIAL *x Plasma estrodlol levels of Illustrated female 7 OFFSET 14 21 f 14 OFFSET 28 I 7 14 21 28 CYCLE x----n Plorma rrtradlol levels of othu female Fig. 3. Upper part: combined data for numbers of ejaculations received by females showed that the pattern of change depended on both the hormonal status of the female herself and on the timing of the estradiol peak of the other hormone-treated female in the group. Each point is the mean of 30-32 observations.Lower part: shows the timing of the plasma estradiol peaks in these females. Each point is the mean of four plasma samples. Symbols as in Figure 1. There were significant differences between offsets for all 11measures of social and sexual interactions between males and females shown in Table 11. ANOVAs using a cycle segment x male x female design for each type of offset showed that the amplitudes of almost all the behavioral changes were greater during offset than during synchronized cycles. Differences between cycle phases were significant for 4 of 11 measures during synchronized cycles, but for 7 measures in -7 offsets, 8 measures in 7 offsets, and 10 measures in 214 offsets. + DISCUSSION If it is accepted that the artificial cycle is an adequate and valid model for the natural cycle, these results fully supported the hypothesis that, when given the opportunity, male rhesus monkeys will maintain stable levels of ejaculatory activity, whereas females will mate most frequently 24-48 hours after the estradiol peak a t the time of ovulation. During synchronized cycles, both the level of and pattern of change in ejaculations made by males were closely similar to those reported previously [Michael & Zumpe, 1970al for intact females in pair tests (Fig. 5), which, incidentally, gave a further behavioral validation of the artificial cycle. Ejaculations made by males were less rhythmic in offset cycles than in synchronized cycles, whereas the amplitudes of changes in ejaculations received by females were dramatically increased. Hormonal factors were clearly causal because in every case the maxima (minima for latency measures) followed the estradiol peak by 1-2 days. This was in good agreement with earlier studies on pairs in which females were undergoing either natural [Michael & Bonsall, 19771 166 I Michael and Zumpe SYNCHRONIZED - 7 OFFSET +7 OFFSET f 14 OFFSET 4 ? h 4 2 4 ru P ? I s Fig. 4., For latencies to the first ejaculation and first mounting attempt received by females, and for proximity times, offsetting the cycles of two females in a group greatly increased the amplitudes of changes in these behavioral measures of consort bonding. Each point is the mean of 30-32 observations. Upward arrows indicate the timing of the estradiol peaks of the females whose data are illustrated and downward arrows indicate those of the other females. Symbols a s in Figure 1. or artificial cycles [Michael et al., 19821. Beyond this, the enhancement of the rhythms in female mating activity in the presence of other females depended on the interaction between the hormonal status of the female herself and that of the other hormone-treated female in the group. Taken together with the male's tendendency toward relatively stable levels of mating activity, this interaction produced well-marked periods of maximum sexual activity in each female, periods that would be further accentuated in a more complex social setting involving several cycling females. Offsetting the cycles affected other measures of behavior that are involved in the formation and maintenance of consortships. Latencies to the first ejaculation and to the first male mounting attempt received by females, as well as numbers of mounting attempts, redirected aggressive gestures, approaches and grooming times, all became more rhythmic during offset cycles. A word about Proximity times which were completely nonrhythmic in synchronized cycles. This observation confirmed an earlier finding that proximity times did not change with hormone treatments of the female [Michael & Zumpe, 19841. The reasons for this somewhat surprising finding are complex but depend on an overriding effect of partner preferences, which create strong social bonds throughout the cycle and tend to mask any effects of changes in sexual motivation on proximity times. In offset cycles, social bonding was overridden by changes in sexual preference, and proximity times became highly rhythmic. "Significantly different from zero offset. Number of ejaculations received by female Time to first ejaculation received by female (seconds) Number of mounting attempts received by female Time to first mounting attempt received by female (seconds) Number of redirected aggressive gestures made by male Number of sexual invitations made by female Number of redirected aggressive gestures made by female Number of approaches by male to female Number of approaches by female to male Time spent in proximity (seconds) Time spent allogrooming - (seconds) 3.36 2 0.24 1136 C 54 6.75 ? 0.32 1588 5 76 4.19 5 0.45 2.12 2 0.25" 3.58 2 0.48 * 0.20 3.34 2.95 5 0.17 1184 i 51 779 t 45 6.55 t 0.33 1632 t 75 3.74 t 0.40 1.33 5 0.13 2.55 -+ 0.32 3.56 t 0.23 2.86 2 0.18 1263 -+ 52 881 t 45 * 0.32" 828 2 48 3.34 2 3.20 f 0.19" 1029 ? 53" 1358 ? 60 2.15 2.50 +- 0.16" 1.81 2 0.20 * 0.41 3.16 1.47 t 0.22 11.99 6.43 10.23 8.40 5.04 4.16 3.78 4.92 * 0.47 6.09 7.59 * 0.31" 5.00 7.48 F(3,1238) 1973 2 81" 5.37 2341 5 73 0.78 f 0.04 14 Offset N = 320 Mean -+ S.E. -t 1.66 ? 0.18 3.29 i 0.32 1863 C 79 5.78 2400 i 72 2104 2 70 0.70 2 0.04 2205 t 68 0.04" N = 318 Mean ? S.E. + 7 Offset 0.92 2 - 7 Offset N = 317 Mean -+ S.E. 0.79 -+ 0.40 Zero Offset N = 320 Mean -+ S.E. 0.0001 0.0001 0.0001 0.0001 0.002 0.006 0.01 0.0001 0.0001 0.002 0.0001 P Significance of differences between offsets TABLE 11. Comparisons of Levels of Behavior During Synchronized and Offset Artificial Menstrual Cycles in Female Rhesus Monkeys Observed in Small Social Groups (ANOVAs, Offset b y Male b y Female Design f o r All T y p e s of Cycles Combined) 168 I Michael and Zumpe I 7 14 21 26 DAYS OF CYCLE 0-0 0-0 Group tests: synchronized artificial cycles Pair tests: natural cycles (Michael 8 Zumpe, 1970) Fig. 5. Both the levels of, and patterns of change in, numbers of ejaculations made by males in group tests during synchronized artificial cycles were closely similar to those in pair tests during natural cycles with intact females [Michael & Zumpe,197OaJ.Group tests, N = 14-16. Pair tests, N = 46-60. Symbols as in Figure 1. The experimental design controlled for differences in levels of behavior between animals, in partner preferences, and in dominance rank, and also minimized the effects of long-term testing, although all these factors contributed to the variance. Wallen & Winston  reported that in a social setting the amplitudes of behavioral rhythms of intact females were greater, while levels of behavior were lower, than those in a pair setting. No distinction was made between the behavior of the male with the experimental female and his behavior with all females in the group, which are quite different entities in a social context. The inference, that mating in a group situation requires substantially greater hormonal stimulation, received no support from the present study in which the hormonal and social variables could be well-controlled. We take a different view, namely, that the presence of other females in different cycle phases greatly enhances the behavioral effects brought about by changes in the individual’s own plasma hormone levels. This sociohormonal integration of behavior is likely to be highly adaptive by ensuring the mating success of both the males and females in a social group. The data imply that effects of ovarian hormones on human sexual activity, notoriously difficult to document in Western societies, might best be found in cultures where polygynous mating systems are acceptable and the male has the opportunity to select between females in different phases of the menstrual cycle. CONCLUSIONS 1. In rhesus monkeys, the rhythms of female mating activity that are associated with changing levels of ovarian hormones during the menstrual cycle are massively enhanced by the presence of other females in different phases of the cycle. 2. This is a consequence both of the varying hormonal status of females and of the male’s tendency to maintain his mating activity at a stable level by changing sexual partners so as to consort with each female in turn as she nears midcycle. 3. This sociohormonal integration of behavior is a highly adaptive phenomenon that ensures that mating by both males and females is largely confined t o periods of maximum fertility. ACKNOWLEDGMENTS This work was supported by USPHS grant MH 19506, and general research support was provided by the Georgia Department of Human Resources. Both are gratefully acknowledged. Behavioral Cycles in Rhesus Monkeys / 169 REFERENCES Altmann, S.A. A field study of the sociobiology of rhesus monkeys (Mucacu muluttu). ANNALS OF THE NEW YORK ACADEMY OF SCIENCES 102:338-435,1962. Bercovitch, F.B. Time budgets and consortships in olive baboons (Pupio unubis). FOLIA PRIMATOLOGICA411180-190,1983. Bonsall, R.W.; Zumpe, D.; Michael, R.P. Menstrual cycle influences on operant behavior of female rhesus monkeys. JOURNAL OF COMPARATIVE AND PHYSIOLOGICAL PSYCHOLOGY 92~846-855, 1978. Bonsall, R.W.; Zumpe, D.; Michael, R.P. Computerized behavioral scoring and analysis system. PHYSIOLOGY AND BEHAVIOR 35:I (Software Package PAB-014-S85), 1985. Carpenter, C.R. Sexual behavior of free ranging rhesus monkeys (Mucacu muluttu).JOURNAL OF COMPARATIVE PSYCHOLOGY 33:113-162, 1942. Darwin, C. THE DESCENT OF MAN, AND SELECTION IN RELATION TO SEX. New York, Appleton, 1871. Dittus, W.P.J. The evolution of behaviors regulating density and age-specific sex ratios in a primate population. BEHAVIOUR 69:265-302, 1979. Ferin, M.; Antunes, J.L.; Zimmerman, E.; Dyrenfurth, I.; Frank, A.G.; Robinson, A.; Carmel, P.W. Endocrine function in female rhesus monkeys after hypothalamic disconnection. ENDOCRINOLOGY 101:16111620,1977. Hartman, C.G. Studies in the reproduction of the monkey Mucacus (Pithecus) rhesus, with special reference to menstruation and pregnancy. CONTRIBUTIONS TO EMBRYOLOGY 23:l-161, 1932. Hausfater, G. DOMINANCE AND REPRODUCTION IN BABOONS (PAPZO CYNOCEPHALUS):A QUANTITATIVE ANALYSIS. (Contributions to Primatology, 7). Basel, Switzerland, S. Karger, 1975. Herndon, J.G.; Turner, J.J.; Ruiz de Elvira, M.C.; Collins, D.C. Silent ovulation in rhesus monkeys ( M . muluttu): dissociation of hormonal and behavioral states. PHYSIOLOGY AND BEHAVIOR 40:665-672, 1987. Hess, D.L.; Resko, J.A. The effects of progesterone on the patterns of testosterone and estradiol concentrations in the systemic plasma of the female rhesus monkey during the inter-menstrual period. ENDOCRINOLOGY 92~446-453, 1973. Johnson, D.F.; Phoenix, C.H. Hormonal control of female sexual attractiveness, proceptivity, and receptivity in rhesus monkeys. JOURNAL OF COMPARATIVE AND PHYSIOLOGICAL PSYCHOLOGY 90~473-483,1976. Kaufmann, J.H. A three-year study of mating behavior in a free-ranging band of rhesus monkeys. ECOLOGY 46500-512, 1965. Knobil, E. On the control of gonadotropin secretion in the rhesus monkey. RECENT PROGRESS IN HORMONE RESEARCH 3011-46, 1974. Knobil, E. Control of ovulation in the rhesus macaque. Pp. 257-262 in ENDOCRINE ASPECTS OF REPRODUCTION. R.L. Norman, ed. New York, Academic Press, 1983. Lindburg, D.G. The rhesus monkey in North India: a n ecological and behavioral study. Pp 1-106 in PRIMATE BEHAVIOR: DEVELOPMENTS IN FIELD AND LABORATORY RESEARCH, Vol. 2 L.A. Rosenblum, ed. New York, Academic Press, 1971. Michael, R.P.; Bonsall, R.W. Peri-ovulatory synchronization of behaviour in male and female rhesus monkeys. NATURE 265: 463-465, 1977. Michael, R.P.; Herbert, J.; Welegalla, J. Ovarian hormones and the sexual behaviour of the male rhesus monkey (Mucucu muluttu) under laboratory conditions. JOURNAL OF ENDOCRINOLOGY 39: 81-98,1967. Michael. R.P.: Richter. M.C.: Cain. J.A.: Zumpe’, D.; Bonsall, R.W. Akificial men: strual cycles, behaviour and the role of androgens in female rhesus monkeys. NATURE 275:439-440,1978. Michael, R.P.; Saayman, G.S.; Zumpe, D. The suppression of mounting behaviour and ejaculation in male rhesus monkeys (Mucacumuluttu) by administration of progesterone to their female partners. JOURNAL OF ENDOCRINOLOGY 41 :421-43 1, 1968. Michael, R.P.; Welegalla, J. Ovarian hormones and the sexual behaviour of the female rhesus monkey (Mucucu muluttu) under laboratory conditions. JOURNAL OF ENDOCRINOLOGY 41:407-420,1968. Michael, R.P.; Zumpe, D. Rhythmic changes in the copulatory frequency of rhesus monkeys (Mucucu muluttu) in relation to the menstrual cycle and a comparison with the human cycle. JOURNAL OF REPRODUCTION AND FERTILITY 21:199-201, 1970a. Michael, R.P.; Zumpe, D. Sexual initiating behaviour by female rhesus monkeys (Mucucu muluttu) under laboratory conditions. BEHAVIOUR 36:168-186,1970b. Michael, R.P.; Zumpe, D. Interaction of so- 170 I Michael and Zumpe cial, spatial and hormonal factors on the behavior of rhesus monkeys (Mucucu mulutta). PRIMATES 25462-474, 1984. Michael, R.P.; Zumpe, D.; Bonsall, R.W. Behavior of rhesus monkeys during artificial menstrual cycles. JOURNAL OF COMPARATIVE PSYCHOLOGY 962375-885, 1982. Norman, R.L.; Resko, J.A.; Spies, H.G. The anterior hypothalamus: how it affects gonadotropin secretion in the rhesus monkey. ENDOCRINOLOGY 9959-71, 1976. Southwick, C.H.; Beg, M.A.; Siddiqi, M.R. Rhesus monkeys in North India, pp 111-159 in PRIMATE BEHAVIOR. I. DeVore, ed. New York, Holt, Rinehart and Winston, 1965. Trivers, R.L. Parental investment and sexual selection. Pp 136-179 in SEXUAL SELECTION AND THE DESCENT OF MAN. B. Campbell, ed. Chicago, Aldine, 1972. Wallen, K.; Goy, R.W. Effects of estradiol benzoate, estrone, and propionates of testosterone or dihydrotestosterone on sexual and related behaviors of ovariectomized rhesus monkeys. HORMONES AND BEHAVIOR 9:228-248, 1977. Wallen, K.; Winston, L.A. Social complexity and hormonal influences on sexual behavior in rhesus monkeys (Mucucu muluttu). PHYSIOLOGY AND BEHAVIOR 32: 629-637, 1984. Williams, G.C. ADAPTATIONS AND NATURAL SELECTION A CRITIQUE OF CURRENT EVOLUTIONARY THOUGHT. Princeton NJ, Princeton University Press, 1966. Wilson, M.E.; Gordon, T.P.; Collins, D.C. Variation in ovarian steroids associated with the annual mating period in female rhesus monkeys (Mucucu muluttu). BIOLOGY OF REPRODUCTION 27530-539, 1982. Winer, B.J. STATISTICAL PRINCIPLES IN EXPERIMENTAL DESIGN. New York; McGraw-Hill Book Company, 1971. Wright, K.; Collins, D.C.; Preedy, J.R.K. The use of specific radioimmunoassays to determine the renal clearance rates of estrone and 17P-estradiol during the menstrual cycle. JOURNAL OF CLINICAL ENDOCRINOLOGY AND METABOLISM 47: 1084-1091,1978. Zumpe, D.; Michael, R.P. Redirected aggression and gonadal hormones in captive rhesus monkeys (Macacu muluttu). ANIMAL BEHAVIOUR 18:ll-19, 1970. Zumpe, D.; Michael, R.P. Effects of ejaculations by males on the sexual invitations of female rhesus monkeys (Mucucu muluttu). BEHAVIOUR 60:260-277, 1977. Zumpe, D.; Michael, R.P. Relation between the hormonal status of the female and direct and redirected aggression by male rhesus monkeys (Mucucu rnulutta). HORMONES AND BEHAVIOR 12:269-279, 1979.