American Journal of Primatology 8:167-181 (1985) Effects of Ovarian Hormones on the Behavior of Captive Macaca fascicularis DORIS ZUMPE AND RICHARD P. MICHAEL Department of P.sychiatry, Emory University School of Medicine, Atlanta, and The Georgia Mental Health Institute. Atlan.ta To examine the effects of ovarian hormones on the behavior of female Macaca fascicularis and their male partners, daily 1-hr behavior tests were conducted while ovariectomized females were (1)untreated, (2) given estradiol benzoate (EB) (5 pg subcutaneously [s.c.]/day), (3) given estradiol benzoate together with increasing doses of progesterone (P)(5 mg, 10 rng, and 20 mg. s.c./day), and (4)given testosterone propionate (TP)(0.25 mg s.c.iday) (six pairs, 540 tests). Weekly blood samples were analyzed by radioimmunoassay for plasma hormone levels (81 samples). Estrogen treatment produced plasma estradiol levels similar to those of intact females during the late follicular phase of the menstrual cycle. Additional progesterone at the lowest dose produced plasma progesterone levels similar to or somewhat higher than those during the midluteal phase, while higher doses produced supraphysiological levels. Androgen treatment resulted in plasma levels well above the physiological range. Hormone treatments produced highly significant effects on the sexual, social, and aggressive interactions of the pairs. As in rhesus monkeys, estrogen increased male and female sexual activity, and increasing doses of additional progesterone reversed these effects. Unlike in rhesus monkeys, testosterone propionate increased both female sexual motivation (invitations) and also male sexual activity and ejaculatory performance. The direction of the hormone-dependent changes in grooming and aggressive interactions confirmed earlier results with intact females and indicated that aggressive interactions and male grooming times were highest, and female grooming times were lowest, when copulatory activity was at its height. Key words: Macaca fascicularis, estradiol, progesterone, testosterone, sexual behavior, grooming, aggression INTRODUCTION In a previous study [Zumpe & Michael, 19831, we compared the behavioral interactions of oppositely-sexed pairs of Macaca fascicularis under controlled laboratory conditions with those of M. mulatta tested by the same observers under Received August 4, 1984; revision accepted October 21, 1984 Address reprint requests to Dr. Richard P. Michael, Department of Psychiatry, Emory University School of Medicine, Post Office Box AF, Atlanta, GA 30322. 0 1985 Alan R. Liss, Inc. 168 I Zumpe and Michael identical circumstances. The results suggested that M. fascicularis would be a suitable species for studying the neuroendocrine mechanisms underlying the behavior of higher primates, since the behavioral interactions of the pair appeared to be strongly influenced by the changing hormonal status of the female during the menstrual cycle. In M. fascicularis, ejaculatory performance and thrusting rates were higher than in M. mulatta during the follicular phase of the menstrual cycle and declined more sharply after the expected time of ovulation. In addition, there was also evidence that female sexual motivation was consistently higher during the follicular than during the luteal phase of the cycle. However, there were some notable behavioral differences between the two species, and these were thought t o have their origins in the fact that M. fasuicularis is less sexually dimorphic than M. mulatta. In particular, a n analysis of their agonistic interactions demonstrated that female M. fascicularis were less dominated by their male partners than were female M. mulatta, and changes in the female’s motivational state were more readily expressed by changes in overt behavior in M. fascicularis than in M. mulatta. Thus, M. fascicularis, under our test conditions, showed larger changes in measures of female sexual motivation and smaller changes in measures of male sexual motivation during the menstrual cycle than did M. nulatta. The present study was undertaken to extend these findings in M. fascicularis to ovariectomized females by examining the behavioral effects of exogenous estradiol benzoate, progesterone, and testosterone propionate in doses comparable with those used previously in M. muZatta [Michael et al, 1968; Michael et al, 1972; Michael and Zumpe, 19771, whose body weights are about twice those of cynomolgus monkeys. METHODS Animals Four mature male (weighing 4.9-5.1 kg) and three adult female (weighing 2.63.7 kg) M. fascicularis were obtained as adults through dealers directly from the wild in Malaysia. Animals were housed in individual cages together in a room in which natural daylight was supplemented by artificial lighting, which was activated by a photocell controlled by the natural daylength. Temperature was maintained between 20°C and 24°C. Food consisted of Purina Monkey Chow, supplemented by vitamins, fresh fruit, and vegetables; water was available ad libitum. Operative Prncedures and Hormone Treatments All females were ovariectomized bilaterally through a midline subumbilical incision several weeks before the experiments started. Histological examination confirmed that ovariectomy was complete. All hormone treatments were given for periods of four weeks and were administered daily subcutaneously (s.c.) in oil at 0800 hr. However, the first injection of each treatment was given at 1200 hr after testing. In conformity with earlier studies on M. mulatta, treatments were given in the following order: (i) 5 p g estradiol benzoate alone; (ii) 5 pg estradiol benzoate together with 5 mg progesterone; (iii) 5 pg estradiol benzoate together with 10 mg progesterone; (iv) 5 p g estradiol benzoate together with 20 mg progesterone; (v) 5 pg estradiol benzoate alone; and after a n interval of 5 weeks, (vi) 0.25 mg testosterone propionate. New hormone treatments were started at 1200 hr on Fridays after that day’s behavioral test, and data collection began 3 days later on the following Monday. Collection of Plasma Samples and Steroid Assays Blood (3 ml) was obtained at 1600 h r once a week (Thursdays) from the untranquillized females that had previously been adapted to the sampling procedure. Since Hormones and Behavior in M . fascicularis I 169 the saphenous veins of females were small and not always suitable for repeated blood sampling, the tail vein was sometimes used instead. Plasma levels of estradiol, progesterone, and testosterone were analyzed in the same duplicate 0.5 ml plasma aliquots by a method that involved chromatography on Sephadex LH-20. Steroid concentrations were measured by radioimmunoassay with highly specific antisera, and results were calculated using a four-parameter mass-action equation and a computer. Inter- and intra-assay coefficients of variation were less than 12% (estradiol) and less than 6% (progesterone and testosterone), and water blanks read -2.6 to 2.6 pg (estradiol and testosterone) and 0 to 13.4 pg (progesterone). Plasma testosterone levels during androgen treatment were estimated by radioimmunoassay without chromatography as previously described for males [Bonsall et al, 19761. We report here on the last three samples from each treatment period when conditions were fully stabilized (81 samples). Behavioral Testing Procedures Observations were made during 60-min test sessions on oppositely-sexed pairs of animals behind one-way vision mirrors. Tests were conducted 5 days a week in special observation cages 1.19 m wide by 1.07 m deep by 1.14 m high, into which first the male and then the female were introduced a t the beginning of each test session. At all other times animals were caged singly. Each female was tested with each of two male partners on alternate test days. Two females and two males were tested in a cross-over design controlling for individual differences and partner preferences, while the third female (initially together with one that died) was tested with two other males (six pairs, 540 1-hr behavior tests). Definitions and Terminology The following measures of behavior, described in detail elsewhere [Zumpe & Michael, 19831, receive numerical treatment here: (1)number of ejaculations per test; (2) time to first ejaculation-time in seconds from the start of the test to the first ejaculation (in the absence of any ejaculations, a default value of 3,600 sec was assigned); (3) number of mounts per test; (4)number of male mounting attempts per test-sum of male-initiated mounts and attempts to mount by the male that were refused by the female; (5) number of male-initiated mounts per test; (6) latency to first male mounting attempt-time in seconds from the start of the test to the first male mounting attempt; (7) number of female sexual invitations per test; (8)number of female-initiated mounts per test; (9) latency to first female invitation-time in seconds from the start of the test to the first female sexual invitation; (10) male success ratio-percentage of male mounting attempts that initiated a mount (in the absence of any male mounting attempts, a default value of 100% was assigned for statistical purposes only); (11)male grooming time per test fsecf;(12) female grooming time per test (sec); (13) male direct aggression-number of aggressive gestures directed by the male a t the female; and (14) female direct aggression-number of aggressive gestures directed by the female at the male. Selection of Tests for Numerical Treatment For each pair of animals, the data from the last ten tests with the female ovariectomized and untreated (6 months post-ovariectomy) were compared with the data from each of the ten test periods (4weeks) that followed when females were treated with ovarian hormones. After estrogen withdrawal, and again after androgen withdrawal, 10 days were allowed to elapse between the last hormone injection and the first of the ten tests used to re-establish behavioral baselines. Thus, quantitative comparisons were made between data from ten tests per pair during each of "Single value. 0.25 mg TP Untreated Untreated 5pgEB + 5mgP 5ggEB + 10 mg P 5pgEB + 20 mg P 5 !G EB Ovariectomized 5 Pg EB * 10.5 & 3.59 85.0 k 28.60 95.8 k 29.44 81.9 k 14.79 71.8 k 19.12 194.9 k 53.94 9.5 4.79 17.3 k 2.71 3.2 1.62 26.1 i 12.64 127.8 f 44.78 68.3 & 5.37 58.2 k 4.99 48.5 & 8.54 65.7 & 6.22 18.5 & 5.59 18.6 & 13.00 29.6 & 6.13 No data No data No data 6.4 1.07 110.3 ?i 28.33 61.2 -t 11.62 37.3 t 2.29 44.6 k 11.21 48.6" Plasma Ez (pg/ml) Female 1 Female 2 Female 3 2.9 0.32 2.3 0.03 63.8 & 5.10 60.2 & 8.50 118.5 f 11.9 0.9 & 0.41 0.3 + 0.00 1.0 I 0.27 0.4 i 0.09 * 0.3 k 0.30 0.0 & 0.00 32.2 i 3.92 69.8 & 9.11 102.6 k 6.10 0.3 & 0.15 1.4 f 1.04 0.8 k 0.06 0.5 & 0.27 * * 0.8 f 0.22 1.1 0.12 17.6 +_ 2.65 49.4 & 8.94 68.6 f 6.85 0.9 k 0.15 1.2 f 0.46 0.4 & 0.12 1.1 0.20 Plasma P (ng/ml) Female 1 Female 2 Female 3 96.1 _+ 5.94 94.7 & 2.18 119.0 + 14.89 134.3 k 5.42 188.0 & 17.47 141.2 k 22.73 126.9 + 10.96 17934 i 2831 68.3 f 13.31 Female 1 80.2 & 2.44 100.8 k 11.04 118.3 5 4.77 119.5 k 8.50 147.9 & 7.70 73.6 k 3.87 130.0 I 12.50 5596 i; 791 251.1 _C 36.96 31.7 & 6.41 69.4 i 11.87 33.4 & 6.95 48.8 j~ 12.87 59.3 k 5.05 67.8 35.00 72.7 f 42.50 11519 + 719 63.9 k 13.55 Plasma T (pg/ml) Female 2 Female 3 TABLE I. Changes in Plasma Hormone Levels for Each of Three Ovariectomized iM. fasciculuris in Response to Different Hormone Treatments (Means + S.E.M. for Three Samples) L r 5n ! z s C N -4 0 . Y Hormones and Behavior in M. fasciculuris I 171 nine experimental conditions when the female was (1) untreated, (2) treated with estradiol benzoate, (3-5) treated with estradiol together with three increasing doses of progesterone, (6) treated with estradiol alone, (7) untreated (estrogen withdrawal), (8) treated with testosterone propionate, and (9)untreated (androgen withdrawal). We have also provided data for the purpose of comparison h o t analyzed statistically) from 149 one-hr behavior tests conducted during two menstrual cycles with each of the same six pairs of animals when the females were still intact. These latter data are a subset of those published previously [Zumpe & Michael, 19831. Statistical Treatment of Results The significance of the differences in behavior among the nine experimental conditions was assessed by standard analysis of variance for repeated measures using a treatment by animal-pair by test-order design. The Scheffi. test was used to make comparisons between means [Winer, 19711. These results are given in the tables, where underlining groups data not significantly different from each other. Latencies to the first male mounting attempt and to the first female invitation, which showed a bimodal distribution, were analyzed by the median test [Siegel, 19561. RESULTS Plasma Hormone Levels Table I gives the mean plasma estradiol, progesterone, and testosterone levels for each ovariectomized female during each treatment condition. Estrogen treatment resulted in plasma estradiol levels similar to those during the latter part of' the follicular phase of the menstrual cycle when these females were intact (about 100 pgiml). Additional treatment with 5 mg progesterone produced plasma progesterone levels that were similar to or higher than those occurring during the midluteal phase when these females were intact (5.0-16.0 ng/ml), and higher doses produced progesterone levels well above the physiological range. Testosterone propionate administration resulted in plasma testosterone levels greatly in excess of the physiological range for these females when intact (about 500 pgjml). Sexual Behavior Table I1 gives, for each of the nine treatment conditions, the means per test for numbers of ejaculations, time t n ejaculation, and numbers of' mounts for all six pairs combined. For the purpose of comparison, the top line in the table gives data from these pairs when the females were intact. Treating ovariectomized females with estradiol benzoate, progesterone, and testosterone propionate produced highly significant changes in all three measures of behavior. Estradiol restored ejaculatory and mounting activity, and the additional administration of increasing doses of progesterone progressively reversed these effects. When progesterone was withdrawn, behavior reverted to the levels observed during treatment with estradiol alone. Testosterone propionate produced significant changes in all three behavioral measures. In summary, the sexual performance of male M. fascicularis was enhanced by treating their female partners with both estradiol and testosterone and was progressively suppressed by increasing doses of progesterone. Examination of data from individual pairs of animals revealed that the behavioral effectiveness of the various hormone treatments depended very much on the identity of the female of the pair. This is illustrated in Figure 1, which shows changes in the ejaculations of males paired with different females. It can be seen that with female 1 there was a major effect with estrogen but a trivial effect with both progesterone and testosterone. With female 3, both estrogen and progesterone P F8,40 Intact (1) Ovariectomized (2) 5 p g EB (3) 5 p g E B + 5 m g P (4) 5 pgEB t 1 0 m g P (5) 5 p g EB + 20 mg P (6) 5 p g EB (7) Untreated (8) 0.25 mg TP (9) Untreated S.E. * 2.2 k 0.07 0.9 k 0.11 2.4 k 0.10 1.8 k 0.14 1.4 + 0.15 1.2 i 0.13 2.3 k 0.12 1.0 0.10 1.6 f 0.10 0.8 k 0.10 12.1 < 0.001 9 1 7 5 4 8 3 6 2 149 60 60 60 60 60 60 60 60 60 N No. of ejaculations Mean 149 60 60 60 60 60 60 60 60 60 N * 448.8 & 77.83 1785.9 209.12 95.2 f 49.37 874.7 i 177.16 1266.8 212.73 1476.3 k 218.36 414.4 k 139.28 1505.1 f 201.52 635.9 i 139.50 1874.1 i 213.35 4.8 < 0.001 2 6 8 3 4 5 7 1 9 Time to ejaculation (sec) Mean k S.E. * * 12.3 k 0.53 5.8 0.53 9.8 k 0.62 7.3 k 0.67 5.4 f 0.61 4.6 0.59 8.6 0.65 4.4 k 0.47 8.5 i 0.60 3.9 k 0.47 4.3 < 0.001 9 7 5 4 1 3 8 6 2 149 60 60 60 60 60 60 60 60 60 N No. of mounts Mean & S.E. TABLE 11. Effects on Male Sexual Performance of Treating Ovariectomized M . fasciculuris With Gonadal Hormones (Six Pairs) k z$ P 5a 9 N . !si h3 Hormones and Behavior in M. farsciculuris I 173 I FEMALE I c, G h 8 2I I o No. of tests. 23 I 0 10 I0 1010 10 10 10 10 10 10 10 I 0 0 10 10 10 10 I0 24 10 :O 10 : O K 10 10 10 10 28 10 10 101010 10 10 I0 10 23 FEMALE 2 2 N o . o f tests. 23 10 10 I 0 0 I0 10 10 LO FEMALE 3 No. of t e s t s , 28 I0 I0 1010 10 10 10 10 10 HORMONAL STATUS OF FEMALES. 0 Intact Untreated 5 pg E B 2 5 mg T P 5 p g EB + 5, 10 and 20 rng P respectively DO Fig. 1. The effects on the ejaculatory performance of male M. foseicrtlnris of' treating nvariectomized females with exogenous hormones. With female 1 (top) estrogen treatment (EB) increased ejaculations, hut additional progesterone (P)and testosterone (TP)had rather weak effects. With female 2 (middle) both EB and P had very strong behavioral effects, while with female 3 (butturn) the effects of TP were particularly obvious. Vertical bars give standard errors of the means. P F8,40 Intact (1) Ovariectomized (2) 5 pgEB (3) 5 p g E B + 5 m g P (4) 5 pg El3 + 10 mg P (5) 5 pg ER + 20 mg P (6) 5 pg EB (7) Untreated (8) 0.25 mgTP (9) Untreated 9.8 k 0.59 5.2 & 0.60 8.6 k 0.70 6.1 + 0.58 4.5 k 0.50 3.2 k 0.39 7.0 & 0.84 4.4 k 0.67 6.5 k 0.77 3.2 k 0.49 3.2 < 0.01 5 9 7 4 1 3 8 6 2 149 60 60 60 60 60 60 60 60 60 N No. of male mounting attempts Mean k S.E. 46.0 k 13.85 504.1 k 145.52 13.0 i 3.22 107.9 k 68.46 627.0 + 181.11 811.6 k 198.27 140.6 & 87.49 560.4 k 161.89 121.2 & 70.58 660.5 182.48 1.8 N.S. 2 3 8 6 1 7 4 9 5 140 49 57 56 51 48 53 47 52 48 N Latency of first male mounting attempt (sec) Mean k S.E. 149 60 60 60 60 60 60 60 60 60 N * 9.3 k 0.55 4.1 k 0.46 7.9 i 0.62 5.2 k 0.51 3.5 & 0.44 2.6 0.39 6.1 k 0.68 2.3 i 0.27 5.2 i 0.51 1.6 0.22 4.1 < 0.005 9 7 5 4 1 3 8 6 2 No. of male-initiated mounts Mean 5 S.E. TABLE 111. Effects on Male Sexual Motivation of Treating Ovariectomized M . fuscieuZurisWith Gonadal Hormones (Six Pairs) E 1 zi;’ B lu i2 3 F Ib . s Hormones and Behavior in ill. fascieularis I 175 exerted massive effects, while that of testosterone was quite minor. With female 3 , all three hormones exerted quite well-marked effects. Male and Female Sexual Motivation Table I11 gives, for each of the nine treatment conditions, the means per test for three measures of male sexual motivation (mounting attempts, latency to the first mounting attempt, and male-initiated mounts) for all six pairs combined. There were significant treatment effects for mounting attempts and male-initiated mounts, which increased during estradiol and testosterone treatments and declined during treatment with even the lowest dose of progesterone. Latencies to the first male mounting attempt did not reach statistical significance by analysis of variance because of high within- and between-pair variance and a bimodal distribution. Because of this, the data were re-analyzed by median test (chi-square = 49.2, df 8, P < 0.001). Latencies shortened significantly with estradiol and testosterone, and lengthened progressively with increasing doses of progesterone. Table IV gives, for each of the nine treatment conditions, the means per test for four measures of female sexual motivation (female sexual invitations, latency to the first female invitation, female-initiated mounts, and male success ratio) for all six pairs combined. While these measures showed trends similar to those for male sexual motivation, results were not strongly significant and were not significant at all for two indices by analysis of variance (median test on latencies to the first female invitation, chi-square = 18.4, df 8, P < 0.02), suggesting that female sexual motivation was less affected by the hormone treatments. However, examination of data from individual pairs showed that this was not quite so; effects depended upon the identity of the female. This is illustrated in Figure 2, which shows data for females 2 and 3 from tests with male 4.With both females there were systematic, hormone-related changes in male mounting attempts. Female 2, which received consistently more mounting attempts than female 3, made few sexual invitations and many refusals, so that there were marked hormone-dependent changes in male success ratios (Fig. 2, middle). Female 3, on the other hand, received fewer mounting attempts and made relatively few refusals, so that male success ratios were high throughout (Fig. 2, middle). However, this female made large numbers of sexual invitations, which increased with both estrogen and androgen treatment and declined with progesterone administration; these contributed importantly to the changes in ejaculatory activity (Fig. l., bottom right). Both females, therefore, demonstrated hormone-dependent changes in sexual motivation but differed in the behavioral mechanisms by which they were expressed: by refusing in female 2 and by inviting in female 3 . Grooming and Agonistic Behavior Table V gives, for each of the nine treatment conditions, the means per test for male and female grooming times and male and female aggression for all six pairs combined. There were significant hormone effects on both male and female grooming times and on male aggression. Estrogen alone significantly decreased male grooming times and increased female grooming times, and these effects were reversed when females received 10 mg progesterone in addition. Androgen, like estrogen, tended to decrease male grooming and increase female grooming, but these changes were not statistically significant. Estrogen also significantly increased male aggression; additional progesterone, even at the lowest dose, resulted in a significant decline that was reversed when progesterone was withdrawn. Examination of data from individual pairs revealed a large amount of variability, although most pairs reflected the changes outlined here. Figure 3 gives the results for one representative P F8,40 + + + Intact (1) Ovariectomized (2) 5 p g EB (3) 5 p g E B 5 m g P (4) 5 pg EB 10 mg P (5) 5 gg EB 20 mg P (6) 5 pgEB (7) Untreated (8) 0.25 mg TP (9) Untreated Mean f S.E. 149 60 60 60 60 60 60 60 60 60 + 19.2 + 1.63 5.9 f 0.64 9.0 i: 1.14 7.8 f 0.88 6.9 f 0.80 7.3 & 0.73 8.4 k 0.96 7.8 k 0.92 10.2 1.16 6.2 f 0.69 2.2 < 0.05 1 9 4 5 3 7 6 2 8 N No. of female sexual invitations N.S. 376.2 i 117.49 572.5 f 156.39 10.3 5.74 430.8 i 155.92 656.6 C! 197.93 835.2 k 197.56 437.6 f 205.44 302.5 i 115.20 275.2 i 152.83 385.9 i 122.39 1.9 Mean k S.E. 2 8 7 9 3 6 1 4 5 43 52 14 34 43 52 30 46 30 56 N Latency to first female sexual invitation (sec) 3.0 0.29 1.7 0.25 1.9 i: 0.28 2.1 0.29 1.9 & 0.28 2.1 i 0.28 2.5 f 0.32 2.1 k 0.33 3.3 0.46 2.3 & 0.35 2.3 < 0.05 1 2 4 5 7 3 9 6 8 149 60 60 60 60 60 60 60 60 60 No. of female-initiated mounts N Mean + S.E. 97.1 0.62 84.6 & 3.24 93.7 k 1.50 87.2 i 3.52 83.2 f 4.56 74.8 -t 5.83 93.0 f 2.11 72.5 + 5.08 89.9 f 2.17 65.1 k 6.02 2.0 N.S. 9 7 5 4 1 3 8 6 2 147 58 59 58 54 51 59 54 58 50 N Male success ratio Mean & S.E. TABLE IV. Effects on Female Sexual Motivation of Treating OvariectomizedM . fmciculariv With Gonadal Hormones (Six Pairs) h 1 s i;' BEU fi 8 N s 3. Intact + P F8,40 (5) 5 pg EB + 20 mg P (6) 5 p g E B (7) Untreated (8) 0.25 mg TP (9) Untreated (1) Ovariectomized (2) 5 pg EB (3) 5 p g E B + 5 m g P (4) 5 fig EB 10 mg P 149 1575.4 47.14 60 1270.7 f 79.65 60 2085.2 & 76.78 60 1860.5 k 105.61 60 1686.1 i 96.49 60 1687.5 & 79.30 60 1752.1 k 96.77 60 1612.7 k 74.44 60 1833.2 k 110.14 60 1602.0 93.54 2.6 < 0.025 1 9 7 4 5 6 8 3 2 169.5 k 28.27 788.3 f 96.79 495.4 f 71.66 688.7 & 87.13 1066.1 & 99.84 982.7 k 90.84 618.4 k 77.84 821.6 -t 96.56 674.3 k 77.17 797.8 i 104.77 4.5 < 0.001 2 6 8 3 1 9 7 5 4 149 60 60 60 60 60 60 60 60 60 Female grooming time (sec) N Mean k S.E. Male grooming time (sec) N Mean S.E. * 2.3 0.27 0.0 f 0.00 1.1 k 0.28 0.1 k 0.04 0.1 0.03 0.1 0.04 1.3 k 0.30 0.0 i 0.02 0.4 i 0.13 0.1 k 0.09 4.8 < 0.003 1 7 4 5 3 9 8 2 6 149 60 60 60 60 60 60 60 60 60 Male direct aggression N Mean i S.E. N.S. * * 1.6 k 0.24 0.7 k 0.41 1.1 & 0.30 1.3 f 0.35 1.6 0.50 1.1 i 0.31 1.2 k 0.32 1.3 0.36 0.6 k 0.12 0.7 k 0.19 0.9 8 9 1 5 2 6 3 7 4 149 60 60 60 60 60 60 60 60 60 Female direct aggression N Mean + S.E. TABLE V. Effects of Treating Ovariectomized M . fascicuZarisWith Gonadal Hormones on Grooming and Aggressive Interactions With Their Male Partners (Six Pairs) . 2 5. 178 I Zumpe and Michael MALE 4 FEMALE I 2 I f T 20 10 ,mm 0 N o of tests. 24 10 10 10 10 10 10 10 10 10 28 10 I0 10 10 IC I0 10 li IC HORMONAL STATUS OF FEMALES: a5 pg .Intact 0Untreated m 5 pg E B EB t 5, 10 and 20 rng P reSp8CtlVeIy DO 25 rng T P Fig. 2. “ k o different mechanisms by which two females, both paired with the same male, expressed hormone-dependent changes in sexual motivation. Female 2 (left), receiving many male mounting attempts, showed hormone-dependent changes in receptivity (changes in male success ratios); female 3 (right), receiving fewer male mounting attempts, generally accepted them but showed hormone-depcndcnt changes in proceptivity (changes in sexual invitations). Vertical bars give standard errors of the means. pair and shows that the male directed most aggression at the female, and groomed her least, when she was treated either with estrogen or with androgen, that is, during treatments when copulatory activity was maximal (Fig. 1). In contrast, female grooming times were shortest when copulatory activity was minimal. DISCUSSION The present study, as well as the earlier one [Zumpe & Michael, 19831, using M. fmcicularis stemmed from the fact that feral-reared M. ntulatta are generally less Hormones and Behavior in M. fasciculuris I 179 MALE 4 WITH FEMALE 3 1000 7 NO. of tests 1 I 28 10 10 10 10 10 10 10 10 10 28 10 10 10 1010 10 10 10 10 HORMONAL STATUS OF FEMALE Intact 5 y g EB t nuntreated 5 , 10 and 20 rng P respectively 59g EB 00 - 2 5 mg TP Fig. 3. Data from one pair illustrating hormone-dependent changes in grooming and aggressive interactions. During hormone treatments that produced high levels of ejaculatory activity (EB and TP), male grooming times were short, and aggression, especially by the male, was high. Female grooming times tended to increase during treatments producing high ejaculatory activity. Vertical bars b<ve standard errors of the means. available for behavioral studies now than in the past. All the results reported here confirm observations made upon intact females paired with males during normal menstrual cycles. Although only six pairs were used, the results were remarkably clear-cut for a behavioral project involving primates, and further replications with more animals seemed unwarranted at this stage. Despite significant differences between individual pairs (Fs,72 = 6.4-241.3, P < 0.001), hormone treatments had very predictable and highly significant effects on the sexual, social, and aggressive interactions of the pair (Tables 11-V). As in rhesus monkeys, estrogen increased ejaculatory activity and measures of both male and female sexual motivation, while additional progesterone reversed these effects. In fact, M. fusciculuris appeared to be more sensitive than rhesus monkeys to progesterone. With 2 pgikg estradiol benzoate (EB), most of the behavioral effects of progesterone were established at 2 mg/ kg, whereas in rhesus monkeys receiving 1 K g k g EB, 5 mgikg progesterone was needed [Michael et al, 19681. Changes in male sexual motivation were more conspicuous throughout than were changes in female sexual motivation, and this finding was in line with results obtained in rhesus monkeys. The old observation in rhesus monkeys [Michael & Welegalla, 19681 that two behavioral mechanisms exhibited by 180 I Zumpe and Michael females contributed to the decline in male ejaculatory activity during the luteal phase was readily confirmed in M. fascicularis. The first mechanism depended upon loss of female receptivity (increased refusals), and the second depended upon the loss of female proceptivity (decreased invitations). This leaves out the question of declining male motivation, which could involve pheromones, as large amounts of vaginal aliphatic acids are produced by this species (Michael, unpublished results). However, several aspects of behavior do differ in the two species. First, levels of sexual activity with ovariectomized, untreated, wild-caught females were generally considerably higher in M. fascicularis, and 1-2 ejaculations per test were quitc common; thus, baselines were higher than in M. mulatta [Michael, 19681. Of course, the copulatory pattern in this species appears to consist mainly of a single intromission [Shively et al, 1982; Zumpe & Michael, 19831, and when females were intact, M. fascicularis males were more potent (more ejaculations per test) than M. mulatta males. Second, as described above, the inhibitory effects of progesterone on behavior were more marked in M. fascicularis. Third, treating ovariectomized rhesus females with androgen usually does not increase male ejaculatory performance [Michael el al, 1972; Michael & Zumpe, 19771, but in M. fascicularis both male mounting attempts and ejaculations were increased. This could be because increased female invitational behavior in this less sexualIy dimorphic species more effectively influences the male’s behavior. Fourth, male grooming times decreased and female grooming times increased when sexual interactions increased, whereas rhesus males tend to groom more when sexual interactions are high. Fifth, direct aggression by male M. fascicularis was significantly higher when females were treated with estrogen alone than during any other treatment condition, whereas male rhesus monkeys show little direct aggression but high levels of redirected aggression with estrogenized partners [Michael & Zumpe, 1970; Zumpe & Michael, 19791. Again, these results are in good agreement with observations on intact M. fascicularis, in which male aggression tended to be highest around midcycle. Finally, treating ovariectomized females with 5 pg estradiol benzoate alone and in combination with 5 mg progesterone produced plasma hormone levels in the physiological range as judged by our own studies and by those of others [Goodman et al, 1977; Goodman & Hodgen, 19791. The plasma hormone data should be helpful in determining physiological dose-ranges for future experiments, and the observation that the same hormone doses gave different plasma hormone levels in different animals is consistent with recent data from rhesus monkeys [Michael et al, 19841. CONCLUSIONS 1. Administering estradiol benzoate, progesterone and testosterone propionate to ovariectomized female M. fascicularis had highly significant effects on their sexual, social, and aggressive interactions with male partners. 2. As in M. rnulntta, estradiol increased ejaculations and male and female sexual motivation in M. fascicularis, while additional progesterone reversed these effects. 3. In contrast to results in M. mulatta, treating M. fascicularis females with large doses of testosterone increased ejaculatory activity, and M. fascicularis appeared to be more sensitive to the effects of progesterone than did M. mulatta. 4. Other differences between M. fascicularis and M. mulatta in their behavioral responses to hormones confirm previous observations on intact females during the menstrual cycle. 5. Thus, M.fascicularis is a suitable alternative to M. mulatta as a model for investigating the neuroendocrine mechanisms underlying the behavior of higher primates. Hormones and Behavior in In. fasciculark I 181 ACKNOWLEDGMENTS This w o r k was supported b y USPHS grant MH 19506, and general research support was provided by the Georgia Department of Human Resources. Both are gratefully acknowledged. P r e l i m i n a r y data were presented at the IXth Congress of the International Primatological Society, August 8-15,1982, in Atlanta, Georgia. REFERENCES Bonsall, R.W.; Baumgardner, D.G.; Michael, R.P. A computerized semi-automated radioimmunoassay for plasma testosterone. JOURNAL OF STEROID BIOCHEMISTRY 79353-858,1976. Goodman, A.L.; Hodgen, G.D. Menstrual cycle characteristics in chronically hemiovariectomized cynornolgus monkeys (Macucu fasciculuris). JOURNAL OF CLINICAL ENDOCRINOLOGY AND METABOLISM 48:345-347,1979. Goodman, A.L.; Descalzi, C.D.; Johnson, D.K.; Hodgen, G.D. Composite patterns of circulating LH, FSH, estradiol and progesterone during the menstrual cycle i n cynomolgus monkeys. PROCEEDINGS OF THE SOCIETY FOR EXPERIMENTAL BIOLOGY AND MEDICINE 155:479-481, 1977. Michael, R.P. Gonadal hormones and the control of primate behaviour, pp 69-93 in ENDOCRINOLOGY OF HUMAN BEHAVIOUR. R.P. Michael, ed. Oxford University Press, Oxford, 1968. Michael, R.P.; Welegalla, J. Ovarian hormones and the sexual behaviour of the female rhesus monkey (Mucaca muluttu) under laboratory conditions. JOURNAL OF ENDOCRINOLOGY 41:407420,1968. Michael, R.P.; Zumpe, D. Aggression and gonadal hormones in captive rhesus monkeys (Mucacu rnulutta). ANIMAL BEHAVIOUR 18:1-10,1970. Michael, R.P.; Zumpe, D. Effects of androgen administration on sexual invitations by female rhesus monkeys (Mucucu muluttu). ANIMAL BEHAVIOUR 25:936-944.1977. Michael, R.P.; Bonsall, R.W.; Zumpe, D. The behavioral thresholds of testosterone in castrated male rhesus monkeys (Mucaca muluttu). HORMONES AND BEHAVIOR 18:161-176,1984. Michael, R.P.; Saayman, G.S.; Zumpe, D. The suppression of mounting behaviour and ejaculation in male rhesus monkeys (Mucucu rnuluttu)by administration of progesterone to their female partners. JOURNAL OF ENDOCRINOLOGY 41:421-431,1968. Michael, R.P.; Zumpe, D.; Keverne, E.B.; Bonsall, R.W. Neuroendocrine factors in the control of primate behavior. RECENT PROGRESS IN HORMONE RESEARCH 28:665-706, 1972. Shively, C.; Clarke, S.; King, N.; Shapiro, S.; Mitchell, G. Patterns of sexual behavior in male macaques. AMERICAN JOURNAL OF PRIMATOLOGY 2:373-384, 1982. Siegel, S. NONPARAMETRIC STATISTICS FOR THE BEHAVIORAL SCIENCES. New York, McGraw-Hill, 1956. Winer, B.J. STATISTICAL PRINCIPLES IN EXPERIMENTAL DESIGN. New York, McGraw-Hill, 1971. Zumpe, D.; Michael, R.P. Relation between the hormonal status of the female and di rect and redirected aggression by male rhesus monkeys (Macucu rnuluttu). HORMONES AND BEHAVIOR 121269279,1979. Zumpe, D.; Michael, R.P. A comparison of the behavior of Mucam fiwciculuris and Mucucu rnuluttu in relation to the menstrual cycle. AMERICAN JOURNAL OF PRIMATOL OGY 4:55-72, 1983.