American Journal of Primatology 30169-174 (1993) Milk Secretion in Pregnancy Among Free-Ranging Japanese Macaques ICHIROU TANAKA', SHIN-ICHI HAYAMA', AND HIDE0 NIG12 'Department of Anthropology, Faculty of Science, The University of Tokyo; 'Division of Wild Animal Medicine, Nippon Veterinary and Animal Science University, Musashino, Tokyo, Japan Milking under anesthesia in pregnant free-ranging Japanese macaques (Macaca fuscata) directly revealed lactation in gestation at Jigokudani Monkey Park, the Shiga Heights, Nagano Prefecture, Japan, from 12 to 14 February 1992. Multiparae secreted milk at 76-97 days of estimated fertilization age when the birth intervals t o the next offspring were 2 years. The observation of sucking behavior from February 1991 to March 1992 indicated that concurrent suckling by these multiparae terminated approximately 70 days before the next parturition after the growth of fetuses had accelerated and the embryos survived the crisis of abortion. Thus, Japanese macaque mothers appear to hedge maternal investment with concurrent lactation against possible miscarriage. Two nulliparous pregnant females secreted milk 3 months before the first parturition although they had no suckers. The first preparation of lactation appears to require the duration of longer than 3 months in nulliparae although worked mammary glands appear to be able to resume within 1month in multiparae. 0 1993 Wiley-Liss, Inc. Key words: Japanese macaque, duration of lactation, pregnancy, parity, suckling behavior INTRODUCTION Most Old World primates usually give birth to single young, which develop slowly and have a long period of breast-feeding [review in Hearn, 19841. The pattern of changes in suckling behavior, however, had not been determined in primates [Hearn, 19841.Although Hearn [19841conjectured that rhesus macaques mothers ceased to suckle their infant before entering the next pregnancy, the observation of suckling behaviour in free-ranging Japanese macaques suggested lactation in pregnancy [Tanaka, 19921.We investigated lactation in gestation with milking under anesthesia and estimated the duration of lactation. METHODS We temporarily (for 2 or 3 h) captured free-ranging Japanese macaques (Macaca fuscata) in Jigokudani Monkey Park, the Shiga Heights, Nagano Prefec- Received for publication J u n e 9, 1992;revision accepted February 9,1993. Address reprint requests to Dr. Ichirou Tanaka, Department of Anthropology, Faculty of Science, The University of Tokyo, Hongo 7-3-1, Bunkyo-ku, Tokyo, 113 Japan. 0 1993 Wiley-Liss, Inc. 170 / Tanaka et al. ture, Japan, from 12 to 14 February 1992. There is a distinct mating season in autumn from November to early December and a distinct birth season in spring from late April to June. After the injection of two anesthetics (xylazine hydrochloride and ketamine hydrochloride) intramuscularly, we diagnosed pregnancy with ultrasonography. Then we wrung the nipples of pregnant female macaques. As soon as milk seeped, we stopped squeezing breasts to minimize the influence of milking on sucking infants. Therefore we did not use oxytocin. After diagnosis and milking, we injected intravenously the antagonist (tolazoline hydrochloride) of the former anesthetic into macaques to recover from anesthesia quickly. We had no sooner released captured monkeys than they awoke. None of the anesthetics or the antagonist affects fetal development. The suckling behavior of a free-ranging troop of Japanese macaques was studied in Jigokudani Monkey Park from February 1991 to March 1992. The monkeys can be observed within a distance of 30 cm at closest approach. Using the focalanimal sampling (continuous recording) method [Martin & Bateson, 19861 for mother-infant dyads, we recorded when the infant began to suck its mother’s nipple and when it stopped. Based on the local rate of sucking (the number of sucks per second), we discriminated nutritive sucking (sucking with milk flow) from non-nutritive sucking (sucking without milk flow) because nutritive sucking occurred a t a constant low local rate; non-nutritive sucking occurred a t a n increasing and high local rate [Tanaka, 19921. To measure the local rate of sucking, we used 8 mm video cameras (SONY CCD-TR75 and CCD-TR705). During the video playback of suckling, we calculated local rates and distinguished nutritive sucking from non-nutritive sucking. In Japanese macaques, in the second phase of suckling (after infants were 7 months old), suckling occurred cyclically with the mean cycle length of 2 h, 2 min, and 58.5 seconds (mean 5 SEM = 7,378.5 481.2 seconds, N = 36), whereas the estimated standard deviation is 2,807.0 seconds. A cycle length is defined as the time from a n increase in sucking rate (the end of nutritive sucking) to a n increase in sucking rate in the next sucking bout [Tanaka, 19921, because infants sucked milk that had been secreted by and stored in mammary glands. We analyzed only the temporal lengths in which the previous nutritive sucking bout, the interval from the previous bout to the bout, and the nutritive sucking bout were observed continuously. We excluded from analysis those observations where we could not distinguish the local rate of sucking because the mother tightly embraced and hid her infant. Because our records of suckling cycle length were complete data, survival analysis [Nelson, 19821 estimated that the 99 percentile of the cycle length distribution was 3 h , 51 min, and 48 seconds (13,908 seconds). If no nutritive sucking of a n infant to its mother is observed during 5 h of continuous focal sampling which includes a t least two normal suckling cycles, suckling was considered finished. As the criterion of suckling cessation, we used not 4 h (just beyond the 99 percentile) but 5 h because, in the different phase (the first phase), there was a n outlier whose cycle length was 4 h, 29 min, and 39 seconds [see Tanaka, 19921 in which a n infant continued whimpering 3 h after the previous nutritive sucking bout. This frustration behavior also suggested that the duration of more than 3 h was abnormally long for the suckling cycle. We continually observed the target mother-infant dyad from birth for at least continuous 5 h. We defined the duration of suckling a s the age of the target infant between the last observation with nutritive sucking and the first observation without nutritive sucking. As to the technical terms, we used lactation in physiological phenomena based on milk secretion and suckling in behavioral phenomena based on the observation of nutritive sucking by infants. * Macaque Lactation in Gestation / 171 TABLE I. Milking Under Anesthesia in Pregnant Japanese Macaques Mother Name Toimu Mesumu Taiko Mekako Tachiyo Narumi Kerimi Tomano Ringokamikiri Dokuga Mesuguro Tobira Knife Nadeshiko Momiji Meakan Tarai Age (years) Parity 5 5 6 6 7 8 8 10 11 11 11 12 12 13 13 15 12 0 0 1 1 1 1 2 3 3 3 3 4 4 4 4 5 3 Milk secretion Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Embryo Interbirth interval (years)" Days before parturition 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 88 103 78 88 85 92 81 79 93 97 81 88 76 87 93 77 73 Estimated fertilization age (days)b 85 70 95 85 88 81 92 94 80 76 92 85 97 86 80 96 100 "Interbirth interval (years) is a period from the parturition of the suckling infant (except Tarai) to the parturition of the embryo. In Tarai, interbirth interval is a period from the parturition of the previous infant to that of the embryo. bEstimated fertilization age (days) is calculated by taking away days before parturition from 173 because the mean gestation length is 173 2 1.8 (SEM) days among Japanese macaques [Nigi, 19761. RESULTS In 16 pregnant female Japanese macaques milk secretion was revealed directly in milking under anesthesia (Table I). Two nulliparous pregnant females (named Toimu and Mesumu) secreted milk (70 and 85 days after estimated fertilization) 3 months before the first parturition. Multiparous lactation lasted until approximately 100 days into gestation: milk secretion was documented in 14 females examined 70-97 days after estimated fertilization but was not present in one female examined 100 days after fertilization (Table I). In multiparae, interbirth intervals to the subsequent offspring (the fetuses) varied our results: when interbirth intervals from suckers to fetuses were 2 years, pregnant mothers (N = 14) secreted milk, and when an interbirth interval was 3 years, a pregnant mother (N = 1)did not secrete milk. A month after milking, Dokuga, one of the milking subjects, suckled her 658day-old infant on March 13, 1992, 68 days before the next parturition, and no suckling was observed during 5 h and 20 min of continuous focal sampling on March 19, 1992, 62 days before parturition. A pregnant mother named Namako suckled her 640-day-old infant on February 15, 1991, 110 days before the next parturition, and termination of suckling was confirmed on April 6, 1991, 60 days before parturition. DISCUSSION Milking under anesthesia revealed lactation in gestation in 16 Japanese macaque females. Fourteen multiparae had embryos and suckled their infants simultaneously. Two nulliparae secreted milk although they had no sucker. Milk 172 / Tanaka et al. secretion in pregnancy appears to differ between multiparous and nulliparous mothers, because multiparous suckling behaviour ceased at late gestation when the interbirth intervals were 2 years. Moreover, a pregnant multipara did not secrete milk a t the 100 days of estimated fertilization age when the interbirth interval was 3 years. Two multiparous mothers ceased suckling just before conception when the birth intervals were 3 years [Tanaka, 19921. Thus, in multiparae with 2-year-old infants lactation appears to terminate before gestation. Even if multiparous suckling (the second phase with a low level of milk energy output) lasted until parturition when the next offspring was born in the following year, the other phase of suckling (the third phase with a high level of milk energy output) occurred a month before parturition [Tanaka, 19923. These cessations of suckling behavior show that multiparous lactation in gestation appears to be the extension of suckling for the previous offspring; nulliparous lactation in gestation appears to be the expansion of preparation for the first neonates. Milk secretion in two nulliparous females indicated that nulliparae prepared for lactation at least 3 months before parturition although multiparae prepared for lactation only a month before parturition because of the timing of the third phase of lactation in Japanese macaques [Tanaka, 19921. Thus, the first preparation of milk secretion appears to need longer duration than the next. Once mammary glands worked, lacation might resume only within a month. Milking under anesthesia in pregnant free-ranging Japanese macaques indicates that lactation lasted until approximately 100 days into gestation when the next offspring was born after 2 years. Besides the above two cases of cessation of suckling (60 and 62 days before parturition), two terminations of suckling were confirmed a t 77 and 82 days before parturition when the next offspring was born after 2 years [Tanaka, 19921. Suckling, hence, ceased a t approximately 110 days after estimated fertilization (see Table I). The data from different subjects and the rough estimates of fertilization age appear to overlap the duration of lactation and the termination of suckling behavior. Further study of hormones and continuous milking on the same subject is required. In rhesus macaques whose gestation length (168 days) is approximately equal to that of Japanese macaques (173 days), the growth of fetuses accelerates at 70 days of gestation [van Wagnen et al., 19651. From 90-100 days to 155-160 days of gestation, the rhesus fetal growth curve follows a n almost linear curve where fetal weight change is approximately 5.7 glday [van Wagenen et al., 19651. Assuming that prenatal growth in Japanese macaques follows a similar pattern, this suggests that lactation ceases when fetuses begin to grow linearly after organogenesis (embryonic period). Human mothers tend to abort embryos in organogenetic period [Hoshi, 19881. Boklage [19901 estimated that in human beings 73% of natural single conceptions do not survive to 6 weeks of gestation. In baboons, which are similar to humans in many important aspects of early reproductive biology, 60% of embryos were aborted spontaneously in early pregnancy [Kuehl et al., 19921. In Japanese macaques abortion peaks 1 month after conception [Goto, personal communication] although Japanese macaques are seasonal breeders, differing from baboons, which are non-seasonal breeders. Thus, after fetuses survive the crisis of miscarriage, macaque mothers appear to cease suckling yearlings. If mothers abort, mothers can increase survivability of the previous offspring in compensation for miscarriage with concurrent lactation. This suggests that Japanese macaques hedge maternal investment in yearlings and embryos with lactation in pregnancy. This coexistence of lactation and gestation is also seen in humans [the Kalahari desert San: Konner, 1977; rural Guatemalans: Klein et al., 19771. Jones [ 19881 reported rural Indonesian women who experienced resumption of menses Macaque Lactation in Gestation / 173 while breast-feeding. Ellison [ 19901 suggested that in human beings ovarian responsiveness has evolved to optimize the probability of a successful reproductive outcome in any individual instance of reproductive effort based on long-term energy balance. In rodents, simultaneous gestation and lactation of two litters may be a n adaptation to ecological constraints [Gilbert, 19843: rodents (voles and lemmings), whose breeding season is seasonally restricted and whose conception-toweaning interval is short relative to the breeding season, tend to overlap litters to increase fecundity with reduced interlitter spacing. In Japanese macaques, when the next offspring is born after a n interval of 1 year, suckling lasts until the subsequent parturition; when the next offspring is born after a n interval of 2 years, suckling continues into the early part of the next pregnancy; when the next offspring is born after a n interval of 3 years, suckling ceases before gestation [Tanaka, 19923. This concurs with milking data in multiparae (Table I). Therefore, the duration of lactation may be related to the ages of the previous offspring. In toque macaques, age-specific mortality peaked from birth to 1 year old, and secondly from 1year old to 2 years old [Dittus, 19791. This is consistent with Japanese macaques in Jigokudani Monkey Park [Tokida, personal communication]. When the interbirth interval to the subsequent offspring is 1year, mothers who desert suckling might tend to lose the previous infant, because infants less than a year old do not get milk like orphans. Mothers, hence, continue lactation until parturition. Because yearling orphans had better chances of surviving than newborn orphans in free-ranging Japanese macaques [Hasegawa & Hiraiwa, 19801, yearlings appear to need less maternal care. Therefore, mothers cease suckling 2 months before parturition. The rhesus prenatal growth curve [van Wagenen et al., 19651 shows a small cost of embryos on mothers in early pregnancy. This small expense may enable mothers to continue investment in the previous offspring and to increase its chance of survival. However, if the burden of fetuses on mothers increases after the growth of fetuses accelerates, mothers stop suckling yearlings. Since 2-year-old orphans are much more likely to survive than l-year-old orphans, lactation ceases before gestation when the next offspring is born after a n interval of more than 2 years. CONCLUSIONS 1. When the interbirth intervals to the subsequent offspring were 2 years, pregnant Japanese macaque multiparae (N = 14) secreted milk a t 76-97 days of estimated fertilization age. 2 . When the next offspring was born after 2 years, suckling lasted until approximately 100 days into gestation when the growth of fetuses accelerated. 3. Nulliparous pregnant females (N = 2) could secret milk 3 months before the first parturition. ACKNOWLEDGMENTS We thank Mr. Sougo Hara for permission to study a t Jigokudani Monkey Park; Mr. Eishi Tokida, Mr. Haruo Takefushi, Mr. Shigenori Nishizawa, Mr. Toshio Hagiwara, and the other staff of Jigokudani Monkey Park for their assistance; and Dr. Kenichi Aoki, Professor Toshisada Nishida, Dr. Yukio Takahata, and anonymous reviewers for their valuable advice. This study was supported in part by a grant in aid from the Japanese Ministry of Education, Science, and Culture. 174 / Tanaka et al. REFERENCES Boklage, C.E. Survival probability of human conceptions from fertilization to term. INTERNATIONAL JOURNAL OF FERTILITY 35:75-94, 1990. Dittus, W.P.J. The evolution of behaviors regulating density and age-specific sex ratios in a primate population. BEHAVIOUR 69:265-302, 1979. Ellison, P.T. Human ovarian function and reproductive ecology: New hypotheses. AMERICAN ANTHROPOLOGIST 92: 933-952, 1990. Gilbert, A.N. Postpartum and lactational estrus: A comparative analysis in rodentia. JOURNAL OF COMPARATIVE PSYCHOLOGY 98:232-245, 1984. Hasegawa, T.; Hiraiwa, M. Social interactions of orphans observed in a free-ranging troop of Japanese monkeys. FOLIA PRIMATOLOGICA 33:129-158, 1980. Hearn, J.P. Lactation and reproduction in non-human primates. SYMPOSIA OF THE ZOOLOGICAL SOCIETY OF LONDON 51:327-335, 1984. Hoshi, H. HUMAN GROWTH AND AGING. Tokyo, Therapeia, 1988 (in Japanese). Jones, R.E. A biobehavioral model for breastfeeding effects on return to menses postpartum in Javanese women. SOCIAL BIOLOGY 35~307-323,1988. Klein, R.E.; Lasky, R.E.; Yarbrough, C.; Habicht, J.; Sellers, M.J. Relationships of infantkaretaker interaction, social class and nutritional status to developmental test performance among Guatemalan infants. Pp. 385-403 in CULTURE AND INFANCY: VARIATION IN THE HUMAN EXPERIENCE. P.H. Leiderman, S.R. Tulkin, A. Rosenfeld, eds. New York, Academic Press, 1977. Konner, M. Infancy among the Kalahari desert San. Pp. 287-328 in CULTURE AND INFANCY: VARIATION IN THE HUMAN EXPERIENCE. P.H. Leiderman, S.R. Tulkin, A. Rosenfeld, eds. New York, Academic Press, 1977. Kuehl, T.J.; Kang, IS.; Siler-Khodr, T.M. Pregnancy and early reproductive failure in the baboon. AMERICAN JOURNAL OF PRIMATOLOGY 28:41-48, 1992. Martin, P.; Bateson, P. MEASURING BEHAVIOUR. Cambridge, Cambridge University Press, 1986. Nelson, W. APPLIED LIFE DATA ANALYSIS. New York, John Wiley & Sons, 1982. Nigi, H. Some aspects related to conception of the Japanese monkey (Macuca fuscata). PRIMATES 17:81-87, 1976. Tanaka, I. Three phases of lactation in freeranging Japanese macaques. ANIMAL BEHAVIOUR 44:129-139, 1992. van Wagenen, G.; Catchpole, H.R.; Negri, J.; Butzko, D. Growth of the fetus and placenta of the monkey (Macaca mulattu). AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 23:23-34, 1965.