Asynchrony within estrous synchrony among ringtailed lemurs (Primates: Lemuridae)

Physiology & Behavior, Vol. 49, pp. 47-52. ©Pergamon Press plc, 1991. Printed in the U.S.A.

0031-9384/91 $3.00 + .00

Asynchrony Within Estrous Synchrony Among Ringtailed Lemurs (Primates: Lemuridae) M I C H A E L E. P E R E I R A

Duke University Primate Center, 3705 Erwin Road, Durham, N C 27705 R e c e i v e d 19 April 1990

PEREIRA, M. E. Asynchrony within estrous synchrony among ringtailed lemurs (Primates: Lemuridae). PHYSIOL BEHAV 49(1) 47-52, 1991.--Data from wild and semifree-ranging groups of ringtailed lemurs (Lemur catta) were combined to evaluate the hypothesis that female ringtailed lemurs exhibit asynchrony of estrus within seasonal synchrony of estrous cycles. Photoperiodic and probably social entrainments lead all females within social groups of ringtailed lemurs to experience estrus annually within periods of 7 to 20 days. Among an average of five adult females in each of 11 group-years, however, only 2 separate instances of dyadic estrous overlap were observed. Computer simulations of estrus occurring independently among group-living females revealed that this was unlikely to have occurred by chance. Recent research on social enhancement and suppression of ovarian cycles in mammals suggests that a single pheromone-based signal-response system could mediate both ringtailed lemurs' remarkable annual estrous synchrony and the subsidiary asynchrony reported here. Asynchrony of estrus probably functions to maximize each female's ability to exercise mate choice by circumventing temporal conflict among females. Asynchrony of estrus and female mate choice cause current models to explain male membership in primate groups to fail for ringtailed lemurs. Such results highlight the need for detailed information on behavioral and physiological reproductive tactics before generally applicable models of reproductive strategies can be developed. Estrous cycle Mating systems

Primates

Synchrony

Asynchrony

Entrainment

THE Malagasy lemurs show the most seasonal breeding of all primates (30). In all populations yet studied, virtually all conceptions occur within an annual period of less than two months [e.g., (12, 26, 28, 29, 32, 35)]. Predictable seasonality of nutriment availability in Madagascar is widely believed to be responsible for this pattern: Across species of diverse sizes (<100 g to >5 kg), breeding is timed such that infants are weaned when food is readily available to them (14). The annual schedule of reproductive activity in ringtailed lemurs (Lemur catta) has been shown to be entrained to changes in photoperiod (8), with estrus occurring 4 to 5 months after the summer solstice (27). Within social groups, the reproductive activities of lemurs are typically extremely clustered in time. First, individual estrous periods usually last between several hours and just less than a day (4,10). Second, all females within groups usually pass through estrus within a period of one to three weeks (12, 13, 18, 32). Failing to conceive, female ringtailed lemurs experience a second estrous period about 39 days later (4, 7, 8, 24). Thus, mating within a group sometimes extends to 50 to 55 days: 7 to 20 days for the first round of estrus, plus the 39-day interval to one or more females' second estrous periods (32). Most conceptions, however, occur during females' first cycles (Fig. 1), making mating seasons as short as one week within social groups. In no anthropoid primate are mating opportunities so restricted for males. The mating seasons of the seasonally breeding macaques (Macaca spp.), for example, normally last three to four months (3,30), and individual females copulate across periods of several days or more (10).

Pheromones

Mate choice

Remarkably, all researchers of mating activity in wild ringtailed lemurs have noted that females were not observed in estrus on the same days (12, 18, 32). This also has been true generally for the semifree-ranging groups at the Duke University Primate Center [DUPC (24)]. Thus, it seemed possible that within the seasonal estrous synchrony of ringtailed lemurs, on a finer scale, lies deterministic asynchrony of estrus. To test this hypothesis quantitatively, I pooled all available information on estrus dates within free-ranging and semifree-ranging social groups of ringtailed lemurs. METHOD AND RESULTS

Observations The data came from three sources: 1) an ongoing program of research (16, 23-25, 36, 38) on two forest-living groups of ringtailed lemurs at the DUPC (Lcl and Lc2 Groups), 2) a recent 13-month study of two groups of individually identified ringtails living in the Beza-Mahafaly Special Reserve, Madagascar (32), and 3) previously published reports on the wild ringtailed lemurs at Berenty, Madagascar (12, 13, 18). Female ringtailed lemurs exhibit changes in vulval condition that allow human observers to predict the onset of sexual receptivity (13). During the final week preceding ovulation and receptivity, genital labia swell, becoming turgid, and the vaginal orifice opens and flushes pink or red in coloration. Also, female ringtails solicit copulation from particular males only during estrus (24,32), using their invariable agonistic dominance over males (12,25) to

47

48

PEREIRA

30, tO 25. r'-

Physiological Considerations and Levels of Synchron~ To determine whether female ringtailed lemurs exhibited nonrandom asynchrony of estrus, an estimate of the usual interval of days within which all females experience estrus was require& Two sets of observations indicated that 20 days would be a maximum value. First was the present dataset: in only one of 12 sets of first estrous cycles (Lc1-1986, Table 1) did more than 20 days elapse before all females experienced estrus. Second, the estrous cycles of ringtailed lemurs are known to be 39 days long. Thus, any estrus occurring more than 20 days after first estrus in a group reflects a cycle closer to and best considered pan of the group's second set of estrous cycles. This view accommodates existing knowledge on the reproductive physiology of lemurs and explains related behavioral observations. Van Horn and his colleagues (8,9) demonstrated that reductions in day length similar to those that occur at the summer solstice and autumnal equinox lead ringtailed lemurs to resume ovarian cycling [see also (27)]. Low-level female gonadal cycling probably begins anew annually just after the solstice. Each year at this time, formerly juvenile females begin genital marking and rates of gratuitous agonistic conflict between adult males increase sharply (unpublished data). Factors yet unknown gradually further synchronize females (see belowl and lead to fully functional cycling across the intervening 4 to 5 months. This would explain recurring observations of "pseudo-estrus" preceding true estrus in lemurs [e.g., (13, 28, 37)], as well as an unusual set of estrous events in Lcl Group. Just as pubertal (34) and early postlactational ovarian cycles are not fully functional in anthropoids (2), female lemurs probably often " w a s h b y " the first possible dates for estrus, while becoming progressively more attractive to males. Despite daily monitoring of Lcl Group in 1985. for example, estrus could be detected in only three of the seven mature females at the time when mating normally begins (Table 1). First estrus among the four remaining females occurred precisely 39 days at'-

Total births: 103

"~ 20. 015. L. Q~ "~10

E

z

5

J ~ ~ a M~ ~ . uAv JuN JuL ,uG sEP OCT Nov OEc T i m e of Y e a r ( w k s ) FIG. 1. Seasonality of breeding in the DUPC ringtailed lemurs. All term births from 1977 through 1990 plotted (ringtailed lemurs were permanently maintained outdoors at DUPC beginning in 1977). Note peaks of births occur roughly at 39-day intervals, the length of estrous cycles.

repel unpreferred males and all male attempts to copulate outside of estrus. In experimental settings, estrus in ringtailed lemurs can last 24 h (9), whereas it typically lasts only 4 to 12 h in natural settings (7,9). Nonetheless, females usually copulate with several males before termination of estrus (18,32). Vaginal sperm plugs, formed minutes after ejaculation [(7), see also (6)], are often retained for one or two days following estrus (personal observation). Observation of plugs in situ often helps to confirm a particular female's day of estrus. Observers in all studies from which data were taken for this report made use of all these behavioral, anatomical, and ejaculatory indices to determine estrus dates. Table 1 shows the complete set of observations used in the following analyses.

TABLE 1 FIRST ESTROUS DAYS IN 11 GROUP-YEARS (DAY FOLLOWING AUTUMNAL EQUINOX)

Female Study Group

1

2

3

4

5

Jolly 1966 Koyama 1988 Beza Black Beza Green Lc 1- 1985

27 30 49 49 43

32 32 50 54 46

33 33 56 55 60

34 35 57 59

39 36

Lc1-1986 Lc1-1987 Lc1-1988 Lc1-1989 Lc2-1988 Lc2-1989

40 30 28 37 38 37

41 41A 28 42 40 43

44 41P 34 43 41 45

82 45 48 35 44 47 52

83 64 49 39 45

6

7

68 86t

46

47

*Estrous date taken for only fully adult females not being targeted for social expulsion [see (38)]; A denotes e s t m s began in a.m.; P denotes estrus began in p.m. (see also text); both cases of estrous overlap (Lcl-1987,1988) overestimated to help ensure conservative statistical results, tCycles of three females appeared suppressed until time when the others' second cycles would have occurred, barfing conception; only one set of three females used for each analysis of asynchrony.

ESTROUS ASYNCHRONY AMONG LEMURS

WITHIN 20-DAY INTERVALS

6O0

49 A.

WITHIN 12-DAY INTERVALS

D.

500 400 300

20O

200

100

100 0

C WITHIN 12-DAY INTERVALS

C.

E 5ooi

7O0 600

~'O 400,

500

t-~ 300.

400 300

200 Z

B.

600

E 400 o 300

o

WITHIN 20-DAY INTERVALS

700

(n 5OO

o

800

200 1O0 0

1O0 0

5

10

2o 25 30 35 Number of half-days

15

0

5 10 15 20 of e s t r o u s o v e r l a p

0

25

3o

35

FIG. 2. Expected frequencies of estrous overlap assuming all females experience estrus independently with 20- or 12-day intervals (see text). Each analysis using data from 11 groupyears (panels A and C; females per group-year shown at right) repeated allowing only one set of values from each social group (panels B and D; females per group shown; see also Table 1 and text). Note that maximum possible frequencies of estrous overlap range between 102 (panel D) and 184 half-days (pane~ A).

ter their group's first estrus that year--the duration of one estrous cycle. Because many of the seasonally breeding anthropoids are photoperiodically entrained (19) and all have mating seasons much longer than those of lemurs, it seemed likely that a second mechanism acts to synchronize lemur estrous cycles further than the effect of photoperiod alone. I tested this by comparing the variability in timing of estrus within and between the 5 years of data available for Lc 1 Group (Table 1; note insufficient data yet available to compare timing of estrus in Lc 1 versus Lc2 Groups). In particular, because the nutrition available to this semifree-ranging group has remained constant over the years, interannual variability in the timing of estrus probably fairly represents the range of flexibility in females' responsiveness to changes in photoperiod. Between-year variability significantly exceeded within-year variability in timing of estrus (two-tailed Kruskal-Wallis test, H = 17, p<0.01), conrn'ming that a second mechanism further synchronized estrous cycles within years beyond the basic effect of photoperiod. The range of dates for first estruses in the outdoor history of the DUPC ringtailed lemurs, as reflected by the dates of earliest births (Fig. 1), suggests that declining photoperiods annually determine a season of about 5 weeks for the occurrence of first estruses (range of viable gestation lengths in past 5 years = 8 days).

Iterative Simulations To test for significance in frequency of asynchronous estrus, a computer program was developed to simulate the independent occurrence of estrus among female ringtailed lemurs within social groups. The program accepts values for the numbers of (a) social groups observed, (b) adult females per group, (c) days required for all females within groups to experience estrus, and (d) iterations desired. The program uses a random number generator [Turbo Pascal, version 4.0; see (5)] to select for every female an onset time for estrus within the specified interval of days. Estrus periods are assigned 24-h durations. Iteration of the model generates for a specified population the expected distribution of fre-

quencies of estrous synchrony, thereby allowing estimation of the exact probability of having obtained any given result by chance alone. Lemur estrus can begin at any time of day. When necessary (see Table 1), females first seen in estrus just after dawn were considered to have entered estrus between midnight and midday; those whose estrus clearly began after the onset of morning observations were considered to have initiated estrus between midday and midnight. The simulation program was adjusted to select at random for each female a half-day for the onset of estrus within the specified interval of days inclusive of all estrous periods. Simulated females whose estrous periods began during the same half-day experienced full estrous overlap (two half-days), whereas those whose estrous periods began during adjacent half-days experienced partial overlap (one half-day). Several factors ensured that the exact probability testing conducted was conservative (alpha level = 0.05). First, the null model assumes that females experience estrus independent of one another within specified intervals of days. The previous analysis of variance indicated that, for long inclusive intervals (e.g., maximum value, 20 days), the most likely violation of this assumption would have been a synchronizing effect that moved estrous cycles closer together. I tested the data against the null model, however, for significant asynchrony of estrus. Second, whereas the model's simulation of estrous overlap is coarsely discrete, estrous periods are likely free to overlap in a more continuous manner, and refinement of the program's algorithm toward continuity increases the significance of any given result. For example, the probability of 1 day of estrous overlap using a model that chooses a full, 24-h estrous period for each female exceeded that of two half-days of overlap using the present model, which, in turn, exceeded that of four quarter-days of overlap using an appropriately modified third model. Limited information on onsets and terminations of estrus precluded use of more refined models. Finally, the degree of both observed instances of estrous overlap was purposely overestimated (Table 1).

PEREIRA

50

The dataset was analyzed twice using the maximum possible interval needed for all females to experience estrus (20 days). First, separate years of data from DUPC groups were considered independent (Table 1, 11 group-years). The second time, each group contributed only one set of values to the analysis (6 social groups; mean values for DUPC groups). In each case, the amount of estrus simultaneity observed (3 half-days) was significantly less than could be expected to have occurred by chance alone (Fig. 2A and B). Usually, all females experienced estrus within 12 or fewer days; in each remaining case, all but one female did so (Table 1). Therefore, the hypothetical interval within which all estrus occurs was reduced to 12 days and the number of females in each groupyear that experienced estrus within that interval was represented for a second simulation (Fig. 2C). It proved exceedingly unlikely to have seen only 3 half-days of estrous overlap by chance among females cycling so closely in time, and the result retained significance when only a single set of values from each social group was accepted (Fig. 2D). DISCUSSION Within remarkable annual estrous synchrony, female ringtailed lemurs exhibit asynchronous estrus within social groups. Whereas estrous synchrony has been described for many species, including humans (17,20), this is the first demonstration of estrous synchrony in lemurs beyond the effect of photoperiod and the first report of deterministic asynchrony of estrus among group-living mammals. Potential Mechanisms McClintock (20,21) recently developed a model, based on empirical work with rats, that potentially unifies the phenomena of suppression, enhancement, and synchrony of ovarian cycling in mammals. By presenting cycling female rats with controlled airstreams from other cycling females, she demonstrated that odors from females in the follicular phase of estrous cycles phase advance (shorten) the cycles of recipient females, whereas odors from ovulating females phase delay (lengthen) cycles. These resuits were predicted by McClintock's model of synchrony among ovarian cycles as the progressive coupling (mutual entrainment) of a set of oscillators (40). Pheromonal signalling among female lemurs is likely responsible for both the tightness of estrous synchrony and the ultimate asynchrony of estrus shown by ringtailed lemurs. Lemurs scent mark using diverse glands and urine (16), and as mating seasons approach, female ringtaited lemurs appear to increase the rates at which they olfactorily investigate each other's genitals and scent marks (personal observation). One hundred thirty-two days, or time enough for just over three estrous cycles, occur between the summer solstice and the annual time of first estrus in ringtailed lemurs. This is the amount of time apparently needed for social coupling of estrous cycles (20). Ultimately, one of each pair of females whose cycles become most closely synchronized in phase must respond to their partner's signals of impending ovulation by delaying their own ovulation. This conforms to the common observation of pairs and trios of ringtailed females experiencing estrus on consecutive days (12, 13, 18, 24, 32) (see esp. Lc1-1989, Table 1). Estrous Asynchrony, Female Mate Choice, and Male Group Membership Whatever the function of annual estrous synchrony among ringtalled lemurs [see (17) for general review of hypotheses], the

final asynchrony of estrus most likely functions to n~axim(ze ttlc efficiency with which individual females can choose their mates. From separate research projects, Sauther 132) and Pereira and Weiss (24) concluded independently that female ringtailed lemurs exercise considerable, if not complete, freedom in choosing their mates. At Beza-Mahafaly, Madagascar. estrous females commonly repel particular males from attempting to copulate, while showing receptivity and sometimes proceptivity toward other males (32). These same patterns are seen each year among the forestliving ringtailed lemurs at Duke, where complementary DNA fingerprinting analyses have provided additional information about female mating preferences. Across a five-year study period, no female in Lc I Group ever allowed a son or matrilineal brother to mate with her. whereas such close kin formed the predominant class of available males and they commonly made copulatory advances 124). In the last two years, several estrous females were attracted exclusively to an unrelated pair of males that had recently immigrated. These males were very low-ranking among males and were attacked by resident males whenever estrous females approached them (the females never defended them). Nonetheless, DNA analysis revealed that every female exclusively attracted to the immigrants reproduced with the dominant one of the pair. Estrous asynchrony heightens each female's ability to exercise mate choice by circumventing temporal conflict with other estrous females. It also increases the chance of one-male reproduction within groups of ringtailed lemurs, particularly if certain variable male traits are universally attractive to females. Elsewhere (24), we have argued that, all else equal, ringtailed females should prefer to mate with unrelated males that have displayed long-term dominance over other adult males (note: in Pereira and Weiss's study, only the newly arrived, low-ranking males were unrelated to most females). Such fathers could best help deter potentially infanticidal males [see (24)]. The otherwise tremendous synchrony of estrus within groups and females" predisposition to mate with many males are two other reproductive tactics that shield individual infants from the risk of male infanticide (24). The present results underscore the need for detailed information on basic behavioral and physiological reproductive tactics before generally applicable models of reproductive strategies can be developed [see also (17,22)]. The reproductive biology of ringtailed lemurs contradicts two one-factor models that seek to explain male membership in primate groups. Ridley (30) suggested that primates become more likely to form multimale groups as mating seasons grow shorter because this increases the incidence of simultaneous estrus and thereby reduces any one male's ability to monopolize mating. Existing lemur data were taken in support of this argument because lemur mating seasons are the shortest among primates and the diurnal, gregarious lemurs form multimale groups. While ringtailed lemur groups include more males than do those of any other lemur, it is now clear that asynchrony of brief estrus periods makes possible one-male reproduction within groups, and female choice patterns may often ensure it (24). A second model, most recently discussed by Andelman (3) and Altmann (1), posits that when female primates form small groups (<6), a single reproductive male is most often present because he can monopolize mating with so few females. Groups of ringtailed lemurs typically have six or fewer adult females, but also contain many adult males. Why species like Lemur catta and Cebus monkeys form multimale social groups when one male often dominates reproduction [e.g., (11,31)] is an important question in the study of primate group demography. If an environmental factor, like predation pressure, favors group living for males, why do males not form all-male bands and forcefully usurp reproductive males in bisex-

ESTROUS ASYNCHRONY A M O N G LEMURS

51

ual groups, as do male langur monkeys in some populations? One speculation for Cebus monkeys is that females collaborate to prevent any extragroup male from taking over their group, requiring an initial period of " s e r v i c e " from males as food finders and predator spotters before supporting their transition to reproductive status (33). In ringtailed lemurs, females' absolute freedom to choose mates and their predicted attractedness to consistently dominant males may require males to display long-term agonistic prowess before they can become attractive as mates. This could explain why, among seasonally breeding primates, only in ringtailed lemurs do males transfer between groups during lactation seasons, often many months before the next mating season (15,32). Each of these hypotheses posits a partial queuing by males for reproductive status. Partial queuing differs from strict queuing [e.g., male stripe-backed wrens (39)] in that the particular male that next assumes reproductive status at any given time may not be the male that has been waiting longest. Partial queuing may be robust against the evolution of alternative tactics (e.g., all-male

bands, mating season migration) because males " w a i t i n g " to attain primary reproductive status (a) achieve a small amount of reproduction along the way, and (b) collaborate with females to preclude success for male " c h e a t e r s . "

ACKNOWLEDGEMENTS The staff of the DUPC is appreciated for its caring and expert maintenance of a unique colony of prosimian primates. I thank P. Kappeler for providing the 1987 dates of estrus in Lcl Group and M. Sauther for the estrus dates in her two study groups in Madagascar. D. Burdick encouraged me to develop the "bootstrapping" technique of analysis used in the simulation. P. Kappeler and C. van Schaik contributed stimulating discussions of behavioral phenomena associated with mating in primates. M. McClintock and J. Vandenbergh criticized a previous draft and provided critical references. T.-J. Pyer generously facilitated all aspects of the work, which was financially supported by NIH (R29-HD23243). Duke Primate Center publication #494.

REFERENCES 1. Altmann, J. Primate males go where the females are. Anim. Behav. 39:193-195; 1990. 2. Altmann, J.; Altmann, S. A.; Hausfater, G. Primate infant's effects on mother's future reproduction. Science 201:1028-1030; 1978. 3. Andelman, S. J. Ecological and social determinants of Cercopithecine mating patterns. In: Rubenstein, D. I.; Wrangham, R. W., eds. Ecological aspects of social evolution. Birds and mammals. Princeton, NJ; Princeton University Press; 1986:201-216. 4. Bogart, M. H.; Kumamoto, A. T.; Lasley, B. L. A comparison of the reproductive cycle of three species of Lemur. Folia Primatol. (Basel) 28:134-143; 1977. 5. Borland International Turbo Pascal, 4.0. Scotts Valley, CA: Borland International; 1987. 6. Cranz, C.; Ishak, B.; Brun, B.; Rumpler, Y. Study of morphological and cytological parameters indicating oestrus in Lemur fulvus mayottensis. Zool. Biol. 5:379-386; 1986. 7. Evans, C. S.; Goy, R. W. Social behaviour and reproductive cycles in captive ringtailed lemurs (Lemur catta). J. Zool. (Lond.) 156:181197; 1968. 8. Horn, R. N. van Primate breeding season: photoperiod regulation in captive Lemur catta. Folia Primatol. (Basel) 24:203-220; 1975. 9. Horn, R. N. van; Resko, J. A. The reproductive cycle of the ringtailed lemur (Lemur catta): sex steroid levels and sexual receptivity under controlled photoperiods. Endocrinology 101:1579-1586; 1977. 10. Hrdy, S. B.; Whitten, P. L. Patterning of sexual activity. In: Smuts, B. B.; Cheney, D. L.; Seyfarth, R. M.; Wrangham, R. W.; Struhsaker, T. T. Primate societies. Chicago: University of Chicago Press; 1987:370-384. 11. Janson, C. H. Female choice and mating system of the brown capuchin monkey Cebus apella (Primates: Cebidae). Zeit. Tierpsychol. 65:177-200; 1984. 12. Jolly, A. Lemur behavior: A Madagascar field study. Chicago: University of Chicago Press; 1966. 13. Jolly, A. Breeding synchrony in wild Lemur catta. In: Altmann, S. A., ed. Social communication among primates. Chicago: University of Chicago Press; 1967:3-13. 14. Jolly, A. The puzzle of female feeding priority. In: Small, M., ed. Female primates: Studies by women primatologists. New York: Alan R. Liss, Inc.; 1984:197-215. 15. Jones, K. C. Inter-troop transfer of Lemur catta males at Berenty, Madagascar. Folia Primatol. (Basel)40:145-160; 1983. 16. Kappeler, P. M. The effect of social status on scent marking behavior in ringtailed lemurs (Lemur catta). Anim. Behav.; in press. 17. Knowlton, N. Reproductive synchrony, parental investment, and the evolutionary dynamics of sexual selection. Anim. Behav. 27:10221033; 1979. 18. Koyama, N. Mating behavior of ring-tailed lemurs (Lemur catta) at Berenty, Madagascar. Primates 29:163-175; 1988. 19. Lindberg, D. G. Seasonality of reproduction in primates. In: Mitch-

20.

21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36.

ell, G.; Erwin, J., eds. Comparative primate biology, vol. 2, part B: Behavior, cognition, and motivation. New York: Alan R. Liss, Inc.; 1987:167-218. McClintock, M. K. Pheromonal regulation of the ovarian cycle: enhancement, suppression, and synchrony. In: Vandenbergh, J. G., ed. Pheromones and reproduction in mammals. New York: Academic Press; 1983:113-149. McClintock, M. K. Estrous synchrony: Modulation of ovarian cycle length by female pheromones. Physiol. Behav. 32:701-705; 1984. Packer, C.; Pusey, A. E. Male takeovers and female reproductive parameters: a simulation of oestrous synchrony in lions (Panthera leo). Anim. Behav. 31:334-340; 1983. Pereira, M. E.; Izard, M. K. Lactation and care for unrelated infants in forest-living ringtailed lemurs. Am. J. Primatol. 18:101-108; 1989. Pereira, M. E.; Weiss, M. L. Female mate choice, male migration, and the threat of infanticide in ringtailed lemurs. Behav. Ecol. Sociobiol., in press; 1991. Pereira, M. E.; Kaufman, R.; Kappeler, P. M.; Overdorff, D. Female dominance does not characterize all of the Lemuridae. Folia Primatol. (Basel) 55:96-103; 1990. Pener, J.-J.; Albignac, R.; Rumpler, Y. Faune de Madagascar. Mammiferes lemuriens, vol. 44. Paris: Orstom, CNRS; 1977. Rasmussen, D. T. A comparative study of breeding seasonality and litter size in eleven taxa of captive lemurs (Lemur and Varecia). Int. J. Primatol. 6:501-517; 1985. Richard, A. Patterns of mating in Propithecus verreauxi verreauxi. In: Martin, R. D.; Doyle, G. A.; Walker, A. C. Prosimian biology. London: Duckworth; 1974:49-75. Richard, A. F.; Rakotomanga, P.; Schwartz, M. Demography of Propithecus verreauxi at Beza Mahafaly: Sex ratio, survival and fertility, 1984-1988. Am. J. Phys. Anthropol., in press; 1990. Ridley, M. The number of males in a primate troop. Anim. Behav. 34:1848-1858; 1986. Robinson, J. G. Demography and group structure in wedge-capped capuchin monkeys, Cebus olivaceus. Behaviour 104:202-232; 1988. Sauther, M. L. Reproductive behavior of free-ranging Lemur catta at Beza Mahafaly Special Reserve, Madagascar. Am. J. Phys. Anthropol., in press; 1991. Schaik, C. P. van; Noordwijk, M. van. The special role of male Cebus monkeys in predation avoidance and its effect on group composition. Behav. Ecol. Sociobiol. 24:265-276; 1989. Scott, L. M. Reproductive behavior of adolescent female baboons (Papio anubis) in Kenya. In: Small, M., ed. Female primates: Studies by women primatologists. New York: Alan R. Liss, Inc.; 1984. Sussman, R. W. A preliminary study of the behavior and ecology of Lemurfulvus rufus, Audebert 1800. In: Tattersall, I.; Sussman, R. W. Lemur biology. New York: Plenum Press; 1975:237-258. Taylor, L. L.; Sussman, R. W. A preliminary study of kinship and social organization in a semi-free-ranging group of Lemur catta. Int.

52

J. Primatol. 6:601-614; 1985. 37. Vick, L. G. The role of interindividual relationships in two troops of captive Lemurfulvus. Ph.D. dissertation, University of North Carolina at Chapel Hill; 1977. 38. Vick, L. G.; Pereira, M. E. Episodic targeting aggression and the histories of Lemur social groups. Behav. Ecol. Sociobiol. 25:3-12:

PEREIRA

1989. 39. Wiley, R. H.; Rabenold; K. N. The evolution of cooperative breeding by delayed reciprocity and queuing for favorable social positions. Evolution 38:609-621; 1984. 40. Winfree, A. T. The geometry of biological time. New York: Springer-Verlag, 1980.