International Journal of Primatology

, Volume 28, Issue 1, pp 73–96

Female Defensibility in a Small Troops of Japanese Macaques vis-à-vis Nontroop Males and Copulation on the Periphery of the Troop

Article

DOI: 10.1007/s10764-006-9109-1

Cite this article as:
Hayakawa, S. Int J Primatol (2007) 28: 73. doi:10.1007/s10764-006-9109-1

I provide data compiled over 4 yr on the mating behavior in small troops of wild Japanese macaques on Yakushima Island. The key parameters are the number of sexually receptive females, the number of nontroop males (NTMs), and copulation on the periphery of the troop. I analyzed the following aspects: 1) changes in the proportion of copulation with high-ranking males (HRMs) and NTMs, 2) variations in factors such as fluctuation in the number of sexually receptive females and troop males and their effects on the number of visiting NTMs, 3) the effect of attempted interruption of mounting series by other males, and 4) some aspects of copulation on the periphery of the troop. Throughout the study, 56% of the total number of females mated most frequently with the α-male in a single mating season. However, the relative mating success of HRMs varied over the years and between individuals. The number of visiting NTMs varied depending on the number of receptive females and troop males. Females tended to mate with the NTMs when they appeared around their troops. The direct effect of interruption of the mounting series by other males is equivocal. The females mated with the low-ranking males (LRMs) and NTMs on the periphery of the troop, which increased the possibility of mounting series ending with ejaculation. Females actively sought opportunities for copulation on the periphery of the troop by moving there or initiating close proximity with LRMs and NTMs there. On Yakushima Island, the mating success of HRMs was not always as high as that predicted by the priority of access model. The injury status of the HRM, the number of visiting NTMs, and female choice are all considered to influence a male’s mating success.

KEYWORDS

copulation on the periphery of the troop female choice Japanese macaque nontroop males small troops 

INTRODUCTION

Macaques live in a multimale multifemale social system, and the male dominance rank is often associated with his mating success. The priority of access to estrous female model that Altmann (1962) proposed predicts that the high-ranking males (HRMs) will have priority over the lower-ranking males in access to sexually receptive females. However, researchers have frequently reported exceptions to the prediction (Bercovitch, 1991; Cowlishaw and Dunbar, 1991; Fedigan, 1983; Shively and Smith, 1985; Smuts, 1987; cf. Takahashi, 2004). Ellis (1995) reviewed the literature on dominance and reproductive success, and reported that among 75 studies, 51 concluded that dominant males copulated more frequently with females than other males did. The effect of male rank on mating success may be a conditional probability, and some conditions, i.e., sex ratio, number of females, may vary widely even in a species so that the correlation may not be relevant.

HRMs in Japanese macaques do not attain high mating success consistently in provisioned troops of large size (Enomoto, 1974; Fedigan et al., 1986; Takahata et al., 1999). Female partner selection clearly plays a decisive role in the mating system; female preference for less familiar or novel partners results in higher mating success for younger, middle-ranking males (Huffman, 1991, 1992; Takahata et al., 1999). However, some reports in the habituated wild population show monopolization of copulation by dominant males, especially when the number of estrous females is small (Matsubara, 2004; Soltis, 1999; Takahashi, 2004).

The population of wild Japanese macaques of Yakushima Island is characterized by small-sized troops and high mobility of males during the mating season (Maruhashi, 1982; Sprague, 1991). Thus, many males visit adjacent troops as NTMs during the mating season. Some researchers have reported that NTMs contribute up to 30–40% of a troop’s total copulations during the mating season (Yamagiwa, 1985; Sprague, 1991, 1998). Some NTMs behave aggressively to troop males and acquire the α rank immediately. Therefore, some top-ranking males are also new males, though less commonly than in large provisioned troops on Honsyu, the main island of Japan (Sprague, 1992; Yamagiwa, 1985). The number of NTMs sometimes correlates with the number of estrous females and troop males, so that some troops have no, or only a small number of, NTM visitors (Okayasu, 2001; Soltis et al.,2001).

In general, as the number of estrous females decreases, a few HRMs can easily monopolize them (Cowlishaw and Dunbar, 1991; Smith, 1994; van Schaik, 1983). HRMs monopolized the estrous females in Yakushima, particularly when only 1 or 2 were present (Matsubara, 2004; Soltis, 1999; Soltis et al., 2001). The study of mating activity in small troops on Yakushima Island is important because intertroop mobility of males during the mating season is affected by troop size. Young males preferentially enter troops with many males, while fully grown adult males often enter small troops with fewer resident males as α-males (Suzuki et al., 1998). Thus, males in their prime—optimal age and body size—stay in small troops during their important reproductive stages. Can the males expect high mating success every year? And how does the presence of NTMs affect their mating success?

Paternity becomes less certain when females mate with multiple males, which may result in the reduced risk of infanticide (Hrdy, 1977; Smuts, 1987; Soltis et al., 2001). Conflicts occur when females try to mate with multiple males and HRMs try to monopolize them (Bercovitch, 1995; Soltis, 1999). Such conflict should be strongest in small troops on Yakushima where females can be easily monopolized, while conversely the frequency of troop takeover is likely to be high (Yamagiwa, 1985; Soltis et al., 2001).

I provide an analysis of the data, compiled over 4 yr, on the mating behavior of wild Yakushima macaques in small troops. The focus is on the variation in the frequency of copulation between females and HRMs with reference to factors such as the number of receptive females, the number of NTMs, HRM’s interruption of other male’s mounting series, and the female’s behavior of moving to the periphery of the troop. The differences are expected to play a key role in the variation of mating success across troops over the years in the Yakushima macaques.

MATERIALS AND METHODS

The Study Population

The study site is located on the western coast of Yakushima Island (30°N, 130°E), Japan. I conducted the study on troop B and troop NA in a coastal broadleaf evergreen forest, which lies 0–350 m above sea level (Maruhashi, 1980, 1981). The population is characterized by small troops with overlapping home ranges. The members of troops NA and B tolerated human presence, and on many occasions allowed an observer to approach ≤2 m. I identified all members >4 yr that copulated. The years of study and the corresponding troop compositions are in Table I. The mating season begins in September and lasts until February (Maruhashi, 1982; Soltis et al., 2001). The populations of both study troops decreased between August 1998 and August 1999 (troop B: 21 to 10; troop NA: 47 to 22 individuals) because of localized mass mortality, attributed to poor fruit production and bad nutritional conditions (Hanya et al., 2004).
Table I.

Distribution of adult males in study troops B and NA

  

Troop male

 

Study troop

Year

HRM

LRM

Natal male

Total

Total number of NTM

B

1996

2

3

1

6

13

 

1997

2

4

1

7

5

NA

1999

2

4

1

7

11

 

2000

1

5

4

10

2

Note. HRM, high-ranking male; LRM, low-ranking male; NTM, nontroop male.

The 4-yr study period is outlined as follows: October 10–December 1, 1996; October 1–December 1, 1997; September 29–January 13, 1999–2000; September 27–November 29, 2000.

Sampling of Behavioral Data

I checked the receptivity of all females daily. A female was receptive if the male partner copulated to ejaculation or she had a coagulated seminal plug. I noted copulation frequency and the male partner of each female. I defined mounting series as a male mounting a female with intromission, terminated either by ejaculation or a period of 5 min without mounting. If terminated by ejaculation, I recorded it as a single copulation. I considered ejaculation to have occurred when either a male assumed the ejaculatory posture or I observed a copulatory plug immediately after the male dismounted the female. Other males sometimes interrupted the mounting series, usually by chasing the female or the male partner, or sometimes by just approaching or staring at them from a distance.

The sampling procedure varied over the 4 yr of observation. I carried out observations via ad libitum sampling (Altmann, 1974) in 1996. In 1997, I conducted focal individual sampling on 5 females. I defined a focal session as the gathering of continuous focal data ≤5 min of disappearance of the female. If I did not locate her ≤5 min, I selected and observed another female. Receptive females often move toward the periphery of the troop and return to the center. It is easier to locate them at the center than at the periphery of the troop, and a female’s location is considered to be influenced by that of her copulating partner. In 1999 and 2000, I followed females for 4 hr, each on alternate days. I excluded focal sample sessions of <150 min from the analysis to avoid sample bias. However, if the duration of a focal session was not <150 min and I relocated the missing female, I followed her again either on the same day or the next. When I did not locate the missing female in the troop for the rest of that day and also the next, I recorded her as absent.

I categorized the copulating partners according to the following definitions. Troop males remained in the troop since the previous nonmating season, and classified as HRMs, LRMs, and natal males. HRMs include 1 top-ranking male (α-male) of troop NA (2000) and 2 top-ranking males (α- and β-male) of each troop for the rest of the study period (1996, 1997, 1999). I defined their ranks on the basis of the observed approach-retreat interaction among the troop males. The number of HRMs is determined annually by the number of receptive females in each troop; the priority-of-access model predicts that lower-ranking males should have access to fertile females only when >1 female is in estrus at the same time, e.g., the third-ranking male should consort only if the fertile period of 3 females overlaps. In troop NA (2000), however, only 1 female was receptive throughout more than two-thirds of the period of observation, so I classified only 1 male as HRM.

Natal males are born in the troop under observation. During the study period, no natal male achieved a high-ranking status in the troop.

LRMs are males other than HRMs and natal males, and include the supposed nonnatal immigrant males that do not have a specific maternal kin that they stay with during resting and grooming times.

I modified Takahashi’s (2001) and Horiuchi’s (2005) distinction between troop males and NTMs as follows. NTMs are males that first interacted with the study troop after the onset of a mating season. NTMs that appeared near the study troop during a mating season included solitary males and males belonging to other troops. In the Yakushima population, troop males visited other troops as NTMs and mated during the study period (Sprague, 1992). NTMs include both the visitors that did not remain with the troop after the mating season ended and the new males that remained with the troop. In Yakushima, some NTMs dominate troop males and show a suite of behaviors that researchers call α-male attitude, including tree shaking with barking vocalizations and aggressiveness toward troop males (Sprague, 1992). I also categorized such males as NTMs because 1) they often desert the troop by the end of the mating season because of a takeover by other NTMs with α-male attitude or for unknown reasons (Sprague, 1991; personal observation). Therefore, it is unknown during the mating season whether the new males will remain in the troop after the mating season. 2) Troop males affiliate with each other, especially in Yakushima, and the affiliation is often established with new males after the mating season is over (Horiuchi, 2005). 3) The males that joined the troop after the mating season and interacted only with receptive females and rarely with nonestrous females during the mating season (Tsukahara, 1990).

Number of NTMs, Receptive Females, and Troop Males

I observed the number of NTMs, troop males, and receptive females daily. I recorded a member as present in the troop when he or she was within sight of multiple troop members. Sometimes I observed a receptive female mating, grooming, or sitting together with one of the troop males or NTMs, with no sign of other troop members around her. In such a case, I recorded the pair as absent from the troop. When I did not locate the scheduled receptive female in the troop for 2 h after locating the troop and accounting for all other nonreceptive troop females, I recorded her as absent from the troop and changed the focal individual, or I searched for her throughout the study area if no alternative female was present in the troop. Okayasu (2001) suggested that a period <9 d with no sexual behavior could be regarded as a temporary cessation of sexual activities within 1 estrous spell. I considered females that remained receptive after an absence of <9 d, the females as receptive throughout the period. The troop’s male membership was affected by male immigration and dispersion as well as by aging of the juvenile class into adulthood. Further, some troop males were absent from the troop for 1 to several days, as were females, presumably visiting other troops and exploring mating opportunities.

Copulation on the Periphery of the Troop

In 1999 and 2000, I collected data regarding the visibility of other troop members for troop NA. I recorded data regarding the first and second nearest male neighbors and the nearest female neighbor to the focal female every 5 min of scanning. I considered mounting series or copulations, for which there was no record of neighbors in sight (males and females) other than partners, as performed on the periphery of the troop. Conversely, when records existed for male neighbors in sight, I recorded it as with male neighbors nearby. I also classified the records as with higher-ranking males nearby when the neighbor was higher in rank than the copulating partner.

Some copulations occurring on the periphery of the troop involved the female moving toward the periphery, defined as walking silently for >30 s, apparently following or being followed by a male, to an area that is out of sight of other troop members. Some copulations occurring on the periphery of the troop did not involve the behavior, partly because the close proximity relationship had been started behind other troop members. I defined a close proximity relationship as one beginning with an approach ≤1 m of each other, terminating ≤5 min without approaching ≤1 m or after a close proximity relationship with another male was initiated. I also recorded the initiation of close proximity relationships by females or males as either with neighbors (males and females) nearby or on the periphery of the troop.

Statistical Analysis

In the statistical analysis, I used nonparametric tests, and rejected the null hypotheses at p=.05. The statistical tests are 2-tailed and were conducted via STATISTICA for the Mann-Whitney test, Wilcoxon signed rank test, χ2 test, and Spearman’s rank test. I calculated only the Steel-Dwass test by hand.

RESULTS

Interannual Changes in the Copulation Success of HRMs

The distribution of copulation by females according to the status of their male partner across multiple years and troops is in Table II, with 16 female rows including 12 females and 4 females that were receptive in multiple years during study periods. Females copulated with multiple males during a single mating season (mean=6, SD=2.22, range 3–11) and 56.0% (9/16) of the data show that they mated most frequently with the α-male throughout the single mating season.
Table II.

Females’ copulation partners and copulation frequency

   

Total

HRM

LRM

Natal male

NTM

Total

Status of most mated male

Troop: year

Female

Hr.

ejaculations

%

%

%

%

%

 

%

B: 1996

Bb

*

44

4.9 (2)

4.6 (3)

.0

36.4 (3)

100 (8)

Α-Male

31.8

 

Sr

*

18

16.7 (2)

5.6 (1)

.0

77.8 (5)

100 (8)

NTM

57.9

 

St

*

15

2.0 (1)

6.7 (1)

.0

73.3 (3)

100 (5)

NTM

57.1

 

Tr

*

29

44.8 (2)

1.3 (1)

3.4 (1)

41.4 (4)

100 (8)

α-Male

31.0

 

Average

  

3.6 (1.8)

6.8 (1.5)

.85 (.3)

57.2 (3.8)

100 (7.3)

  

B: 1997

Bb

34.08

25

8.0 (1)

12.0 (1)

.0

8.0 (1)

100 (3)

α-Male

8.0

 

Ks

23:12

17

41.2 (2)

47.1 (4)

5.9 (1)

5.9 (1)

100 (8)

α-Male

35.3

 

Sr

43:34

23

95.7 (2)

.0

4.3 (1)

.0

100 (3)

α-Male

65.2

 

St

33:50

15

10.0 (1)

.0

.0

.0

100 (1)

α-Male

10.0

 

Th

35:28

14

92.9 (2)

7.1 (1)

.0

.0

100 (3)

α-Male

5.0

 

Average

  

82.0 (1.6)

13.2 (1.2)

2.0 (.5)

2.8 (.4)

100 (3.6)

  

NA: 1999

Ht

28:00

17

41.2 (1)

29.4 (2)

17.6 (2)

11.8 (1)

100 (6)

α-Male

41.2

 

Rm

71:30

23

21.7 (2)

21.7 (3)

21.7 (1)

34.8 (2)

100 (8)

NTM

26.1

 

Sj

32:00

18

11.1 (2)

66.7 (3)

5.6 (1)

16.7 (1)

100 (7)

LRM

27.8

 

Tm

52:00

27

7.4 (2)

4.7 (3)

25.9 (2)

25.9 (3)

100 (10)

LRM

33.3

 

Average

  

2.4 (1.8)

39.6 (2.8)

17.7 (1.5)

22.3 (1.8)

100 (7.8)

  

NA: 2000

Hn

46:30

12

33.3 (1)

.0

66.7 (2)

.0

100 (3)

Natal male

5.0

 

Ht

35:06

23

8.7 (1)

34.8 (3)

56.5 (2)

.0

100 (6)

Natal male

21.7

 

Sk

40:00

21

66.7 (2)

14.3 (2)

19.0 (1)

.0

100 (5)

α -male

66.7

 

Average

  

36.2 (1.3)

16.4 (1.7)

47.4 (1.7)

  0 (0)

100 (4.7)

  

aIn 1996, I recorded copulations via ad libitum sampling. However, it is evident that every female mated with multiple NTMs and females were not monopolized by HRMs. The numbers of male partners during the observations are indicated in parentheses. Hr, focal observation time for each female; HRM, high-ranking male; LRM, low-ranking male; NTM, nontroop male.

The relative mating success of HRMs varied over the years and across troops (Table II). In 1996, I selected B as the study group. The α-male was BB. In mid-October, PE, a NTM with α-male attitude, visited B and dominated over all the troop males, including the α-male. His dominance was apparent from approach-retrieve interaction involving him and troop males and the grimaces of other troop males toward him. In the beginning of November, another NTM, MR, appeared in the vicinity of troop B. He was aggressive toward the troop males and PE, and soon all the males showed submissive behavior toward MR. After his appearance, the frequency of visits by PE decreased, and after late November, he stopped visiting troop B. A few cases of cooperative aggression by the troop males toward MR also occurred, though they could not expel him completely. A total of 12 NTMs, including PE and MR, visited troop B in 1996. However, only MR remained in B after the end of the mating season. In 1997, I again selected B as the study troop. MR, who entered the troop in the previous year, was the α-male of the troop. NTMs rarely appeared in the vicinity of the troop, and the α-male monopolized all the females. In 1999, the study troop was NA. The α-male of the troop was ID and had lived in NA for ≥2 yr. In early October, a NTM, BT, visited NA and dominated the α-male and mated with the receptive females. However, he disappeared around late October when no females showed receptive behavior for >10 d. During the period, ID sustained severe injuries, inflicted by an unknown aggressor, on his leg and limped throughout the mating season. In 1999, a total of 11 NTMs visited NA and copulated with the females. The females also left the troop for several days at a stretch, and their whereabouts were usually unknown. In 2000, the study troop was NA, and the α-male was ID. However, in mid-November, the β-male SH challenged ID, and a rank reversal occurred.

I could not compare copulation frequency for 1996 because I recorded copulations via ad libitum sampling. However, the females mated with multiple males, including some NTMs (Table II), and the α-male was unable to monopolize them effectively.

The females copulated more frequently with the α-males than with other males only in 1997 (B, 1997: N=30, Mann-Whitney z=3.136, p<.01; NA, 1999: N=27, z=−.694, p=.9726; NA, in 2000: N=23, z=−1.883, p=.0596). However, the copulation success of HRMs may be influenced by the natal males because natal males probably do not mate much, not because they are low-ranking but because the females are their relatives. Therefore, as a next step, I conducted the same analysis excluding the natal males. However, the analysis too revealed that the females copulated more frequently with the α-male in 1997 and 2000 (B, 1997: N=25, Mann-Whitney z=−2.923, p<.01; NA, 1999: N=23, z=−.327, p=.7439; NA, 2000: N=15, z=−2.021, p<.05). The result suggests that, in 1997, HRMs copulated more than other males whether natal males were included or not. However, in 2000, some natal males mated frequently, though with some females that were not closely related (Table II). In 2000, the HRMs mated more than other males when I excluded natal males from the test, but they may not have mated more than other males when I included natal males.

HRMs mated more than other males only in some years. The mean (±SD) number of receptive females per day for B is 1.44±.80 in 1996 and 1.73±.63 in 1997, whereas that for NA is 1.40±.99 in 1999 and 1.00±.71 in 200. The figures are smaller than the corresponding data from 1997 for NA (2.42±1.177; Soltis et al., 2001), when the α-male monopolized the females. In 1999, the dominant males could not monopolize copulations in NA; however, it is not possible to explain the result on the basis of the number of receptive females alone.

I further tested the priority of access model by calculating the estimated copulation frequency of α-males, assuming that they can completely monopolize 1 female in a day. I calculated the expected rate of copulations according to the priority of access model (1/the number of receptive females in the troop). On the days when I observed >2 females, I calculated the expected rate of copulations as the number of females sampled/the number of receptive females in the troop.

In 1997, the α-male’s mating success is not significantly different from the prediction of the model (Wilcoxon signed rank test: z=1.29, p=.20; N=25); in 1996, 1997, and 2000, α-males were unable to monopolize copulations as the model predicts, and they failed to monopolize them even when there was only 1 estrous female in the troop (Fig. 1; z=1.29, p=.20; N=25 Wilcoxon signed rank test: 1996: z=4.17, p<.0001, N=24; 1999: z=4.6, p<.0001, N=32; 2000: z=2.9, p<.01, N=21).
Fig. 1.

Observed and expected copulations of the α-males in relation to the number of receptive females in the troop. The number in parentheses represents the number of observation days. On the days when only 1 female was receptive, I calculated the expected rate of copulations (▪) as 1/the number of receptive females in the troop. The observed rate of copulation (□) is the rate of focal females that mated with the α-male. On days when I observed >2 females, I calculated the expected rate of copulation as the number of females sampled/the number of receptive females in the troop. The observed rate of copulations is the rate of the focal females that mated with the α-male. When neither of the focal females mated with the α-male, the observed rate of the copulation is 0 (Wilcoxon signed rank results: **p<.01, ***p<.001).

Some males took over and attained the top-ranking position in B in 1996 and in NA in 1999. The NTMs that dominated over the troop’s α-males may be part of the reason for the lack of monopolization. I again tested the priority of access model in the same manner for α-males and α-attitude NTMs, but only for the period during which the males behaved as top-ranking males (Fig. 2). In 1996, PE, who took over the troop, failed to monopolize copulations (Wilcoxon signed rank test: BB: z=2.02, p<.05; N=6). However, NTM MR’s rate of copulation was not significantly different from the expected rate (z=.533, p=.59, N=15). In 1999, ID, the α-male that sustained severe injuries this year, failed to monopolize copulations after α-attitude NTM BT’s disappearance (ID++: z=3.72, p<.001; N=22); however, BT’s data are not significantly different from the expected rate (z=4.15, p=.68, N=9). I did not test the cases of BB (α-male in 1999) and ID (before BT’s take over) because the sample size is too small, suggesting that the failure of monopolization by α-males is attributed partly to the presence of NTMs that took over the troop temporarily during their visiting. However, one cannot explain all cases in such a manner because some α-males failed to monopolize copulation even when there were no such NTMs with α-male attitude.
Fig. 2.

Observed and expected rate of copulation of α-males and NTMs during their time as top-ranking male in troop B in 1996 (a) and troop NA in 1999 (b). ID+; data before BT took over the troop. ID++; data after BT’s disappearance from troop NA. The number in parentheses represents the number of observation days. I calculated observed and expected rate of copulation in the same way as in Fig. 1 (Wilcoxon signed rank results: *p<.05, ***p<.001).

Number of NTMs

The number of NTMs is positively correlated with the observed copulation rate of the NTMs (Fig. 3; X=22.2±37.3%; Spearman’s rank correlation: rs=.574, p<.0001; N=109 d) and is negatively correlated with that of the HRMs (X=41.8±44.2%, rs=−.198, p<.05). There is no significant correlation between copulation with LRMs and the number of NTMs (X=27.3±34.9%, rs=.161, p=.095). Apparently, on the days when NTMs appeared, the monopolization rate of HRMs decreased because the females copulated with them. The NTMs include ones with α-male attitude and also ones that stay only on the periphery of the troop.
Fig. 3.

Copulation rate of HRMs and NTMs in relation to the number of NTMs visiting the troop. The number of NTMs that visited the study troop was correlated with the rate of NTMs’ copulation activity (X=22.2±37.3%; Spearman’s rank correlation: rs=.574, p<.0001; N=109 days). The number of NTMs that visited the troop correlates negatively with the rate of HRMs’ copulation activity (X=41.8±44.2%, rs=−.198, p<.05).

The number of NTMs that appeared in the vicinity of the study troop changed daily. The number is also positively correlated with the number of receptive females (Fig. 4a; Spearman’s rank correlation rs=.158, p<.05) and negatively correlated with the number of troop males (Fig. 4b; rs=−.228, p<.0001), suggesting that NTMs appeared in the vicinity of the troop on days when there were several receptive females and a small number of troop males. However, even when the number of receptive females was 1 or 0, 2 NTMs appeared in the vicinity of the troop on average, which could disturb the copulation success of HRMs (Fig. 3).
Fig. 4.

Number of NTMs visiting a troop in relation to the number of receptive females or the number of males in the troop. The number in parentheses represents the number of observation days. I excluded females and males that were absent from the troop, as well as nontroop males that interacted with only absent females and not with other troop members.

Table III.

Total number of days and the percentage of absence of the receptive female from the troop

 

Female

Number of days when the female was receptive

Number of days the female was absent

% absence

1996

St

17

2

11.7

 

Sr

8

0

 .0

 

Tr

17

0

 .0

 

Bb

15

1

6.67

1997

Ks

8

0

 .0

 

Th

12

0

 .0

 

St

17

2

11.7

 

Sr

18

0

 .0

 

Bb

15

0

 .0

1999

Ht

25

16

64.0

 

Rm

37

9

24.3

 

Sj

13

6

46.2

 

Tm

43

6

14.0

2000

Hn

16

0

 .0

 

Ht

34

18

53.0

 

Sk

13

0

 .0

Note. I recorded the female as absent from the troop if we did not locate her in the troop for >2 h after locating the troop. We located all the other nonreceptive troop females on a particular day and recorded them as receptive during the absence if they were receptive on the day before and after the absence.

On average, I recorded receptive females as absent from the troop on 15% (0–64%) of all observation days (Table III). In most cases, I could not determine their locations and the identities of their companions. I located them within the home range of the troop on only 10 d. They were with a NTM on 7 d, with a LRM on 3 d, and never with HRMs. They were mating, sitting together maintaining contact, or grooming each other. Focal observations of a receptive female with a NTM were difficult because in most cases the male was not habituated to humans and attempted to move away. I successfully observed a female with a NTM for >4 h on 2 d, and in both cases, she consorted with only 1 NTM, copulating repeatedly without returning to her troop.

Mounting Series with Male Neighbors and the Effect of Copulation Interruptions by Other Males

Interruptions of copulations occurred in 17 mounting series, with higher-ranking males accounting for 94% (16/17) of the interruptions. They accounted for only 14.4% (17/118) of all observed mounting series, excluding the ones involving the α-male. One reason for the low overall rate of interruption is that during the mounting series, the nearest neighbors were most often males ranked lower than the partner males (Table IV). Copulation with a β-male, when no higher-ranking male was nearby, accounted for 66.7% (4/6) of all mounting series, 88% (50/57) with LRMs, and 52.8% (29/56) with natal males.
Table IV.

Number of mounting series with higher-ranking males as the nearest neighbor

Status of male partner

With higher-ranking male nearby

Total

β-Male

2 (1)

 6

LRM

7 (4)

57

Natal male

27 (12)

56

Total

36 (17)

119

Note. The data contain only the mounting series in which the male partner is not an α-male. The number in parentheses is the number of mounting series that is interrupted by other males.

aThe number of mounting series not interrupted by males (including the mounting series that were not completed for unknown reasons).

When the nearest male to the mating pair was higher in rank than the male partner, the percentage of interruptions was 50% (1/2) with β-males, 57% (4/7) in the mounting series with LRMs, and 46.1% (12/27) with natal males. Half of the mounting series were interrupted when a higher-ranking male was the nearest neighbor.

However, the interrupters generally failed to persuade the pairs to discontinue copulation. On 9 occasions (52.9%), the pair resumed, and the mounting series ended in ejaculation. On 2 occasions (11.7%), the female joined the interrupter, and the mounting series ended in ejaculation. Even if the mounting series were not interrupted by other males, only 6.2% (44/73) of the mounting series with male neighbors were completed up to the point of ejaculation, while the rest were left incomplete for unknown reasons. In the 29 mounting series that subjects completed, the females changed their partner for the next mounting series 41.4% of the time. Therefore, it remains unclear whether an interrupter can increase his copulation opportunity through interruption. Conversely, individuals carried out 48 mounting series on the periphery of the troop, without neighbors nearby—including males and females—ending in ejaculation 83% (40/48) of the time. The probability of ejaculation on the periphery is significantly higher when neighbors were not nearby than when other male neighbors were around [6.2% (44/73) mounting series ended in ejaculation (χ2=12.15, p<.001)]. Therefore, the mere presence of other male members may disturb the mounting series and prevent ejaculation regardless of whether interruption was intentional or not.

Next, I tested the longer-term effect of interruption throughout the mating season. Twenty-one males participated in the interruptions. Twelve (57.1%) interrupters had copulated with a female sometime before the interruption occurred during the mating season. Of the interrupters, 7 (58.3%) copulated with the female sometime after the interruption. Conversely, 9 (43.9%) of interrupters had not copulated with the female before the interruption, and 6 (66.6%) of them copulated after the interruption. In brief, 18 of 21 male interrupters (85.7%) copulated before or after the interruption during the mating season; however, the behavior seems to neither increase nor decrease the chance of copulation with that female throughout the mating season.
Table V.

Number and percentage of copulations on the periphery of the troop and the status of male partner

Status of partner

Year

Female

Number of copulations on the periphery

Total number of copulations

%

HRM

1999

Ht

0

5

.0

  

Rm

0

5

.0

  

Sjd

0

2

.0

  

Tm

0

2

.0

 

2000

Hn

0

4

.0

  

Ht

0

2

.0

  

Sk

0

15

.0

  

Average

  

.0

LRM

1999

Ht

1

5

2.0

  

Rm

1

4

25.0

  

Sj

9

12

75.0

  

Tm

6

13

46.2

 

2000

Hn

0

0

  

Ht

2

8

25.0

  

Sk

2

3

66.7

  

Average

  

43.0

Natal male

1999

Ht

0

4

.0

  

Rm

1

5

2.0

  

Sj

0

1

.0

  

Tm

0

7

.0

 

2000

Hn

1

10

1.0

  

Ht

4

9

44.4

  

Sk

1

2

50  

  

Average

  

18.4

NTM

1999

Ht

2

2

100  

  

Rm

4

8

5.0

  

Sj

3

3

100  

  

Tm

3

6

5.0

 

2000

Hn

0

0

  

Ht

0

0

  

Sk

0

0

  

Average

  

66.7

Copulation on the Periphery of the Troop

Copulation on the periphery of the troop involved LRMs, natal males, and NTMs, but not HRMs (Table V). The mean percentage of copulation on the periphery of the troop is 0 with HRMs (N=7), 43.0%±23.5 with LRMs (N=6), 17.8%±21.5 (N=7) with natal males, and 75%±28.9 (N=4) with NTMs. BT, who attained the top-ranking position in early October and remained in the troop until mid-October in 1999, also copulated on the periphery in 5% (5/10) of the instances. Females tended to copulate on the periphery of the troop more frequently when they mated with LRMs or NTMs than with HRMs (Steel-Dwass test, HRMs vs. LRMs; t D 4.61, p<.01. HRMs vs. NTMs; t D 4.33, p<.05). Copulations on the periphery of the troop occurred as much as 40% (10/25) of the day with only 1 receptive female in the troop, 45% (9/20) of the day with 2, and 50% (4/8) with 3. The mean rate of copulation by the α-male is .078±.26 (N=23) on the day when copulation on the periphery of the troop occurred, and .34±.45 (N=30) when it did not occur, which are significantly different (Mann-Whitney U test, U=235.5, p<.05).

I then investigated how active the females were in their participation in the copulations on the periphery. Some close proximity relationships ≤1 m involved female initiation of close proximity to the male on the periphery or her movement toward the periphery with the male. I recorded initiation of a close proximity relationship with males on the periphery of the troop on 64 occasions. I noted close proximity with males of different status on the periphery of the troop, with 5.69% (12/211) accounted for by HRMs, 2% (30/158) by LRMs, 1.7% (14/131) by natal males, and 29.4% (10/34) by NTMs. Whether females tended to initiate close proximity relationships with neighbors nearby or on the periphery of the troop also differed depending on the status of their male partners (Table VI; χ2=18.4, p<.05). Females were present less frequently on the periphery of the troop when they initiated close proximity relationships with LRMs and NTMs than with HRMs (HRM vs. LRM: χ2=12.44, p<.0005; HRM vs. NTM; χ2=14.6, p<.0005). Females tended to initiate more close proximity relationships with LRMs and NTMs on the periphery of the troop than with HRMs.
Table VI.

Times a female approached or was approached by males to start close proximity ≤1 m divided by the presence of neighbors at that time

Status of male partner

On the periphery of the troop

With neighbors nearby

HRM

12

211

LRM

30

158

  

\(\left.\vphantom{\frac{1^*}{2}}\right\}\!*\!* \)

Natal male

14

131

  

\(\left.\vphantom{\frac{1^*}{\int_\int}}\right\}\hbox{ns}\)

NTM

10

34

  

\(\left.\vphantom{\frac{1^*}{\Big(}}\right\}\!*\!*\)

**p<.0005.

Females moved toward the periphery with male partners on 52 occasions. Before moving, the female sometimes glanced at the male, raising her tail and soliciting him to copulate. The frequency of females moving toward the periphery correlates with her copulating frequency with the male she sought or followed (Spearman’s rank correlation; N=60, rs=.419, p<.01), but only when LRMs were involved (Fig. 5; LRMs: N=24, rs=.563, p<.01) and not when HRMs, natal males, and NTMs were involved (Fig. 5; HRMs: N=13, rs=.091, p=.752; natal males: N=12, rs=.056, p=.0622; NTMs: N=11, rs=.194, p=.540).
Fig. 5.

Frequency of females moving toward the periphery and copulating frequency in each dyad. Spearman’s rank correlation: HRMs: N=13, rs=.091, d=.752; LRMs: N=24, rs=.563, p<.01; natal males: N=12, rs=.056, p=.062; NTMs: N=11, rs=.194, p=.540. Each dot represents a male-female dyad.

DISCUSSION

Interannual Changes in the Copulation Success of HRMs

With regard to their male partners, 56% of the cases showed that receptive females copulated most frequently with the α-male throughout the mating season (Table II). Even in small troops with few receptive females (average 1.00–1.73), the mating success of HRMs varies over years and between troops (Table II). The most important novel finding of this study, compared to previous studies, is that some α-males failed to monopolize copulations even when the number of receptive females was only 1 or 2 (Fig. 1), which shows that the model of α-male is insufficient in the population. Further, some females considered to be receptive were absent from the troop (Table IV), and may have copulated with LRMs and NTMs, suggesting that the data may be biased toward copulation with HRMs. Therefore, the small size of a troop may lead to female monopolization by HRMs in most cases, but does not guarantee it. The variation may be explained by injury status of the HRMs, the number of NTMs, and female choice.

Injury Status of HRMs

The injury status of HRMs may explain the difference in copulation success of top-ranking males. The α-male of troop B in 1997, MR, attained the top-ranking position the previous year by taking over the troop. He was considered a prime adult male superior to all the troop males and NTMs that appeared in the vicinity of B in 1996. He did not suffer from any injury or disease in 1997, and there was no sign of decline in his injury status. Contrarily, heavy injuries may explain the low mating activity of ID, the α-male of troop NA in 1999. The injury on the palm of his hand made it difficult for him to walk and to run quickly.

HRM’s mating tactics are associated with long periods of consort and aggressive behavior toward estrous females and other troop males (Huffman, 1991; Matsubara, 2004; Soltis, 1999). Interruption of mounting series is often considered an attempt by interfering individuals to disrupt or delay copulation between a pair, thus improving the interrupter’s own reproductive potential (Dewsbury, 1982; Hrdy, 1977; Stephenson, 1975; Tutin, 1979). Interruption by males did not change the number of incomplete mounting series as the previous study of the same speacies (Huffmann, 1987). However, mounting series on the periphery of the troop ended in ejaculation more often than mounting series with other males in sight. I also showed that the rate of mounting series with males higher in rank than their male partner was small. Mating success of HRMs may be due to the mental pressure on lower-ranking males rather than direct aggressive behavior directed toward the mating pair.

Soltis (1999) and Matsubara (2004) showed that the following approach by HRMs increased their mating success. The tactics require energy and reduce feeding time (Matsubara, 2004). In Yakushima, male rank is associated with age and HRMs are usually in prime age and physical condition (Sprague, 1992). However, at the same time, aggressive NTMs frequently visit small troops and try to take them over, which increases the risk of injury for HRMs (Sprague, 1992; Yamagiwa, 1985). In case of severe injury of HRMs that precludes aggressive behavior, the females may easily move to the periphery of the troop.

Presence of NTMs

On the days when NTMs appeared, monopolization by HRMs decreased. Visits by NTMs diminish the immediate copulation success of HRMs from 3 perspectives. 1) Receptive females are usually attracted to unfamiliar males and if they find one, will try to leave the HRM (Fig. 4; Okayasu, 2001). 2) Some of the NTMs are dominant over HRMs on a one-to-one basis and monopolize receptive females (Fig. 2). If NTMs act aggressively, HRMs may sustain severe injury (Sprague, 1992; Yamagiwa, 1985).

The number of NTMs correlates positively with the number of receptive females in the troop and negatively with the number of troop males. The number of troop males relates to female defensibility by troop males and correlates with the size of the troop and the number of troop males absent from the troop. Yamagiwa (1985) suggested that the possibility of copulation by NTMs depends on the effectiveness of the troop males’ cooperative defense against them. Cooperative aggression, particularly toward the NTMs that had taken over the troops, occurred; however, its effect is unclear (Sprague, 1992).

The females tended to copulate on the periphery of the troop when the partner was not a HRM. The male partner is the only guardian of a receptive female in the periphery. The dispersal of females from the center of the troop makes it difficult for HRMs to monopolize them, and NTMs can solicit females at the periphery easier than at the center of the troop, where multiple guardians are present (Table III). Some receptive females were absent from the troop and mated with NTMs and LRMs. The females’ behavior, together with copulation on the periphery, will increase the chance for NTMs to mate with receptive females.

NTMs that established definite dominance relationships had high mating success (Fig. 2). I categorized them as NTMs but they dominated all the troop males, though they sometimes faced cooperative aggression by troop males. Therefore, high rank has a definite effect on the mating success of males. However, the effects of rank and novelty are difficult to distinguish. The high-ranking males also mated with receptive females on the periphery of the troop before and after they dominated troop males (Table V). Hence without cooperation from the females, they would not have achieved high mating success.

Female Choice and Copulation on the Periphery of the Troop

Copulation on the periphery of the troop increased the probability that a mounting series would end in ejaculation. Previous authors reported copulation out of sight of the other males (rhesus macaque: Brereton, 1992; Japanese macaque: Berard et al., 1994; Okayasu, 2001; Wolfe, 1986). Copulation on the periphery occurred even when the number of receptive females in the troop was only 1, in which case the rate of copulation with the α-male decreased significantly. The female moving toward the periphery with a male increased their probability of copulating (Fig. 5). The initiation of a close proximity relationship on the periphery of the troop was higher with LRMs and NTMs than with HRMs, suggesting that LRMs and NTMs tend to stay on the periphery of the troop compared to HRMs. My study also showed that receptive females sometimes were absent from the troop (Table III), which increased the chance of mating with LRMs and NTMs. Females are never passive partners, but actively solicit males and when necessary mate on the periphery of the troop with LRMs and NTMs.

However, females of troop B in 1997 mated with HRMs significantly more often than with other males. The relative unfamiliarity of the α-male with the females and the small number of NTMs indicate female choice for HRMs. Soltis (1999) observed a variety of interactions between females and HRMs concerning male coercion and female choice. The highly variable length of residency and the injury status of α-males help explain the variety of interactions between HRMs and females.

Interaction Between HRMs and Females in Small Troops on Yakushima

In Yakushima, fully grown adult males often enter small troops from the top-ranking position (Suzuki et al., 1998). My study revealed the new α-males enjoyed high mating success, though it was not guaranteed every year. Success fluctuates according to injury status, presence of NTMs, and female choice. Suzuki et al. (1998) reported that mature males tend to enter troops with fewer resident males, which may be a tactic to decrease the possibility of attack by resident males to ensure mating success the following year.

For females, copulation with multiple males may help to confuse paternity, which can reduce the risk of infanticide (Hrdy, 1977; Smuts, 1987; Soltis et al., 2000). At the same time, copulation with multiple males incurs a cost on females by decreasing their feeding time and increasing the distance they travel in a day to places where other males will not disturb mating and to release them from the aggression of HRMs (Matsubara and Sprague, 2004). Females mated with HRMs that had no injury and when the number of NTMs was small. Females may choose HRMs as copulation partners when the cost of copulation with multiple males is high and the risk of infanticide by other males is small. In Yakushima, though there is only 1 reported case of infanticide, males have harassed infants of mothers they did not mate with, which may influence infant survivorship and also female mate choice (Soltis et al., 2000).

When females mate with LRMs or NTMs, they tend to copulate on the periphery of the troop. The dispersal of females makes it even easier for NTMs to solicit out of sight of troop males. After copulation with NTMs, females return to the center of the troop, sometimes followed by the NTM (personal observation). The behaviors would increase the chance of aggressive behavior between HRMs and NTMs with α-male attitude, and possibly initiate troop takeover by them. The suggestion is supported by the idea of Yamagiwa (1985) that females recruit males as guardians of the troop to defend their home range.

ACKNOWLEDGMENTS

I thank Professors Yukimaru Sugiyama, Osamu Takenaka, Jyuichi Yamagiwa, and the late Shigeo Uehara for their constant support and critical comments. I express special thanks to Dr. Michael A. Huffman for his discussions on mainland Japanese macaques and his advice on the technical aspects of writing an article. Thanks are also due to Dr. Hideo Ohsawa, Professor Akio Mori, Dr. Hideki Sugiura, Dr. Yasuyuki Muroyama, Dr. Chie Hashimoto, Dr. Hiroyuki Takemoto, and colleagues at the Ecology and Behavior Department for their constant advice and encouragement for the work. I also thank Dr. Miki Matsubara, Dr. Shigeru Suzuki, and Dr. Yukio Takahata for guiding me in following and identifying individual wild Japanese macaques of Yakushima. Also, thanks to all members of the Yakushima Research Group for their discussions on Yakushima Island. I express special thanks to Dr. Joseph Soltis of Disney World, USA, with whom I shared the most information and discussions on this topic. A Research Fellowship from the Japan Society for the Promotion of Science for Young Scientists (No. 9611) supported the research.

Copyright information

© Springer Science+Business Media, LLC 2007

Authors and Affiliations

  1. 1.Primate Research Institute, Kyoto University, KanrinInuyamaJapan

Personalised recommendations