Res. Popul. Ecol. (1990)32, 391--406. ~) by the Society of Population Ecology

LIFETIME REPRODUCTIVE

S U C C E S S IN R E P R O D U C T I V E L Y

SUPPRESSED FEMALE VOLES Takashi SAITOH Wildlife Management Laboratory, Hokkaido Research Center, Forestry and Forest Products Research Institute, Hitsujigaoka 1, Toyohira-ku Sapporo 062, Japan

INTRODUCTION

Reproductive suppression and failure under crowded conditions are seen in females of many mammals (reviewed by Wasser and Barash, 1983). Many laboratory studies of rodents have shown that social odors cause pregnancy failure and suppression of estrus (reviewed by Brown, 1985; Milligan, 1980; Richmond and Stehn, 1976; Rogers and Beauchamp, 1976), but some authors (Bronson, 1979; Bronson and Coquelin, 1980; Rogers and Beauchamp, 1976) consider reproductive suppression or failure to be laboratory artifacts. Many studies, however, report on the occurrence of suppressed maturation and pregnancy failure in field or semi natural populations (in an outdoor enclosure) of microtine rodents (Bondrup-Nielsen, 1986; Bujalska, 1970, 1973; Gilbert et al., 1986; Heske and Nelson, 1984; Ims, 1985;Jensen and Gustafsson, 1984; Kawata, 1987; Mallory and Clulow, 1977; Nakata, 1989; Saitoh, 1981; Westlin and Nyholm, 1982), and even laboratory workers have recently begun to pay attention to the ecological significance of reproductive suppression and failure (Bronson, 1985; Haigh, 1987; Vandenbergh, 1987). Why is reproduction suppressed in some females? Wasser and Barash (1983) proposed a model for the phenomenon: females can optimize their lifetime reproductive success by suppressing reproduction when future conditions for the survival of offspring are likely to be sufficiently better than present ones as to exceed the costs of the suppression itself(the Reproductive Suppression Model). Do some females, otherwise, sacrificially suppress their own reproduction for a population not to be overcrowded? The latter "group selection" idea has been severely criticized in the theoretical field (e.g., Maynard Smith, 1982; Wilson, 1975), but there have been few adequate observational or experimental studies on this point. Empirical data are needed to test these ideas. The grey red-backed vole, Clethrionomys rufocanus bedfordiae (Thomas), is the most common small mammal in Hokkaido, Japan. Maturation of young female voles is suppressed in high density populations (Nakata, 1989; Saitoh, 1981) and female voles whose home ranges overlap with other females fail to become pregnant (Kawata, 1987).

392

T h e purposes of this study are to observe reproductive suppression in females of a grey red-backed vole population in an outdoor enclosure, to compare lifetime reproductive success between reproductively suppressed and non-suppressed females, and to test the Reproductive Suppression Model.

MATERIALS AND METHODS

An outdoor enclosure was used for the experiment in the T o m a k o m a i Experimental Forest of Hokkaido University, iocated in the south-central part of Hokkaido (the northernmost island of J a p a n ; 42~ 141~ T h e climate is intermediate between temperate and subfrigid. T h e enclosure with an area of 2.1 ha (100 by 210 m) was constructed in a secondary broad-leaved forest having almost uniform undergrowth, and was fenced with galvanized iron sheets extending 0 . 9 m u n d e r g r o u n d and 0.9 m above ground. The outer fences were successful in limiting vole movement. This enclosure was divided into five parts with either a high (0.9 m) or a low fence (0.45 m) (broken lines in Fig. 1) for other experiments which had been done, but each inner fence had a total o f t e n small holes (20 by 10 cm) at 10 m intervals so that the voles could move among the parts. T h e grey red-backed vole commonly occurs in the surrounding area of the enclosure, although it is less dominant than the wood mouse, Apodemus argenteus. Some species of small mammals entered the enclosure and were removed when they were trapped. T h e i r influence on the vole population was presumed to be negligible, because their numbers were small and their ecological niches were different from that of the vole. Food (about 150 g), consisting of oats (80%) and rabbit chow (20%), was provided at 100 trap stations on grid B (100 by 100 m; Fig. 1) at 10-14 day intervals as extra food throughout the study. Grid A (100 by 100 m) as a control was intact except for trapping and grid AB (0.1 ha) connecting grids A and B was distinguished from grid A and grid B by different vegetation because this part was heavily disturbed at the construction of the enclosure. Twenty-six voles (17 females and 9 males) were taken from a windshelter-belt of the Ishikari Plain (about 80 km N of the study area) on 15 September 1984 and released into the enclosure on 16 September 1984. T h e population densities, body mass of individuals, and reproductive condition were similar between control and experimental grids at release. T r a p p i n g in the enclosure was conducted from September 1984 to N o v e m b e r 1986. T h e catch-mark-release method was followed using Sherman-type live traps at 210 trap stations distributed in a 10 by 10 meter grid pattern. T w o traps were set at each station and baited with oats. Cotton was placed in each trap for bedding year around, and all traps were covered by a wooden box with two holes to protect animals captured and to facilitate trapping in deep snow. A 5 day census was usually conducted twice a m o n t h (once in September 1984, February 1985, a n d N o v e m b e r 1986), and traps were examined twice daily (morning and night). T o avoid the chance of par-

393

T E

o

o

o

o

o

0

0

0

0

0

o

o

o

o

o

0

0

0

0

0

o

0

o

0

o

o

o

o

o

9

o

o

o

o

o

o

o

=

o

0

o

o

o

o

o

o

o

o

o

!

0

o

o

o

o

o

o

o

o

o

o

9

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

0

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

o

~

o

9

o

o

o

o

o

9

i

9

9

9

e

9

9

e

9

e

:

9

6

~ :6

0 0

I

9

9

e

e

e

9

9

9

I

o

o

o

l

A: control

i

9

9

1

4

9

~

e

I e

AB

e

e

9

B: experiment

Fig. 1. The enclosure in the Tomakomai Experiment Forest of Hokkaido University, Hokkaido, Japan. Outer fence (solid line); inner fence (broken lines) had small holes allowing the voles to move between grids. Trap stations (circles) were distributed at 10 m intervals with two live traps at each station. Trap stations on grid B (shaded) were simultaneously feeding sites.

turition by a vole in a trap and interference with lactation, traps were not set in areas of the home range occupied by females regarded to be in the late stages of pregnancy or at the most active stages of lactation, after their reproductive conditions had been observed o n a capture d u r i n g a t r a p p i n g period. V o l e s were t o e - c l i p p e d for i n d i v i d u a l identification. Upon first capture in any given trapping session, each vole's body weight was m e a s u r e d to the nearest 0.5 g u s i n g a b e a m b a l a n c e .

A n i m a l n u m b e r , sex,

reproductive condition (males: testes scrotal or abdominal; females: vagina perforate or imperforate; nipples small or elongated and bare indicating lactation; pregnancy recorded by body form and palpation), and location of capture were recorded for each capture. The vole then was released at the point of capture. The gestation period in this vole is 18-19 days under laboratory conditions (Abe, 1968). Pregnancy can be detected by palpation from about the 10th day of pregnancy and the parturition date can be predicted by the size of the embryo within an error of three days (Saitoh unpublished). Juveniles are weaned at around their 17th day of life and become trappable. Thus, if trapping is carried out twice a month, the reproductive career of each female can be recorded, because she must be trapped either at late pregnancy stage or during lactation. Mother-offspring relationship was inferred when a juvenile (body mass < 25 g and/or darkish pelage) was trapped in the territory of an adult female who had large nipples and/or had been pregnant 20-30 days before. I counted the number of juveniles that appeared on a given female's territory as the number of her offspring. Lifetime reproductive success was measured according to three components for each female; the number of pregnancies, the number of successful litters ([a successful litter] -- [a pregnancy which produced at least one trappable offspring]), a n d the n u m b e r o f offspring w h i c h b e c o m e trappable (the a p p r o x i m a t e

394

value of lifetime reproductive success). T h e birth date of a juvenile was estimated from its body mass and the reproductive career of its estimated mother. T h e age at the start (implantation) of the first pregnancy was used as the index of the start of reproduction. This age was estimated from the stage of pregnancy and/or the estimated parturition date. Age at the first pregnancy and lifetime reproductive success were analyzed for females who were born on a grid and bred on the same grid; the data of females who traveled between grids and whose home range covered different grids was omitted, because lifetime reproductive success was affected by population density.

DEFINITION AND ASSUMPTIONS

T h e r e are some difficulties for testing the model. I must compare the lifetime reproductive success between the reproductively suppressed females and females who attempt to breed at a younger age in an inadequate environment for reproduction. However, it is almost impossible to divide the environment surrounding every individual into adequate and inadequate. Thus, I will compare the lifetime reproductive success between reproductively suppressed females and non-suppressed ones; these females are defined as the females who become pregnant at an older age (late reproducing females), and as the females who become pregnant at a younger age (early reproducing females), respectively (see the results for practical procedures). T h e above mentioned definition is based on the following assumptions: all female voles have a character on reproduction which stereotypically reacts to an environment; female voles always breed as soon as possible in an adequate environment, whereas they always suppressed their reproduction in an inadequate environment. But actually, early reproducing females probably involving two types of females; the females who breed early in an adequate environment and females who attempt to breed at an younger age in an inadequate environment.

RESULTS

Density and Reproduction M i n i m u m n u m b e r alive is a reliable index of the actual density when trap responsiveness is high (Hilborn et al., 1976; the n u m b e r of trapping sessions during which an individual was captured divided by the n u m b e r of sessions when the individual was exposed to capture during its life span), which it was in my study ( > 9 7 . 5 % on each grid). Densities on the experimental grid (1.0 ha) were always greater than those on the control grid (Fig. 2; Wilcoxon's signed-ranks test, ts=6.22, P < 0 . 0 0 1 ) , although density differences changed seasonally. Reproductive output (the n u m b e r of juvenile recruits) was the most effective factor determining the density differences in comparison with survival rate and dispersal (immigration and emigration) (see Saitoh,

395

Fig. 2. Densitiesof vole population on the control grid (solid line), and the experimental grid to which food was added (broken line). Periods of snow cover (> 10 cm) are shaded. 1989a for details). Almost all (96.5%) of u n m a r k e d voles which appeared in the enclosure were juveniles which could be born in the enclosure before they attempted to disperse. In m a n y cases, some juveniles which were similar body size to each other, were captured at the same or near trap sites during the same trapping period. T h e y could be litter mates. T h e juveniles appeared discontinuously on both grids (Fig. 3). T h e total n u m b e r of juvenile recruits was 1,093 (males: 583; female: 510). T h e experimental grid had 3.7-fold recruits of the control grid. Age at First Pregnancy Age at the first pregnancy was recorded for 197 of 510 females who were born in the enclosure; 49 females born after J u n e 1986 were not observed during enough periods to record their age at the first pregnancy and the remainder died before their pregnancies. Age at the first pregnancy varied widely from 30 to 198 days (Fig. 4). Although half of the females (50.3%) became pregnant before their 70th day, it was not u n c o m m o n for a female to become pregnant at an age of over 100 days. I analyzed age at the first pregnancy, and further on two generations (generation A: females born from F e b r u a r y to J u n e 1985, and generation B: females born from September to N o v e m b e r 1985) which had much variation in age at the first pregnancy and whose lifetime reproductive success could be measured. Age at the first pregnancy was related to the birth date in generation A. Almost all of the females born in February 1985 became pregnant at a younger age. However, in females born from M a r c h to J u n e 1985, females born earlier became pregnant at a later age whereas those born later became pregnant at a younger age (Fig. 5). T h e relationship between age

396

Fig. 3. Number of recruits (juveniles) first trapped during each period on the control and the experimental grids. Periods of snow cover (> 10 cm) are shaded. at the first p r e g n a n c y and birth date was linear in females b o r n from M a r c h to J u n e 1985 ( Y = 210.5-0.90 X , Fs = 239.3, P < 0.001).

These relationships were not different

between the control and the experimental grid, except for the fact that females who b e c a m e p r e g n a n t at over 150 days old were not observed on the control grid. Age at the first p r e g n a n c y varied widely regardless of the birth date in generation B (Fig. 5). Its variation was conspicuous on the experimental grid, while age at the first p r e g n a n c y exceeded over 100 days in all females of the control grid except for one. Because the distribution of age at the first p r e g n a n c y was continuous in the both generations, females who became p r e g n a n t at an age of less than 80 days and females who did at an age of over 100 days were tentatively classified into the early and the late reproductive type, respectively; the data on females who b e c a m e p r e g n a n t between the ages of 81 and 100 days were eliminated in the following analyses. Lifetime R e p r o d u c t i v e Success Lifetime reproductive success was analyzed only for females on the experimental grid, because age at the first p r e g n a n c y was less variant on the control grid. Lifetime reproductive success was shown by the three c o m p o n e n t s for each female; i.e., the n u m b e r of pregnancies, the n u m b e r of successful litters, and the n u m b e r of offspring which b e c a m e trappable. Those values could be m e a s u r e d for most of the 196 females b o r n by N o v e m b e r 1985 in the enclosure; the reproductive

397

Fig. 4. The distribution of age at the first pregnancy. The whole histogram shows the distribution of the ages of all females whose age at the first pregnancy could be measured; i.e., for females born from November 1984 to May 1986 on grid A, gird B, and grid AB. The light shaded areas show the distribution of the ages for females born from February to June 1985 (generation A) on the experimental grid, and the dark shaded areas show that for females born from September to November 1985 (generation B) on the experimental grid. career of females b o r n after that could not be traced completely because a considerable n u m b e r of those females were still alive at the end of the investigation.

Females whose

ages at the first p r e g n a n c y were highly variant, were ones b o r n from F e b r u a r y to J u n e 1985 (generation A), and ones b o r n from S e p t e m b e r to N o v e m b e r 1985 (generation B; Fig. 4). I classified t h e m into the early reproductive type and the late reproductive type, as mentioned above.

I added six females who had lived for m o r e than 100 days

without b e c o m i n g p r e g n a n t to the late reproductive type (two in generation A and four in generation B), because their reproduction could be suppressed until the age of their death. T w e n t y of thirty-two females of the early reproductive type (62.5%; 11 of 17 in generation A and 9 of 15 in generation B) succeeded in raising their offspring; the reproductive success of the two early reproducing females (generation A) could not be m e a s u r e d owing to the uncertainty of the mother-offspring relationship.

A similar

percentage (60.9%) of the late r e p r o d u c i n g females also succeeded in raising offspring (8 of 13 in generation A and 6 of 10 in generation B). T h e n u m b e r of pregnancies was higher in the late reproductive type in generation

398

200 -

A

''",,.,'".'".'"')'"'"",,,,,,,,,,., B

u ce-

150

.~100 O t-

000 9

9

,<

9 50

;

Oe

9

~Q

6, .$

30 ; ,;, +o ; ;o 20 Feb.

Mar.

; ,o ~;, ; ,o ~o Apr.

; ,o ;o 3o

May

Date

9 ,o ;o

Jun.

of

Birth

Sep.

; ,3, ;o Oct.

', ,'o ~o 3'o Nov.

in 1985

Fig. 5. The relationship between age at the first pregnancy and birth date. Left half: females born from February to June 1985 (generation A); right half: females born from September to November 1985 (generation B); triangles: females born and breeding on the control grid; circles: females born and breeding on the experimental grid. The circle with the figure of 1 denotes the female who was born on the experimental grid, bred on the control grid, and bred again on the experimental grid. Symbols plotted on over 200 days of age indicate females who died at over 100 days old before becoming pregnant. An regression anaoysis was carried out on females enclosed by a dotted line (Y=210.5-0.90X, Fs=239.3, P< 0.001). A, whereas that was lower in the late reproductive type in generation B; the difference was significant only in generation B ( M a n n - W h i t n e y ' s U-test, U = 9 9 . 0 , P > 0 . 1 0 for generation A; U = 105.0, P < 0 . 0 0 5 for generation B; T a b l e 1).

T h e n u m b e r of suc-

cessful litters was also higher in the late reproductive type in generation A, whereas it was lower in the late reproductive type in generation B; both of these differences were not

significant ( M a n n - W h i t n e y ' s

U=71.5,

P)0.10

U-test,

for generation B).

U=83.0,

P)0.10

for generation

A;

Although the n u m b e r of pregnancies was

significantly higher in the early reproducing females than in the late reproducing females in generation B, the n u m b e r of successful litters was not significantly different between them.

This m e a n s that m o r e pregnancies failed to produce any offspring in

the early reproducing females of generation B; the proportion of successful litters ([the n u m b e r of successful litters]/[the n u m b e r of pregnancies]) of the early reproducing females was lower (0.37) than that of the late reproducing females (0.70) in generation B (Fisher's exact probability test, P = 0.08); the proportion of successful litters of the early and the late reproductive types were similar to each other in generation A (0.49

399 a n d 0.45 for t h e e a r l y a n d the l a t e r r e p r o d u c t i v e t y p e , r e s p e c t i v e l y ) . T h e n u m b e r o f o f f s p r i n g was h i g h e r in the late r e p r o d u c t i v e t y p e in g e n e r a t i o n A , w h e r e a s it w a s l o w e r in t h e late r e p r o d u c t i v e t y p e in g e n e r a t i o n B; these differences w e r e n o t significant ( M a n n - W h i t n e y ' s

U-test, U = 8 3 . 0 ,

U = 71.5, P > 0.10 for g e n e r a t i o n B; T a b l e 1).

P>0.10

for g e n e r a t i o n A ;

T h e v a r i a t i o n o f t h e n u m b e r o f offspr-

i n g was l a r g e in e v e r y r e p r o d u c t i v e t y p e a n d g e n e r a t i o n ; t h e r a n g e ( C . V . ) was 0 - 1 5 ( 1 2 3 . 8 % ) a n d 0 - 1 4 ( 1 2 1 . 3 % ) for t h e e a r l y a n d the l a t e r e p r o d u c t i v e t y p e , r e s p e c t i v e l y , in g e n e r a t i o n A , a n d it w a s 0 - 1 2 ( 1 1 7 . 7 % ) a n d 0 - 8 ( 1 2 6 . 8 % ) for the e a r l y a n d the late r e p r o d u c t i v e t y p e , r e s p e c t i v e l y , in g e n e r a t i o n B.

Variance of the number of offspring

was h i g h e s t in t h e late r e p r o d u c i n g f e m a l e s o f g e n e r a t i o n A (34.16), a n d it w a s lowest in the late r e p r o d u c i n g s i g n i f i c a n c e was n o t P>0.05),

females

of generation

detected between

B (8.75).

these two

However,

generations

statistical

(F-test, F=3.904

as well as in a n y o t h e r c o m p a r i s o n s .

M e a n life s p a n s were l o n g e r in the l a t e r e p r o d u c i n g f e m a l e s in the b o t h g e n e r a tions (Mann-Whitney's P>0.10

U-test,

U=165,

P<0.001

for g e n e r a t i o n

A;

U=80.5,

for g e n e r a t i o n B).

O n t h e c o m b i n e d d a t a o f t h e two g e n e r a t i o n s , the n u m b e r o f p r e g n a n c i e s , the n u m b e r o f successful litters, a n d t h e n u m b e r o f o f f s p r i n g w e r e n o t different b e t w e e n t h e e a r l y a n d late r e p r o d u c i n g f e m a l e s ( W i l c o x o n t w o s a m p l e test, ts = 1.543, P > 0.10; t, = 0.247, P > 0.50; t, = 0.2 74, P > 0.50, r e s p e c t i v e l y ) , w h e r e a s life s p a n was significant-

Table 1. Lifetime reproductive success (the number of pregnancies, the number of successful litters, and the number of offspring) of females becoming pregnant earlier (the early type) and of females whose reproduction was suppressed (the late type). Type of No. of Generation reproduction females

A

early late

B

late early late

ar,,

Total

19 13a

Mean age at Mean No. Mean No. of Mean No. first of successful of pregnancy (days) pregnancies litters offspring 47.8 137.5

2.06 c ] 3.00 a INS

1.13 u } 1"36a NS

10"

122.8

1.11 f

0.78 f

2.33 f NS 221.2f J

34 23

57.0 132.4

2.52 }N S 2.15

1.14 }N S 1.10

3.79 }N S 150.1}**** 3.70 255.9

15

3.73 b } 4.82 a NS

Mean life span (days) 131 lc~ 28413ai****

s

s

Generation A: females born from February to June 1985; B: females born from September to November 1985. Generation B involved females who were still alive at the end of this study; one in the early and one in the late type. Asteriskes and "NS" represent a statistical significant level in a difference; three:0.5%; four:0.1%; a Females who died at age over 100 days before becoming pregnant, were considered reproductively suppressed ones; two in generation A and four in generation B. The mean ages at the first pregnancy were calculated without these females. b Two females were killed by trapping accidents and the reproductive success of two other females could not be observed. Because data of these four females were omitted for the calculation, the sample size of this value was 15. r d,e, and f The data of females killed by trapping accidents were omitted for the calculations. Thus, the sample size of the value was 17,11,14,and 9,respectively.

400 Table 2. Life span and results in reproduction of offspring born from females of the early and late reproductive type. Figures in parentheses are the numbers excluding the voles who were killed by trapping accidents; i.e., these are sample sizes of the mean life spans. Mother's generation

Mother's reproductive type

No. of voles

Life span mean-----S.D. (days)

No. of voles becoming pregnant

No. of successful voles in reproduction a

females A

B

early

30(26)

109.9•

11

9

late

24(23)

117.3-+ 86.5 )

93.9 IN S

13

8

early late

25 7

107.6+ 85.2u) 152.0+ 89.6cJ~NS

14 4

8 3

early late early late

26(22) 44(42) 29 14(13)

140.5+--105.6.~NS 110.0+ 79.5cJ 74.7--+ 56.5a~N S" 103.4• 71.9eJ

-----

males A B

D

m

"NS" represents a statistical signigicant level in a difference (non-significant). the number of voles who reared their offspring till becoming trappable. b,c,a, and e involve the data of the voles who were still alive at the end of the investigation; the number of these voles are four, one, one, and two, respectively. ly longer in the late reproducing females than in the early reproducing females (t, =3.657 P<0.001). Life span of female offspring was longer in the late reproducing females than in the early reproducing females in the both generations; that of male offspring was longer in the early reproducing females than in the late reproducing females in generation A, whereas the relationship was reversed in generation B; these differences were not significant (Wilcoxon two sample test, t , = 0 . 7 2 2 , P > 0 . 5 ; t , = 1 . 3 4 5 , P > 0 . 2 for daughters of generation A and B, respectively; t , = 0 . 9 7 6 , P > 0 . 2 , t, = 1.538, P > 0 . 1 for sons of generation A and B, respectively; Table 2). In daughters, the proportions of voles'who successfully bred were 30.0% and 33.3% in the early and the late reproducing females, respectively, in generation A, and they were 32.0~ and 42.9~ in the early and the late reproducing females, respectively, in generation B. T h e combined value of the two generations was not different between the reproductive types (30.9% and 35.5~ in the early and the late reproducing females, respectively; G-test, G = 0 . 1 3 9 , P > 0.50). Reproductive results of sons could not be measured. Sex Ratios of Offspring Sex ratio of offspring were nearly even in the early reproducing females (sons/daughters = 0.87 for generation A and 1.16 for generation B), whereas they were biased toward males in the late reproducing females (1.83 for generation A and 2.00 for generation B; Table 2); combined sex ratio of offspring in the late reproducing females was significantly deviated from a 1 : 1 ratio (G-test, G--8.275, P < 0 . 0 0 5 ) .

401 DISCUSSION Young female voles are reproductively suppressed by social factors in the genus of Clethrionomys; their maturation is delayed in high density populations (BondrupNielsen, 1986; Bujalska, 1970, 1973; Gilbert etal., 1986; Nakata, 1989; Saitoh, 1981) and females who have no territories fail to become pregnant (Kawata, 1987). The following facts of this study conformed to the above reports; age at the first pregnancy varied more in higher density populations on the experimental grid, and there was a negative regression between age at the first pregnancy and birth date when population density was decreasing (Fig. 5). Almost all of the females on the regression line in Fig. 5 began to breed, regardless of their birth dates, in the middle of September 1985 when the population density had decreased to its lowest point. By that time the competition for breeding resources may have eased. The late reproducing females could be considered as reproductively suppressed ones. However, the reproductive suppression of the females on the regression line cannot be explained by the breeding territoriality alone, because the females suppressed their reproduction till the middle of September 1985, although the number of adult females had already decreased, and the vacant areas might have been provided before that time. Other factors (e.g., food nutrition or an unknown social factor) must influence the time when a vole start to breed addition to the breeding territoriality. Did these females have any benefit for themselves in any case? Did they, otherwise, sacrificially inhibit their reproducing for a population not to become superabundant? Three components of lifetime reproductive success (the number of pregnancies, the number of successful litters, and the number of offspring) were not different between the early and the late reproducing females on the combined data of the two generations (Table 1). T h e relationship was, however, reversed between the generations; the measurements of the late reproducing females was higher than those of the early reproducing females in generation A, whereas those of the late reproducing females was lower than those of the early reproducing females in generation B, although these differences were not significant, except for the number of pregnancies in generation B. The reversed difference might be caused by the difference in the factors suppressing the reproduction; the breeding territoriality and other unknown factors. But, the combined data sets of the two generations could be better indicators for lifetime reproductive success because of non-significance in statistical tests (although there are some problems in that combination; fitness is influenced by population density and the time of reproduction). Thus, lifetime reproductive success could be similar between the two reproductive types, or it could be somewhat lower in the late reproducing females than in the early ones. The number of pregnancies was significantly higher in early reproducing females than in late reproducing females in generation B. However, the number of successful

402 litters was not significantly different between them, because the proportion of successful litters of the early reproducing females was lower than that of the late reproducing females. In generation B, all of the late reproducing females who became pregnant succeeded in raising their offspring, whereas six out of fifteen early reproducing females failed to produce offspring despite their pregnancies. Some of the six females may have attempted to breed at a younger age in an inadequate condition for reproduction; namely, the conditions where the females' reproduction must be suppressed. This tendency, however, was not observed in generation A. The late reproducing females were living for longer periods than the early reproducing ones. The longer life span of the late reproducing females may be influenced by the analysis method, because the data on the late reproducing females did not involve the data on the females who died at an age younger than 100 days. Thus, it is necessary to compare another measurement. The breeding period (the period after subtracting age at the first pregnancy from the life span) tended to be longer for the late reproducing females (146.8 days) than for the early reproducing females (83.3 days) in generation A, and the breeding period was similar between the late reproducing females (98.4 days) and the early reproducing females (104.5 days) in generation B. Thus, the delayed start in reproduction did not shorten the duration of the breeding period. The longer life spans of the late reproducing females could compensate for their delayed start in reproduction. To be exact for testing the adaptation of reproductive suppression, I must have compare the lifetime reproductive success between the late reproducing females and females who attempted to breed at a younger age in inadequate conditions for reproduction. However, many females of the early reproductive type may have bred in adequate conditions for reproduction. Thus, I probably compared the lifetime reproductive success of the late reproducing females with that of the superior ones. Nevertheless, the difference of lifetime reproductive success was so small that the late reproducing females achieved reasonable lifetime reproductive success by suppressing their reproduction. These facts suggest that the reproductively suppressed female voles did not sacrifice themselves for a population and support the Reproductive Suppression Model by Wasser and Barash (1983). Life spans of offspring were not different between the reproductive types, whereas the sex ratio of offspring was different between them (Table 2). McShea and Madison (1986) reported a biased sex ratio in juvenile recruitment in a population of meadow voles, Microtus pennsylvanicus; more female juveniles appear in the spring, whereas more male juveniles appear in fall. They explained the biased sex ratio in connection with the social factors. Female offspring at the onset of the breeding season (i.e. low densities) would have less difficulty securing territories than would female offspring later in the breeding season, who would have to compete with their mother and other larger females for space. Male offspring born in the spring would have to compete with overwintered (probably superior) males, whereas male offspring born in the fall

403 would successfully breed as overwintered males during the next spring. Thus, a mother would get more grandchildren by producing more daughters in the spring and more sons in the fall. The social system of meadow vole is similar to the grey red-backed vole; breeding females of the grey red-backed vole establish territories against other females and the home range of a breeding male usually covers territories of several breeding females (Kawata, 1985; Saitoh, 1989b), and most female offspring stay at their birth place and breed on the territory neighboring their mother's, whereas a male offspring usually disperses from its birth place (Saitoh, unpublished). Although the offspring of the early and the late reproducing females in the present results were not equivalent to juvenile recruitment in the spring and the fall reported by McShea and Madison (1986), it is likely that the late reproducing females bred in the environment where it would have been difficult for their daughters to acquire their territories. There are many reports in mammals about biased sex ratio of offspring (CluttonBrock and Iason, 1986). The biased sex ratio is basically interpreted by the general form of the Trivers and Willard's theory (1973) that mothers invest more in the sex which would most successfully breed, but this is controversial in detail (Clutton-Brock and Iason, 1986; Hrdy, 1987). The present study provides interesting facts to that discussion. I cannot, however, fruitfully discuss the biased sex ratio of offspring here, because I could not evaluate the lifetime reproductive success of sons and daughters. In conclusion, the present results support the Reproductive Suppression Model by Wasser and Barash (1983), because lifetime reproductive success and the life span of offspring were not different between the early and the late reproducing females; the losses observed in the females which delayed the start of reproduction were compensated for by their longer life span. Although the late reproducing females produced more sons, discussing the adaptive significance of this biased sex ratio of offspring was beyond the scope of this study. To test the adaptive significance of the biased sex ratio, it is necessary to manifest ,the relationship between mother-offspring, and to measure the lifetime reproductive success of sons and daughters. These researches are also important for testing the Reproductive Suppression Model more precisely, because the technique adopted in this study was simple. The mother-offspring relationship can be determined (not estimated) by using genetic techniques (e.g., DNA fingerprinting). More intensive trapping of juveniles is required for more accurate counting of the number of offspring; although trappability was very high in this study, some unmarked voles firstly appeared in adult or subadult size.

SUMMARY

A population of the grey red-backed vole, Clethrionomys rufocanus bedfordiae, was investigated on.a 1 ha control grid and a 1 ha grid on which the voles were fed within a 2.1 ha outdoor enclosure in Hokkaido, Japan by live trapping from 1984 to 1986, for testing the Reproductive Suppression Model of Wasser and Barash (1983)-females can

404 o p t i m i z e t h e i r l i f e t i m e r e p r o d u c t i v e success b y s u p p r e s s i n g r e p r o d u c t i o n w h e n f u t u r e c o n d i t i o n s for t h e s u r v i v a l o f o f f s p r i n g a r e l i k e l y to b e sufficiently b e t t e r t h a n p r e s e n t ones as to e x c e e d t h e costs o f t h e s u p p r e s s i o n itself.

A g e at the first p r e g n a n c y m o r e

v a r i e d in a h i g h e r d e n s i t y p o p u l a t i o n o n the e x p e r i m e n t a l g r i d a n d f e m a l e s c o u l d b e classified i n t o the e a r l y a n d t h e l a t e r e p r o d u c t i v e t y p e in two g e n e r a t i o n s (A: f e m a l e s b o r n f r o m F e b r u a r y to J u n e 1985).

1985; B: f e m a l e s b o r n f r o m S e p t e m b e r to N o v e m b e r

L i f e t i m e r e p r o d u c t i v e success (the n u m b e r o f p r e g n a n c i e s , the n u m b e r of suc-

cessful litters, a n d t h e n u m b e r o f offspring) w a s n o t different b e t w e e n the e a r l y a n d t h e late r e p r o d u c i n g females.

T h e late r e p r o d u c i n g f e m a l e s l i v e d for l o n g e r p e r i o d s t h a n

t h e e a r l y r e p r o d u c i n g f e m a l e s , so t h a t t h e loss b y d e l a y e d s t a r t o f r e p r o d u c t i o n w a s c o m p e n s a t e d for b y a l o n g e r life s p a n .

Life s p a n w a s n o t different b e t w e e n o f f s p r i n g o f

t h e e a r l y a n d the l a t e r e p r o d u c i n g females.

T h e s e facts s u p p o r t e d t h e R e p r o d u c t i v e

Suppression Model. ACKNOWLEDGMENTS: I thank M. Kawata, T. Inoue and Y. It6 for their valuable comments on drafts of the manuscript and K. Ota, H. Mori, H. Abe, and K. Ishigaki for their kind encouragement, and I am indebted also to the staff of the Tomakomai Experimental Forest of Hokkaido University for their kind assistance with my field work. This work was partly supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture, Japan (No. 61790226).

REFERENCES

Abe, H. (1968) Growth and development in two forms of Clethrionomys: 1. External characters, body weight, sexual maturity and behavior. Bull Hokkaido Forest Experiment Station 6: 69-89. (in Japanese with English summary) Bondrup-Nielsen, S. (1986) Investigation of spacing behaviour of Clethrionomysgapperi by experimentation. J. Anim. Ecol. 55: 269-280. Bronson, F. H. (1979) The reproductive biology of the house mouse. Q. Rev. Biol. 54: 265-299. Bronson, F. H. (1985) Mammalian reproduction: an ecological perspective. Biol. Reprod. 32: 1-26. Bronson, F. H. andA. Coquelin (1980) The modulation ofreproduction by priming pheromones in house mouse: speculations 'on adaptive function. 243-265. In D. Muller-Schwarze and R. M. Silversten (eds) Chemical Signals; Vertebratesand Aquatic Invertebrates. Plenum, New York. Brown, R. E. (1985) The rodent I: effects of odours on reproductive physiology (primer effects). 245344. In R. E? Brown and D. W. MacDonald (eds) Social Odours in Mammals, vol 1. Oxford University Press, Oxford. Bujalska, G. (1970) Reproduction stabilizing elements in an island populations of Clethrionomysglareolus (Schreber, 1780). Acta Theriol. 15: 381-412. Bujalska, G. (1973) The role of spacing behaviour among females in the regulation of reproduction in the bank vole. J. Reprod. Fert. Supple. 19: 461-472. Clutton-Brock, T. H. and G. R. Iason (1986) Sex ratio in mammals. Q. Rev. Biol. 61: 339-374. Gilbert, B. S., C.J. Krebs, D. Talarico, and D. B. Cichowski (1986) Do Clethrionomys rutilus females suppress maturation of juvenile females? J. Anim. Ecol. 55: 543-552. Haigh, G . R . (1987) Reproductive inhibition of female Peromyscus leucopus: female competition and behavioral regulation. Amer. Zool. 27: 867-878.

405 Heske, E . J . and R . J . Nelson (1984) Pregnancy interruption in Microtus ochrogaster! laboratory artifact or field phenomenon?

Biol. Reprod. 31: 97-103.

Hilborn, R., J. A. Redfield, and C. J. Krebs (1976) On the reliability of enumeration for mark and recapture census of voles. Can. J. Zool. 54: 1019-1024. Hrdy, S. B. (1987) Sex-biased parentalinvestment among primates and other mammals: a critical evaluation of the Trivers-Willard hypothesis. 97-147.

In J. Lancaster and R. Geddes (eds) Child Abuse and

Neglect: a Biosocial Perspective. Aldine, New York. Ires, R. A. (1985) The effect of pregnancy failure on the onset of reproduction in sympatric populations of

Clethrionomys rutilus and Clethrionomys rufocanus. Ann. Zool. Fenn. 38: 190-193. Evidence for pregnancy failure in young female Lemmus lemmus and Microtus oeconomus. Can. J. Zool. 62: 2568-2570.

Jensen, P. M. and T. O. Gustafsson (1984)

Kawata, M. (1985) Mating system and reproductive success in a spring population of the red-backed vole,

Clethrionomys rufocanus bedfordiae. Oikos 45: 181-190. Pregnancy failure and suppression by female-female interaction in enclosed populations of the red-backed vole, Clethrionomys rufocanus bedfordiae. Behav. Ecol. SociobioI. 20: 89-97. Mallory, F. F. and F. V. Clulow (1977) Evidence of pregnancy failure in the wild meadow vole, Microtus

Kawata, M. (1987)

pennsylvanicus. Can.J. Zool. 55: 1-17. Maynard Smith, J. (1982) The evolution of social behaviour-a classification of models. 29-44. In King's College Sociobiology Group (ed) Current Problems in Sociobiology. Cambridge University Press, Cambridge. McShea, W . J . and D. M. Madison (1986) Sex ratio shifts within litters of meadow voles (Microtus penn-

sylvanicus). Behav. Ecol. Sociobiol. 18: 431-436. Milligan, S. R. (1980)

Pheromones and rodent reproductive physiology.

Syrup. Zool. Soc. London 45:

251-275. Nakata, K. (1989)

Regulation of reproduction rate in a cyclic population of the red-backed vole,

Clethrionomys rufocanus bedfordiae. Res. Popul. Ecol. 31: 185-209. Richmond, M. E. and R. Stehn (1976) Olfaction and reproductive behavior in microtine rodents. 197217. In R . L. Doty (ed) Mammalian Olfaction, ReproductiveProcessesand Behavior. Academic Press, New York. Rogers,J. G . J r . and G. K. Beauchamp (1976) Some ecological implications of primer chemical stimuli in rodents. 181-197.

In R. L. Doty (ed) Mammalian Olfaction, Reproductive Processes and Behavior.

Academic Press, New York. Saitoh, T. (1981)

Control of female maturation in high density populations of the red-backed vole, Clethrionomys rufocanus bedfordiae. J. Anita. Ecol. 50: 79-87.

Saitoh, T. (1989a) Effects of added food on Some attributes of an enclosed vole population. J. Mamm. 70: 772-782. Saitoh, T. (1989b) Communal nesting and spatial structure in an early spring population of the grey redbacked vole, Clethrionomys rufocanus bedfordiae. J. Mamm. Soc. Japan. 14: 27-41. Trivers, R. L. and D. E. Willard (1973) Natural selection of parental ability to vary the sex ratio of offspring. Science 179: 90-92. Vandenbergh, J. G. (1987) Regulation of puberty and its consequences on population dynamics of mice. Amer. Zool. 27: 891-898. Wasser, S. K. and D. P. Barash (1983) Reproductive suppression among female mammals: implications for biomedicine and sexual selection theory. Q. Rev. Biol. 58: 513-538. Westlin, L. M. and E. Nyholm (1982)

Sterile matings initiate the breeding season in the bank vole,

406

Clethrionomys glareolus: a field and laboratory study. Can. J. Zool. 60: 387-391. Sociobiology, the New Synthesis. Harvard University Press, Cambridge.

Wilson, E. O. (1975)

N N

@

19849aD~g19869a~-~'N.~.b, Wasser & Barash (1983) 9 [ - ~ N l - e ~ ' ~ > J (~,, 9

o]',2.