- Email: [email protected]
Studies of pericopulatory pregnancy blockage and the gestation period in deer mice (Peromyscus maniculatus)
HORMONES
AND
BEHAVIOR
19, 164-173 (1985)
Studies of Pericopulatory Pregnancy Blockage and the Gestation Period in Deer Mice (Peromyscus maniculatus) DONALD Department
of Psychology,
A. DEWSBURY
University
of Florida,
Gainesville,
Florida,
32611
It has previously been shown that female deer mice in cycling estrus that copulate with or are exposed to more than one male at the time of mating have a lower probability of delivering a litter than when mating with and being exposed to just one male. In the present research the effect of post-mating exposure to the bedding of a strange versus a familiar male was studied in females in postpartum estrus. In four experiments females in postpartum estrus were protected from pericopulatory pregnancy blockage, presumably by suckling-induced elevated prolactin levels. Varying female experience, number of ejaculations, and the presence of suckling pups had no effect on this protection. Thus, both the pericopulatory block and the Bruce effect (a more delayed type of pregnancy blockage) occur in cycling, but not suckling, females. This suggests that the two may have a common mechanism and leads to a new perspective in the search for adaptive significance. In addition, the subsequent gestation periods were prolonged in females nursing young; the length was affected by the number of pups weaned from the previous litter and the number of pups being carried. Q 1985 Academic
Press. Inc.
Female deer mice (Peromyscus maniculatus) that copulate with, or are exposed to, more than one male at the time of mating have a lower probability of successful pregnancy than do females mating an equivalent amount with and being exposed to just one male (Dewsbury, 1982). The effect was even obtained in response to exposure to a strange male’s bedding. A similar effect has recently been reported in Djungarian hamsters by Wynne-Edwards and Lisk (1984). Such pregnancy blockages bear at least superficial resemblance to the well-known “Bruce effect” (Bruce, 1959), in that in both cases exposure of the female to a second male interferes with pregnancy. However, the pericopulatory blockage differs from traditional demonstrations of the Bruce effect in at least three important ways: (1) exposure to the novel male occurs shortly after copulation rather than several days later, (2) the duration of exposure to the first male is quite brief, and (3) the duration of exposure to the second male is relatively brief. The existence of the pericopulatory preg164
0018-506x/85 $1.50 Copyright All rights
(0 1985 by Academic Press. Inc. of reproduction in any form reserved.
PERICOPULATORYPREGNANCYBLOCK
165
nancy block has important implications for reproduction in deer mice. However, the relationship to the Bruce effect is not yet clear; this research was undertaken in an effort to clarify that relationship. The Bruce effect has been repeatedly demonstrated in deer mice (e.g., Bronson and Eleftheriou, 1963;Eleftheriou, Bronson, and Zarrow, 1%2). Typically, a female is exposed to the first male for a period of days before mating and a “strange” male is introduced 24-48 hr after detection of either a copulatory plug or sperm in a vaginal smear. Arguing in favor of the possibility of a single mechanism underlying the two effects are results from other species demonstrating that the temporal patterning of events can be markedly altered and a Bruce effect obtained (e.g., Chipman, Holt, and Fox, 1966; Lott and Hopwood, 1972;Keverne and de la Riva, 1982;Milligan, 1979).That the two effects appear dissociable is suggested by data from prairie voles revealing a Bruce effect (Stehn and Jannett, 1981;Stehn and Richmond, 1975)but apparentlyno pericopulatoryblockage (Dewsbury and Baumgardner, 1981;Gray, Zerylnick, Davis, and Dewsbury, 1974).Should the two phenomena be found to reflect a single mechanism, the adaptive significance of the Bruce effect, a subject of much recent debate (e.g., Dawkins, 1976; Schwagmeyer, 1979) might be viewed in different perspective. The question of commonality of mechanism can be approached by manipulatingvariablespreviously shown to affect one of the two phenomena to determine whether they atfect the second. In a variety of species it has been found that lactating females mated in postpartum estrus show no Bruce effect (e.g., Bruce and Parkes, 1961;Mallory and Clulow, 1977; Rohrbach, 1982).The failure to obtain the Bruce effect appearsattributable to high prolactin levels (Bruce and Pa&es, 1960;Dominic, 1966).Prolactin administration also eliminates the Bruce effect in deer mice (Bronson, Eleftheriou, and Dezell, 1%9). In the present research the pericopulatory effect was studied in females in postpartum estrus in order to determine whether the postpartum hormonal milieu would protect females from such blockages as it does from the Bruce effect. In the course of such studies, data are generated relating to litter sizes and gestation periods that are relevant to questions of the determinants of gestation period in deer mice. These include the effect of lactation, the number of pups suckled, and the number of pups being carried in the uterus during gestation. Such effects are examined in supplementary analyses. METHODS
Subjects. The subjects were deer mice, Peromyscus maniculatus bairdi, from the colony maintained at the University of Florida (see Dewsbury, 1982).All were between 85 and 180days of age at the beginning of testing and were without prior copulatory or breeding experience. Sixteen “ex-
166
DONALD
A. DEWSBIJRY
perimental” males were used in tests and an additional 9 males were used as breeding males to impregnate females to be studied in postpartum estrus. Females were run as two squads. Twenty-eight of 31 females in the first squad completed Experiment I ; 26 of these females completed Experiment 2. Nineteen of 20 females in the second squad completed Experiment 3; 17 of these females completed Experiment 4. Mice were housed individually in a windowless, air-conditioned room on a reversed 16L:8D photoperiod. Males were housed in clear polycarbonate cages measuring 48 x 27 x 13 cm; those of females measured 29 x 19 x 13 cm. Wood shavings were provided as bedding and water and Purina laboratory rodent chow were available at all times. Apparatus. Tests were conducted in the males’ home cages. Behavioral patterns were recorded on an Esterline-Angus operations recorder. Procedure. Four experiments were conducted; two with each squad of females. These experiments were patterned after Experiment 3 of Dewsbury (1982), in which a 2-hr period of exposure to the bedding of a strange male immediately after mating yielded a pregnancy blockage. In each experiment in the present study each female received two tests, one with a familiar male treatment and one with a strange male treatment, in a within-subjects design. For each female the study began with a 4-day placement in the home cage of a breeder male so that she might be impregnated. Tests were conducted with females in postpartum estrus resulting from either such placement or from a pregnancy initiated in an experimental test. The cages of mated females were checked for litters each morning. When pups were found they were immediately counted and kept or sacrificed according to the appropriate experimental treatment. On the afternoon of the day on which a litter was found the female was tested by being placed in the home cage of an experimental male for a predetermined number of ejaculatory series. At that point the female was removed from the male’s cage and either replaced in the cage (familiar male condition) or placed in the cage of a different experimental male (strange male condition), where she remained, without the male, for 2 hr. Females were then returned to their home cages and checked for litters when parturition was expected. The females serving in Experiments 1 and 2 were divided into two groups; in one the litter was kept and in the other it was sacrificed. This was done in an effort to utilize procedures found to produce differential effects in studies of the Bruce effect (e.g., Bruce and Parkes, l%l). In Experiment 1 the 16 females whose litters were kept mated for three ejaculations with the experimental male. Each was mated twice; once in the familiar-male condition and once in the strange-male condition; half of the females were tested first in the strange-male condition and half first in the familiar-male condition. The 12 females whose litters were sacrificed were tested identically. Experiment 2 was conducted to
PERICOPULATORY
PREGNANCY
BLOCK
167
test the possibility that the three ejaculations allowed in Experiment 1, a level identical to that used by Dewsbury (1982), might have provided sufficient stimulation to override any pregnancy blockage effect. Female deer mice in postpartum estrus have a lower threshold for pregnancy initiation than they do in cycling estrus (Dewsbury, 1979b). Procedures were identical to those of Experiment I, except that just one ejaculatory series was permitted. The experiment was completed by 15 females whose litters were kept and 11 females whose litters were sacrificed. Because Experiment 2 yielded the suggestion of an effect, the procedures were replicated in Experiment 3 with a new group of females. Nineteen females completed the experiment; all litters were kept. Experiment 4 was conducted in an effort to test the possibility that parity and/or prior experience affects the response to a strange male. In Experiment 4, Experiment 3 was replicated using the same females-who now had the previous experience gained by serving in Experiment 3. Seventeen females completed Experiment 4. Measures. The copulatory pattern of deer mice is a multiple-intromission, multiple-ejaculation pattern similar to that in laboratory rats. The standard measures of copulatory behavior-mount latency, intromission latency, ejaculation latency, intromission frequency, mount frequency, mean interintromission interval, and postejaculatory interval-(Clemens, 1969; Dewsbury, 1979a) were determined for each experimental test. As available and appropriate, the following measureswere also obtained: number of pups born in the pregnancy generating the postpartum estrus for the test (i.e., first litter), number of pups weaned from that litter, number of pups in the litter resulting from the experimental test (i.e., second litter), and gestation period for the litter resulting from the experimental test. RESULTS Pregnancy
Blockage
The results relating to the occurrence of pregnancy and pregnancy blockage are presented in Table 1. There was no suggestion of blockage in Experiment 1; litters were delivered after 55 of the 56 tests. A suggestion of an effect was generated in Experiment 2 among females permitted to keep their litters. Litters were delivered by 87% of the females when exposed to the bedding of familiar males but just 60% when exposed to strange-male bedding (P = 0.11). It was this result that prompted the completion of Experiments 3 and 4. There was no such difference among females whose litters were sacrificed. There was no significant treatment effect in Experiment 3; the suggestion of such an effect is in the direction opposite that of Experiment 2. Similarly, when, in Experiment 4, the procedures were repeated on females
168
DONALD A. DEWSBURY
TABLE I Results of Four Experiments on Pregnancy Blockage after Multimale Exposure at the Time of Copulation in Female Deer Mice in Postpartum Esters Percentage delivering litters Experiment
Pups
No. ejacs.
Prior exper.
Familiar male
Strange male
P( l-tailed binomial)
1
Keep Sac. Keep Sac. Keep Keep
3 3 I I 1 I
No No Yes Yes No Yes
100(16/16) 100(12/12) 87(13/15) 82( 9/11) 74(14/19) 70(12/17)
94(15/16) 100(12/12) 60( 9/15) 73( 8/11) 95(18/19) 82(14/17)
0.11 0.50 0.1 I -
2 3 4
that had now gained the experience of Experiment 3, there was no effect of bedding type on pregnancy. In order to ensure lack of bias with respect to the amount of copulatory stimulation received, paired t tests were conducted comparing measures of copulatory behavior in the first ejaculatory series as a function of treatment condition (familiar versus strange male cage). There were six blocks of data, corresponding to the six lines in Table 1, and six measures per block. Only one of the 36 t tests was significant at the 0.05 level. A series of unpaired t tests was conducted in order to determine whether the copulatory behavior differed in tests in which females became pregnant as compared to those after which they delivered no litter. The only significant effects were found in Experiment 4. In Experiment 4, females becoming pregnant copulated in tests with males having shorter mount latencies (315.3 vs 507.1, t = 2.88, P = O.OI), intromission latencies (343.9 vs 510.9, t = 2.27, P < 0.05), and mean inter-intromission intervals (45.7 vs 72.6, t = 2.52, P < 0.02) than males in tests following which females bore no litter. Gestation Period
The effects of keeping versus sacrificing a litter on gestation period can be examined in tests with exposure to familiar and strange male bedding in Experiments 1 and 2. These results are presented in Table 2. It is clear that suckling prolongs the gestation period. In each of the four cases the gestation periods are sign&zantly longer for females keeping their litters than for females whose litters were sacrificed. The range of mean prolongation of the gestations periods was 1.5-2.9 days in different data sets. Multiple linear regression analyses were used to examine the relationship among three variables-the number of pups born in the first litter, the
PERICOPULATORY
169
PREGNANCY BLOCK
TABLE 2 Gestation Period as a Function of the Presence of Suckling Young in Experiments 1 and 2 Gestation (days) Experiment
Male
Keep litter
Sac. litter
r
df
P
I
Familiar Strange Familiar Strange
26.1 26.3 24.9 25.6
23.6 23.4 23.4 23.8
5.65 4.32 3.04 2.10
19 16 20 9
0.001 0.001 0.01 0.05
2
number of pups weaned from the first litter, and the number of pups in the second litter-and the length of the gestation period of the second litter. Various analyses were completed using various sets and subsets of the data from this study. The simplest of these analyses is consistent with the general trend of the more complex ones. A single analysis was done in which the data from all tests in all four experiments were included. There were 111 cases in which data on all four variables were available. Although multiple entries from individual females are included, the results are substantially the same as when many smaller analyses are conducted. There was no significant effect of the number of pups born in the first litter on the gestation period of the next litter. The most important determinant of gestation period was the number of pups weaned from the first litter; the standard regression coefficient was 0.59 (t = 6.00, P < 0.0001). There was also a significant effect of the number of pups born in the second litter; the standard regression coefficient was -0.32 (t = 3.64, P < 0.001). Thus, whereas gestation period was increased when there were more young suckled, long gestation periods were associated with relatively small second litters. The multiple r was 0.52. DISCUSSION The Bruce Effect and Pericopulatory
Blockage
The primary result from this research is the failure to obtain a pericopulatory pregnancy block when females were exposed to the bedding of strange males shortly after mating in postpartum es&us. Female deer mice in cycling estrus show a significantly lower probability of delivering a litter if they mate with or are exposed to more than one male, or the bedding from a second male, at the time of mating than if they mate with and are exposed to just one male (Dewsbury, 1982). Similarly, when female house mice in cycling estrus are exposed to a strange male or his odor l-2 days after mating, their pregnancies can be blocked [Bruce effect (Bruce, 1959)]. However, the Bruce effect is not obtained in lactating females (Bruce and Parkes, 1961). The parallel effects of lactation on
170
DONALD
A. DEWSBURY
the Bruce effect and the pericopulatory pregnancy block suggest that the two phenomena may be manifestations of a single process. This possibility is strengthened by results such as those of Milligan (1979), who found that “familiarity” for purposes of obtaining a Bruce effect could be achieved through contact with a male for just 1 hr at the time of mating. One respect in which the two effects differ, however, is that removal of suckling pups from postparturient females eliminates the Bruce, but not the pericopulatory, pregnancy block. Clearly, one must be cautious in reaching firm conclusions based on a single parametric manipulation; other variables will need to be found to produce parallel effects for the hypothesis of a unitary mechanism to be confirmed. At present, however, the weight of the evidence suggests a single mechanism. The Bruce effect has been reported in a number of rodent species and has been found in nonlactating females but not in lactating females in both meadow voles and Mongolian gerbils (Mallory and Clulow, 1977; Rohrbach, 1982). In addition, females of some species sometimes show a postimplantation pregnancy block in reaction to a strange male (Kenney, Evans, and Dewsbury, 1977; Stehn and Richmond, 1975).In both montane voles and prairie voles, however, the postimplantation pregnancy block is not reduced by concurrent lactation (Stehn and Jannett, 1981). This suggests that the postimplantation pregnancy block may be mediated by a mechanism different from that underlying the pericopulatory block and the Bruce effect. Endocrine
Mechanisms
Milligan (1980) reviewed evidence regarding the endocrine basis of the Bruce effect and noted that it is poorly understood. The effect reflects a failure of implantation that is due in turn to a failure of luteal function. Milligan suggested that the Bruce effect may be caused by interference with a “luteotrophic mnemonic” that is normally triggered by stimulation from copulation and is necessary for successful pregnancy in many species of rodents. The endocrine basis for the protection afforded lactating females appears attributable to high prolactin levels; administration of prolactin to nonlactating female house mice or deer mice mimics the lactation-generated protection from blockage (Bronson et al., 1969; Bruce and Parkes, 1960; Dominic, 1966). However, as Milligan (1980) noted, this does not necessarily mean that the effect of strange males in the Bruce effect is normally mediated by prolactin. In studies of the Bruce effect protective effects of postpartum estrus are lost if suckling is prevented by removal of the young soon after birth (e.g., Bruce and Parkes, 1961). In the present study, however, removal of pups approximately 8 hr before mating produced no increased incidence of blockage. Although the decrease in serum prolactin levels in rats following termination of suckling can be rather rapid (Amenomori, Chen,
PERICOPULATORY
PREGNANCY
BLOCK
171
and Meites, 1970), it is likely that prolactin levels in deer mice declined too slowly to reduce the protection from blockage. Thus, this difference between the Bruce effect and the pericopulatory block appears more a function of the time at which the second male is introduced in studies of the two effects (l-2 days vs immediately after mating) than an indication of a different mechanism. In studies of the Bruce effect there is a period of 1-2 days after pup removal for prolactin levels to decrease; in studies of pericopulatory blockage there are only a few hours. Gestation Periods Increased gestation periods in lactating females have been reported in a variety of rodents, such as house mice (Bruce and East, 1956), collared lemmings (Mallory and Brooks, 1978), and bank voles (Gustafsson, Andersson, and Westlin, 1980). The effect was reported in deer mice by Svihla (1932) but its ubiquity was questioned by Rood (1966). It is clear from the present data that lactation produces a reliable increase in gestation period. In various rodent species the length of the lactational delay has been found correlated with the number of young being suckled (e.g., Gustafsson et al., 1980; Norris and Adams, 1977, 1979). This appears not to be the case in cotton rats, however (Meyer and Meyer, 1944). In deer mice the gestation period is related to the weight of the previous litter (Myers and Master, 1983). In the present study gestation periods were clearly affected by the number of young suckled; the number of young weaned was a major determinant of gestation period. There are interesting species differences in the effect of the size of the litter in utero on gestation length. In deer mice, Svihla (1932) found no effect in nonlactating females and Myers and Master (1983) found none in lactating females. Norris and Adams (1977, 1979) found long gestation periods to be associated with small litter sizes in both laboratory rats and Mongolian gerbils. The relationship is less clear in house mice (Bruce and East, 1956; Norris and Adams, 1981).The inverse relationship in the present data is consistent with that reported in rats and gerbils. Adaptive SigniJicance The question of the adaptive significance of the Bruce effect has received much attention (e.g., Bronson and Coquelin, 1980; Dawkins, 1976; Schwagmeyer, 1979). It has generally been viewed as functioning in nature when a strange male displaces a resident male in a stable social unit; such occurrences may be rare in nature. Such an interpretation of the Bruce effect is forced by the limited range of procedures used in classical studies of the Bruce effect. A new perspective is provided by studies of the pericopulatory pregnancy block. In nature, such an effect would require sequential multimale mating
172
DONALD A. DEWSBURY
rather than disruption of a stable social unit. Multimale mating is known to occur among deer mice in nature (Birdsall and Nash, 1973). Under natural conditions, one or both of the males may, or may not, be familiar to the female and thus not “strange” in the sensemanipulated in labotatory studies (e.g., Healey, 1%7). If the Bruce effect and pericopulatory pregnancy block reflect the same mechanism, the conditions for occurrence in the field clearly exist and the search for adaptive significance can occur with an altered perspective. ACKNOWLEDGMENTS This research was supported by Grant BNS82-00689from the National Science Foundation. I thank Frank Bronson and Michael Leon for suggesting this line of research.
REFERENCES Amenomori, Y., Chen, C. L., and Meites, J. (1970). Serum prolactin levels in rats during different reproductive states. Endocrinology 6, 506-510. Birdsall, D. A., and Nash, D. (1973). Occurrence of successful multiple insemination of females in natural populations of deer mice (Peromyscus manicularus). Evolution 27, 106-I 10. Bronson, F. H., and Coquelin, A. (1980). The modulation of reproduction by priming pheromones in house mice: Speculations on adaptive function. In D. Muller-Schwarze and R. M. Silverstein (Eds.), pp. 243-265. Chemical Signals: Vertebrates and Aquatic Invertebrates. Plenum, New York. Bronson, F. H., and Eleftheriou, B. E. (1963). Influence of strange males on implantation in the deermouse. Gen. Camp. Endocrinol. 3, 515-518. Bronson, F. H., Eleftheriou, B. E., and Dezell, H. E. (1969). Strange male pregnancy block in deermice: Prolactin and adrenocortical hormones. Eiol. Reprod. 1, 302-306. Bruce, H. M. (1959). An exteroceptive block to pregnancy in the mouse. Nature (London) 184,390.
Bruce, H. M., and East, J. (1956). Number and viability of young from pregnancies concurrent with lactation in the mouse. /. Endocrinol. 14, 19-27. Bruce, H. M., and Parkes, A. S. (1960). Hormonal factors in exteroceptive block to 20, xxix-xxx. pregnancy in mice. J. Endocrinol. Bruce, H. M.. and Parkes, A. S. (l%l). The effect of concurrent lactation on the olfactory block to pregnancy in the mouse. J. Endocrinol. 22, vi-vii. Chipman, R. K., Holt, J. A.. and Fox, K. A. (1966). Pregnancy failure in laboratory mice after multiple short-term exposure to strange males. Narure (London) 210, 653. Clemens, L. G. (1%9). Experimental analysis of sexual behaviour of the deermouse, Peromyscus
maniculatus
gambeli.
Behaviour
34, 267-285.
Dawkins, R. (1976). The Selfish Gene. Oxford Univ. Press, Oxford. manicularus): Dewsbury, D. A. (1979a). Copulatory behavior of deer mice (Peromyscus I. Normative data, subspecific differences. and effects of cross-fostering. J. Camp. Physiol. Psychol. 93, 151-160. Dewsbury, D. A. (1979b). Copulatory behavior of deer mice (Peromyscus maniculatus): III. Effects on pregnancy initiation. J. Comp. Physiol. Psycho/. 93, 178-188. Dewsbury, D. A. (1982). Pregnancy blockage following multiple-male copulation or exposure at the time of mating in deer mice, Peromyscus maniculatus. Behav. Ecol. Sociobiol. 11, 37-42.
Dewsbury, D. A., and Baumgardner, D. J. (1981). Studies of sperm competition in two species of muroid rodents. Behav. Ecol. Sociobiol. 9, 121-133.
PERICOPULATORY
PREGNANCY BLOCK
173
Dominic, C. J. (1966). Observations on the reproductive pheromones of mice.: II. Neuroendocrine mechanisms involved in the olfactory block to pregnancy. J. Reprod. Fertil. 11, 415-421.
Eleftheriou, B. E., Bronson, F. H., and Zarrow, M. X. (1962). Interaction of olfactory and other environmental stimuli on implantation in the deer mouse. Science (Washington. D.C.)
137, 764.
Gray. G. D., Zerylnick, M., Davis, H. N., and Dewsbury, D. A. (1974). Effects of variations in male copulatory behavior on ovulation and implantation in prairie voles, Microfus ochrogaster.
Harm. Behav. 5, 389-3%.
Gustafsson, T., Andersson, B., and Westlin, L. (1980). Reproduction in a laboratory colony of bank vole, Cfethrionomys glare&s. Canod. J. Zoo/. 58, 1016-1021. Healey, M. C. (1%7). Aggression and self-regulation of population size in deermice. Ecology 48, 377-392. Kenney, A. McM., Evans, R. L., and Dewsbury, D. A. (1977). Postimplantation pregnancy disruption in Microtus ochrogaster, M. pennsylvanicus, and Peromyscus maniculatus. J. Reprod.
Fertil.
49, 365-367.
Keveme, E. B., and de la Riva, C. (1982). Pheromones in mice: Reciprocal interaction between the nose and the brain. Nature (London) 296, 148-150. Lott, D. F., and Hopwood, J. H. (1972). Olfactory pregnancy-block in mice (Mus musculus): An unusual response acquisition paradigm. Anim. Eehnv. 20, 263-267. Mallory, F. F., and Brooks, R. J. (1978). Infanticide and other reproductive strategies in the collared lemming, Dicrosfonyx groenlandicus. Nature (London) 273, 144-146. Mallory, F. F., and Clulow, F. V. (1977). Evidence of pregnancy failure in the wild meadow vole, Microrus pennsylvanicus. Canod. J. Zool. 55, I-17. Meyer, B. J., and Meyer, R. K. (1944). Growth and reproduction of the cotton rat, Sigmodon hispidus hispidus, under laboratory conditions. 3. Mammal. 25, 107-129. Milligan. S. R. (1979). Pregnancy blockage and memory of the stud male in the vole (Microtus
agrestis).
1. Reprod.
Fertil.
57, 223-225.
Milligan, S. R. (1980). Pheromones and rodent reproductive physiology. Symp. Zoo/. Sot. London 45, 251-275. Myers, P., and Master, L. L. (1983). Reproduction by Peromyscus mnniculatus: Size and compromise. J. Mammal. 64, l-18. Norris, M. L., and Adams, C. E. (1977). Delayed implantation in the Mongolian gerbil, Meriones
unguicularus.
27, 486-487.
Norris, M. L., and Adams, C. E. (1979). Mating post partum, concurrent lactation and reproduction in the laboratory rat. Lab. Anim. 13, 167-170. Norris, M. L., and Adams, C. E. (1981). Concurrent lactation and reproductive performance in CFLP mice mated post partum. Lab. Anim. 15, 273-275. Rohrbach, C. (1982). Investigation of the Bruce effect in the Mongolian gerbil (Meriones unguiculatus). J. Reprod. Fertil. 65, 41 l-417. Rood, J. P. (1966). Observations on the reproduction of Peromyscus in captivity. Amer. Midl.
Nat. 76, 4%-503.
Schwagmeyer, P. E. (1979). The Bruce effect: An evaluation of male/female advantages. Amer. Nat.
114, 932-938.
Stehn, R. A., and Jannett, F. J. (1981). Male-induced abortion in various microtine rodents. J. Mammal.
62, 369-372.
Stehn, R. A., and Richmond, M. E. (1975). Male-induced pregnancy termination in the prairie vole, Microtus ochrogasrer. Science (Washington, D.C.) 187, I21 I-1213. Svihla, A. (1932). A comparative life history study of the mice of the genus Peromyscus. Misc. Publ. Mus. Zool. Univ. Michigan 24, l-39. Wynne-Edwards, K. E., and Lisk. R. D. (1984). Djungarian hamsters fail to conceive in the presence of multiple males. Anim. Behov. 32, 626-628.
AND
BEHAVIOR
19, 164-173 (1985)
Studies of Pericopulatory Pregnancy Blockage and the Gestation Period in Deer Mice (Peromyscus maniculatus) DONALD Department
of Psychology,
A. DEWSBURY
University
of Florida,
Gainesville,
Florida,
32611
It has previously been shown that female deer mice in cycling estrus that copulate with or are exposed to more than one male at the time of mating have a lower probability of delivering a litter than when mating with and being exposed to just one male. In the present research the effect of post-mating exposure to the bedding of a strange versus a familiar male was studied in females in postpartum estrus. In four experiments females in postpartum estrus were protected from pericopulatory pregnancy blockage, presumably by suckling-induced elevated prolactin levels. Varying female experience, number of ejaculations, and the presence of suckling pups had no effect on this protection. Thus, both the pericopulatory block and the Bruce effect (a more delayed type of pregnancy blockage) occur in cycling, but not suckling, females. This suggests that the two may have a common mechanism and leads to a new perspective in the search for adaptive significance. In addition, the subsequent gestation periods were prolonged in females nursing young; the length was affected by the number of pups weaned from the previous litter and the number of pups being carried. Q 1985 Academic
Press. Inc.
Female deer mice (Peromyscus maniculatus) that copulate with, or are exposed to, more than one male at the time of mating have a lower probability of successful pregnancy than do females mating an equivalent amount with and being exposed to just one male (Dewsbury, 1982). The effect was even obtained in response to exposure to a strange male’s bedding. A similar effect has recently been reported in Djungarian hamsters by Wynne-Edwards and Lisk (1984). Such pregnancy blockages bear at least superficial resemblance to the well-known “Bruce effect” (Bruce, 1959), in that in both cases exposure of the female to a second male interferes with pregnancy. However, the pericopulatory blockage differs from traditional demonstrations of the Bruce effect in at least three important ways: (1) exposure to the novel male occurs shortly after copulation rather than several days later, (2) the duration of exposure to the first male is quite brief, and (3) the duration of exposure to the second male is relatively brief. The existence of the pericopulatory preg164
0018-506x/85 $1.50 Copyright All rights
(0 1985 by Academic Press. Inc. of reproduction in any form reserved.
PERICOPULATORYPREGNANCYBLOCK
165
nancy block has important implications for reproduction in deer mice. However, the relationship to the Bruce effect is not yet clear; this research was undertaken in an effort to clarify that relationship. The Bruce effect has been repeatedly demonstrated in deer mice (e.g., Bronson and Eleftheriou, 1963;Eleftheriou, Bronson, and Zarrow, 1%2). Typically, a female is exposed to the first male for a period of days before mating and a “strange” male is introduced 24-48 hr after detection of either a copulatory plug or sperm in a vaginal smear. Arguing in favor of the possibility of a single mechanism underlying the two effects are results from other species demonstrating that the temporal patterning of events can be markedly altered and a Bruce effect obtained (e.g., Chipman, Holt, and Fox, 1966; Lott and Hopwood, 1972;Keverne and de la Riva, 1982;Milligan, 1979).That the two effects appear dissociable is suggested by data from prairie voles revealing a Bruce effect (Stehn and Jannett, 1981;Stehn and Richmond, 1975)but apparentlyno pericopulatoryblockage (Dewsbury and Baumgardner, 1981;Gray, Zerylnick, Davis, and Dewsbury, 1974).Should the two phenomena be found to reflect a single mechanism, the adaptive significance of the Bruce effect, a subject of much recent debate (e.g., Dawkins, 1976; Schwagmeyer, 1979) might be viewed in different perspective. The question of commonality of mechanism can be approached by manipulatingvariablespreviously shown to affect one of the two phenomena to determine whether they atfect the second. In a variety of species it has been found that lactating females mated in postpartum estrus show no Bruce effect (e.g., Bruce and Parkes, 1961;Mallory and Clulow, 1977; Rohrbach, 1982).The failure to obtain the Bruce effect appearsattributable to high prolactin levels (Bruce and Pa&es, 1960;Dominic, 1966).Prolactin administration also eliminates the Bruce effect in deer mice (Bronson, Eleftheriou, and Dezell, 1%9). In the present research the pericopulatory effect was studied in females in postpartum estrus in order to determine whether the postpartum hormonal milieu would protect females from such blockages as it does from the Bruce effect. In the course of such studies, data are generated relating to litter sizes and gestation periods that are relevant to questions of the determinants of gestation period in deer mice. These include the effect of lactation, the number of pups suckled, and the number of pups being carried in the uterus during gestation. Such effects are examined in supplementary analyses. METHODS
Subjects. The subjects were deer mice, Peromyscus maniculatus bairdi, from the colony maintained at the University of Florida (see Dewsbury, 1982).All were between 85 and 180days of age at the beginning of testing and were without prior copulatory or breeding experience. Sixteen “ex-
166
DONALD
A. DEWSBIJRY
perimental” males were used in tests and an additional 9 males were used as breeding males to impregnate females to be studied in postpartum estrus. Females were run as two squads. Twenty-eight of 31 females in the first squad completed Experiment I ; 26 of these females completed Experiment 2. Nineteen of 20 females in the second squad completed Experiment 3; 17 of these females completed Experiment 4. Mice were housed individually in a windowless, air-conditioned room on a reversed 16L:8D photoperiod. Males were housed in clear polycarbonate cages measuring 48 x 27 x 13 cm; those of females measured 29 x 19 x 13 cm. Wood shavings were provided as bedding and water and Purina laboratory rodent chow were available at all times. Apparatus. Tests were conducted in the males’ home cages. Behavioral patterns were recorded on an Esterline-Angus operations recorder. Procedure. Four experiments were conducted; two with each squad of females. These experiments were patterned after Experiment 3 of Dewsbury (1982), in which a 2-hr period of exposure to the bedding of a strange male immediately after mating yielded a pregnancy blockage. In each experiment in the present study each female received two tests, one with a familiar male treatment and one with a strange male treatment, in a within-subjects design. For each female the study began with a 4-day placement in the home cage of a breeder male so that she might be impregnated. Tests were conducted with females in postpartum estrus resulting from either such placement or from a pregnancy initiated in an experimental test. The cages of mated females were checked for litters each morning. When pups were found they were immediately counted and kept or sacrificed according to the appropriate experimental treatment. On the afternoon of the day on which a litter was found the female was tested by being placed in the home cage of an experimental male for a predetermined number of ejaculatory series. At that point the female was removed from the male’s cage and either replaced in the cage (familiar male condition) or placed in the cage of a different experimental male (strange male condition), where she remained, without the male, for 2 hr. Females were then returned to their home cages and checked for litters when parturition was expected. The females serving in Experiments 1 and 2 were divided into two groups; in one the litter was kept and in the other it was sacrificed. This was done in an effort to utilize procedures found to produce differential effects in studies of the Bruce effect (e.g., Bruce and Parkes, l%l). In Experiment 1 the 16 females whose litters were kept mated for three ejaculations with the experimental male. Each was mated twice; once in the familiar-male condition and once in the strange-male condition; half of the females were tested first in the strange-male condition and half first in the familiar-male condition. The 12 females whose litters were sacrificed were tested identically. Experiment 2 was conducted to
PERICOPULATORY
PREGNANCY
BLOCK
167
test the possibility that the three ejaculations allowed in Experiment 1, a level identical to that used by Dewsbury (1982), might have provided sufficient stimulation to override any pregnancy blockage effect. Female deer mice in postpartum estrus have a lower threshold for pregnancy initiation than they do in cycling estrus (Dewsbury, 1979b). Procedures were identical to those of Experiment I, except that just one ejaculatory series was permitted. The experiment was completed by 15 females whose litters were kept and 11 females whose litters were sacrificed. Because Experiment 2 yielded the suggestion of an effect, the procedures were replicated in Experiment 3 with a new group of females. Nineteen females completed the experiment; all litters were kept. Experiment 4 was conducted in an effort to test the possibility that parity and/or prior experience affects the response to a strange male. In Experiment 4, Experiment 3 was replicated using the same females-who now had the previous experience gained by serving in Experiment 3. Seventeen females completed Experiment 4. Measures. The copulatory pattern of deer mice is a multiple-intromission, multiple-ejaculation pattern similar to that in laboratory rats. The standard measures of copulatory behavior-mount latency, intromission latency, ejaculation latency, intromission frequency, mount frequency, mean interintromission interval, and postejaculatory interval-(Clemens, 1969; Dewsbury, 1979a) were determined for each experimental test. As available and appropriate, the following measureswere also obtained: number of pups born in the pregnancy generating the postpartum estrus for the test (i.e., first litter), number of pups weaned from that litter, number of pups in the litter resulting from the experimental test (i.e., second litter), and gestation period for the litter resulting from the experimental test. RESULTS Pregnancy
Blockage
The results relating to the occurrence of pregnancy and pregnancy blockage are presented in Table 1. There was no suggestion of blockage in Experiment 1; litters were delivered after 55 of the 56 tests. A suggestion of an effect was generated in Experiment 2 among females permitted to keep their litters. Litters were delivered by 87% of the females when exposed to the bedding of familiar males but just 60% when exposed to strange-male bedding (P = 0.11). It was this result that prompted the completion of Experiments 3 and 4. There was no such difference among females whose litters were sacrificed. There was no significant treatment effect in Experiment 3; the suggestion of such an effect is in the direction opposite that of Experiment 2. Similarly, when, in Experiment 4, the procedures were repeated on females
168
DONALD A. DEWSBURY
TABLE I Results of Four Experiments on Pregnancy Blockage after Multimale Exposure at the Time of Copulation in Female Deer Mice in Postpartum Esters Percentage delivering litters Experiment
Pups
No. ejacs.
Prior exper.
Familiar male
Strange male
P( l-tailed binomial)
1
Keep Sac. Keep Sac. Keep Keep
3 3 I I 1 I
No No Yes Yes No Yes
100(16/16) 100(12/12) 87(13/15) 82( 9/11) 74(14/19) 70(12/17)
94(15/16) 100(12/12) 60( 9/15) 73( 8/11) 95(18/19) 82(14/17)
0.11 0.50 0.1 I -
2 3 4
that had now gained the experience of Experiment 3, there was no effect of bedding type on pregnancy. In order to ensure lack of bias with respect to the amount of copulatory stimulation received, paired t tests were conducted comparing measures of copulatory behavior in the first ejaculatory series as a function of treatment condition (familiar versus strange male cage). There were six blocks of data, corresponding to the six lines in Table 1, and six measures per block. Only one of the 36 t tests was significant at the 0.05 level. A series of unpaired t tests was conducted in order to determine whether the copulatory behavior differed in tests in which females became pregnant as compared to those after which they delivered no litter. The only significant effects were found in Experiment 4. In Experiment 4, females becoming pregnant copulated in tests with males having shorter mount latencies (315.3 vs 507.1, t = 2.88, P = O.OI), intromission latencies (343.9 vs 510.9, t = 2.27, P < 0.05), and mean inter-intromission intervals (45.7 vs 72.6, t = 2.52, P < 0.02) than males in tests following which females bore no litter. Gestation Period
The effects of keeping versus sacrificing a litter on gestation period can be examined in tests with exposure to familiar and strange male bedding in Experiments 1 and 2. These results are presented in Table 2. It is clear that suckling prolongs the gestation period. In each of the four cases the gestation periods are sign&zantly longer for females keeping their litters than for females whose litters were sacrificed. The range of mean prolongation of the gestations periods was 1.5-2.9 days in different data sets. Multiple linear regression analyses were used to examine the relationship among three variables-the number of pups born in the first litter, the
PERICOPULATORY
169
PREGNANCY BLOCK
TABLE 2 Gestation Period as a Function of the Presence of Suckling Young in Experiments 1 and 2 Gestation (days) Experiment
Male
Keep litter
Sac. litter
r
df
P
I
Familiar Strange Familiar Strange
26.1 26.3 24.9 25.6
23.6 23.4 23.4 23.8
5.65 4.32 3.04 2.10
19 16 20 9
0.001 0.001 0.01 0.05
2
number of pups weaned from the first litter, and the number of pups in the second litter-and the length of the gestation period of the second litter. Various analyses were completed using various sets and subsets of the data from this study. The simplest of these analyses is consistent with the general trend of the more complex ones. A single analysis was done in which the data from all tests in all four experiments were included. There were 111 cases in which data on all four variables were available. Although multiple entries from individual females are included, the results are substantially the same as when many smaller analyses are conducted. There was no significant effect of the number of pups born in the first litter on the gestation period of the next litter. The most important determinant of gestation period was the number of pups weaned from the first litter; the standard regression coefficient was 0.59 (t = 6.00, P < 0.0001). There was also a significant effect of the number of pups born in the second litter; the standard regression coefficient was -0.32 (t = 3.64, P < 0.001). Thus, whereas gestation period was increased when there were more young suckled, long gestation periods were associated with relatively small second litters. The multiple r was 0.52. DISCUSSION The Bruce Effect and Pericopulatory
Blockage
The primary result from this research is the failure to obtain a pericopulatory pregnancy block when females were exposed to the bedding of strange males shortly after mating in postpartum es&us. Female deer mice in cycling estrus show a significantly lower probability of delivering a litter if they mate with or are exposed to more than one male, or the bedding from a second male, at the time of mating than if they mate with and are exposed to just one male (Dewsbury, 1982). Similarly, when female house mice in cycling estrus are exposed to a strange male or his odor l-2 days after mating, their pregnancies can be blocked [Bruce effect (Bruce, 1959)]. However, the Bruce effect is not obtained in lactating females (Bruce and Parkes, 1961). The parallel effects of lactation on
170
DONALD
A. DEWSBURY
the Bruce effect and the pericopulatory pregnancy block suggest that the two phenomena may be manifestations of a single process. This possibility is strengthened by results such as those of Milligan (1979), who found that “familiarity” for purposes of obtaining a Bruce effect could be achieved through contact with a male for just 1 hr at the time of mating. One respect in which the two effects differ, however, is that removal of suckling pups from postparturient females eliminates the Bruce, but not the pericopulatory, pregnancy block. Clearly, one must be cautious in reaching firm conclusions based on a single parametric manipulation; other variables will need to be found to produce parallel effects for the hypothesis of a unitary mechanism to be confirmed. At present, however, the weight of the evidence suggests a single mechanism. The Bruce effect has been reported in a number of rodent species and has been found in nonlactating females but not in lactating females in both meadow voles and Mongolian gerbils (Mallory and Clulow, 1977; Rohrbach, 1982). In addition, females of some species sometimes show a postimplantation pregnancy block in reaction to a strange male (Kenney, Evans, and Dewsbury, 1977; Stehn and Richmond, 1975).In both montane voles and prairie voles, however, the postimplantation pregnancy block is not reduced by concurrent lactation (Stehn and Jannett, 1981). This suggests that the postimplantation pregnancy block may be mediated by a mechanism different from that underlying the pericopulatory block and the Bruce effect. Endocrine
Mechanisms
Milligan (1980) reviewed evidence regarding the endocrine basis of the Bruce effect and noted that it is poorly understood. The effect reflects a failure of implantation that is due in turn to a failure of luteal function. Milligan suggested that the Bruce effect may be caused by interference with a “luteotrophic mnemonic” that is normally triggered by stimulation from copulation and is necessary for successful pregnancy in many species of rodents. The endocrine basis for the protection afforded lactating females appears attributable to high prolactin levels; administration of prolactin to nonlactating female house mice or deer mice mimics the lactation-generated protection from blockage (Bronson et al., 1969; Bruce and Parkes, 1960; Dominic, 1966). However, as Milligan (1980) noted, this does not necessarily mean that the effect of strange males in the Bruce effect is normally mediated by prolactin. In studies of the Bruce effect protective effects of postpartum estrus are lost if suckling is prevented by removal of the young soon after birth (e.g., Bruce and Parkes, 1961). In the present study, however, removal of pups approximately 8 hr before mating produced no increased incidence of blockage. Although the decrease in serum prolactin levels in rats following termination of suckling can be rather rapid (Amenomori, Chen,
PERICOPULATORY
PREGNANCY
BLOCK
171
and Meites, 1970), it is likely that prolactin levels in deer mice declined too slowly to reduce the protection from blockage. Thus, this difference between the Bruce effect and the pericopulatory block appears more a function of the time at which the second male is introduced in studies of the two effects (l-2 days vs immediately after mating) than an indication of a different mechanism. In studies of the Bruce effect there is a period of 1-2 days after pup removal for prolactin levels to decrease; in studies of pericopulatory blockage there are only a few hours. Gestation Periods Increased gestation periods in lactating females have been reported in a variety of rodents, such as house mice (Bruce and East, 1956), collared lemmings (Mallory and Brooks, 1978), and bank voles (Gustafsson, Andersson, and Westlin, 1980). The effect was reported in deer mice by Svihla (1932) but its ubiquity was questioned by Rood (1966). It is clear from the present data that lactation produces a reliable increase in gestation period. In various rodent species the length of the lactational delay has been found correlated with the number of young being suckled (e.g., Gustafsson et al., 1980; Norris and Adams, 1977, 1979). This appears not to be the case in cotton rats, however (Meyer and Meyer, 1944). In deer mice the gestation period is related to the weight of the previous litter (Myers and Master, 1983). In the present study gestation periods were clearly affected by the number of young suckled; the number of young weaned was a major determinant of gestation period. There are interesting species differences in the effect of the size of the litter in utero on gestation length. In deer mice, Svihla (1932) found no effect in nonlactating females and Myers and Master (1983) found none in lactating females. Norris and Adams (1977, 1979) found long gestation periods to be associated with small litter sizes in both laboratory rats and Mongolian gerbils. The relationship is less clear in house mice (Bruce and East, 1956; Norris and Adams, 1981).The inverse relationship in the present data is consistent with that reported in rats and gerbils. Adaptive SigniJicance The question of the adaptive significance of the Bruce effect has received much attention (e.g., Bronson and Coquelin, 1980; Dawkins, 1976; Schwagmeyer, 1979). It has generally been viewed as functioning in nature when a strange male displaces a resident male in a stable social unit; such occurrences may be rare in nature. Such an interpretation of the Bruce effect is forced by the limited range of procedures used in classical studies of the Bruce effect. A new perspective is provided by studies of the pericopulatory pregnancy block. In nature, such an effect would require sequential multimale mating
172
DONALD A. DEWSBURY
rather than disruption of a stable social unit. Multimale mating is known to occur among deer mice in nature (Birdsall and Nash, 1973). Under natural conditions, one or both of the males may, or may not, be familiar to the female and thus not “strange” in the sensemanipulated in labotatory studies (e.g., Healey, 1%7). If the Bruce effect and pericopulatory pregnancy block reflect the same mechanism, the conditions for occurrence in the field clearly exist and the search for adaptive significance can occur with an altered perspective. ACKNOWLEDGMENTS This research was supported by Grant BNS82-00689from the National Science Foundation. I thank Frank Bronson and Michael Leon for suggesting this line of research.
REFERENCES Amenomori, Y., Chen, C. L., and Meites, J. (1970). Serum prolactin levels in rats during different reproductive states. Endocrinology 6, 506-510. Birdsall, D. A., and Nash, D. (1973). Occurrence of successful multiple insemination of females in natural populations of deer mice (Peromyscus manicularus). Evolution 27, 106-I 10. Bronson, F. H., and Coquelin, A. (1980). The modulation of reproduction by priming pheromones in house mice: Speculations on adaptive function. In D. Muller-Schwarze and R. M. Silverstein (Eds.), pp. 243-265. Chemical Signals: Vertebrates and Aquatic Invertebrates. Plenum, New York. Bronson, F. H., and Eleftheriou, B. E. (1963). Influence of strange males on implantation in the deermouse. Gen. Camp. Endocrinol. 3, 515-518. Bronson, F. H., Eleftheriou, B. E., and Dezell, H. E. (1969). Strange male pregnancy block in deermice: Prolactin and adrenocortical hormones. Eiol. Reprod. 1, 302-306. Bruce, H. M. (1959). An exteroceptive block to pregnancy in the mouse. Nature (London) 184,390.
Bruce, H. M., and East, J. (1956). Number and viability of young from pregnancies concurrent with lactation in the mouse. /. Endocrinol. 14, 19-27. Bruce, H. M., and Parkes, A. S. (1960). Hormonal factors in exteroceptive block to 20, xxix-xxx. pregnancy in mice. J. Endocrinol. Bruce, H. M.. and Parkes, A. S. (l%l). The effect of concurrent lactation on the olfactory block to pregnancy in the mouse. J. Endocrinol. 22, vi-vii. Chipman, R. K., Holt, J. A.. and Fox, K. A. (1966). Pregnancy failure in laboratory mice after multiple short-term exposure to strange males. Narure (London) 210, 653. Clemens, L. G. (1%9). Experimental analysis of sexual behaviour of the deermouse, Peromyscus
maniculatus
gambeli.
Behaviour
34, 267-285.
Dawkins, R. (1976). The Selfish Gene. Oxford Univ. Press, Oxford. manicularus): Dewsbury, D. A. (1979a). Copulatory behavior of deer mice (Peromyscus I. Normative data, subspecific differences. and effects of cross-fostering. J. Camp. Physiol. Psychol. 93, 151-160. Dewsbury, D. A. (1979b). Copulatory behavior of deer mice (Peromyscus maniculatus): III. Effects on pregnancy initiation. J. Comp. Physiol. Psycho/. 93, 178-188. Dewsbury, D. A. (1982). Pregnancy blockage following multiple-male copulation or exposure at the time of mating in deer mice, Peromyscus maniculatus. Behav. Ecol. Sociobiol. 11, 37-42.
Dewsbury, D. A., and Baumgardner, D. J. (1981). Studies of sperm competition in two species of muroid rodents. Behav. Ecol. Sociobiol. 9, 121-133.
PERICOPULATORY
PREGNANCY BLOCK
173
Dominic, C. J. (1966). Observations on the reproductive pheromones of mice.: II. Neuroendocrine mechanisms involved in the olfactory block to pregnancy. J. Reprod. Fertil. 11, 415-421.
Eleftheriou, B. E., Bronson, F. H., and Zarrow, M. X. (1962). Interaction of olfactory and other environmental stimuli on implantation in the deer mouse. Science (Washington. D.C.)
137, 764.
Gray. G. D., Zerylnick, M., Davis, H. N., and Dewsbury, D. A. (1974). Effects of variations in male copulatory behavior on ovulation and implantation in prairie voles, Microfus ochrogaster.
Harm. Behav. 5, 389-3%.
Gustafsson, T., Andersson, B., and Westlin, L. (1980). Reproduction in a laboratory colony of bank vole, Cfethrionomys glare&s. Canod. J. Zoo/. 58, 1016-1021. Healey, M. C. (1%7). Aggression and self-regulation of population size in deermice. Ecology 48, 377-392. Kenney, A. McM., Evans, R. L., and Dewsbury, D. A. (1977). Postimplantation pregnancy disruption in Microtus ochrogaster, M. pennsylvanicus, and Peromyscus maniculatus. J. Reprod.
Fertil.
49, 365-367.
Keveme, E. B., and de la Riva, C. (1982). Pheromones in mice: Reciprocal interaction between the nose and the brain. Nature (London) 296, 148-150. Lott, D. F., and Hopwood, J. H. (1972). Olfactory pregnancy-block in mice (Mus musculus): An unusual response acquisition paradigm. Anim. Eehnv. 20, 263-267. Mallory, F. F., and Brooks, R. J. (1978). Infanticide and other reproductive strategies in the collared lemming, Dicrosfonyx groenlandicus. Nature (London) 273, 144-146. Mallory, F. F., and Clulow, F. V. (1977). Evidence of pregnancy failure in the wild meadow vole, Microrus pennsylvanicus. Canod. J. Zool. 55, I-17. Meyer, B. J., and Meyer, R. K. (1944). Growth and reproduction of the cotton rat, Sigmodon hispidus hispidus, under laboratory conditions. 3. Mammal. 25, 107-129. Milligan. S. R. (1979). Pregnancy blockage and memory of the stud male in the vole (Microtus
agrestis).
1. Reprod.
Fertil.
57, 223-225.
Milligan, S. R. (1980). Pheromones and rodent reproductive physiology. Symp. Zoo/. Sot. London 45, 251-275. Myers, P., and Master, L. L. (1983). Reproduction by Peromyscus mnniculatus: Size and compromise. J. Mammal. 64, l-18. Norris, M. L., and Adams, C. E. (1977). Delayed implantation in the Mongolian gerbil, Meriones
unguicularus.
27, 486-487.
Norris, M. L., and Adams, C. E. (1979). Mating post partum, concurrent lactation and reproduction in the laboratory rat. Lab. Anim. 13, 167-170. Norris, M. L., and Adams, C. E. (1981). Concurrent lactation and reproductive performance in CFLP mice mated post partum. Lab. Anim. 15, 273-275. Rohrbach, C. (1982). Investigation of the Bruce effect in the Mongolian gerbil (Meriones unguiculatus). J. Reprod. Fertil. 65, 41 l-417. Rood, J. P. (1966). Observations on the reproduction of Peromyscus in captivity. Amer. Midl.
Nat. 76, 4%-503.
Schwagmeyer, P. E. (1979). The Bruce effect: An evaluation of male/female advantages. Amer. Nat.
114, 932-938.
Stehn, R. A., and Jannett, F. J. (1981). Male-induced abortion in various microtine rodents. J. Mammal.
62, 369-372.
Stehn, R. A., and Richmond, M. E. (1975). Male-induced pregnancy termination in the prairie vole, Microtus ochrogasrer. Science (Washington, D.C.) 187, I21 I-1213. Svihla, A. (1932). A comparative life history study of the mice of the genus Peromyscus. Misc. Publ. Mus. Zool. Univ. Michigan 24, l-39. Wynne-Edwards, K. E., and Lisk. R. D. (1984). Djungarian hamsters fail to conceive in the presence of multiple males. Anim. Behov. 32, 626-628.