- Email: [email protected]
The effects of cross-fostering on the development of social preferences in meadow voles (Microtus pennsylvanicus)
BEHAVIORALAND NEURALBIOLOGY47, 167-172 (1987)
The Effects of Cross-Fostering on the Development of Social Preferences in Meadow Voles (Microtus
pennsylvanicus) BETTY M c G U I R E AND MELINDA NOVAK 1
Department of Zoology, University of Massachusetts, Amherst, Massachusetts 01003 To determine the role of the parental environment in the formation of species attachments in the meadow vole (Microtus pennsylvanicus), young of this species were fostered to either meadow vole parents (in-fostered young) or to prairie vole parents (M. ochrogaster; cross-fostered young). At 50 to 60 days of age, subjects were tested for social preference in an evironment that simulated the conditions of a runway system. The testing design assessed the reaction of fostered animals to a novel species (M. pinetorum) as well as to the parental and biological species. The proportion of total test time spent near stimulus animals did not vary as a function of rearing condition. However, when compared to in-fostered controls, cross-fostered meadow voles displayed increased preference for prairie voles. Novelty did not appear to be a significant factor in test animal choice. These results suggest that in meadow voles, preference for the species of social partner is dependent on postnatal experience with parents. © 1987 AcademicPress, Inc.
Early experience dramatically influences the formation of species attachments in rodents (e.g., Huck & Banks, 1980; Hudgens, Denenberg, & Zarrow, 1968; Lagerspetz & Heino, 1970; McCarty & Southwick, 1977; McDonald & Forslund, 1978). This has been demonstrated using interspecific cross-fostering, a technique involving the exchange of young between mothers of different species. When tested in adulthood for social or olfactory preference, cross-fostered animals typically spend more time adjacent to either a stimulus animal of the foster species or the soiled bedding of the foster species than do control animals. In the present study we used the technique of cross-fostering to investigate the role of the parental environment in the formation of species 1 Correspondence and reprint requests should be addressed to Betty McGuire at the Department of Ecology, Ethology, and Evolution, Shelford Vivarium, University of Illinois, 606 E. Healey St., Champaign, 1L 61820. 167 0163-1047/87 $3.00 Copyright© 1987by AcademicPress, lnc. All rightsof reproductionin any form reserved.
168
MC GUIRE AND NOVAK
attachments in the meadow vole (Microtus pennsylvanicus). Meadow vole young were fostered to either meadow vole parents or to prairie vole parents (M. ochrogaster) and tested for social preference at 50 to 60 days of age. Meadow voles and prairie voles are closely related and occur sympatrically in parts of their geographic distribution (e.g., Krebs, Keller, & Tamarin, 1969). Thus, in areas of sympatry, neonatal learning of species characteristics may be important in the maintenance of sexual isolation. This study differs from previous investigations of social or olfactory preference in two ways. First, the testing design assessed the reaction of fostered animals to a novel species, the pine vole (M. pinetorum), as well as to the parental and biological species. Testing with a third species controls for novelty-based preferences and permits a relative ranking of parental vs biological vs novel species preference. Second, experiments were conducted in an environment that simulated the runway conditions under which social encounters might occur in the field. The testing apparatus was lined with a thick covering of hay and provided extensive neutral areas where test animals did not have to interact with stimulus animals. The animals used in this experiment were meadow voles (third or fourth generation offspring of voles trapped in Massachusetts), prairie voles (first generation offspring of voles trapped in Illinois), and pine voles (fourth generation offspring of voles trapped in Connecticut and New York). Experimental groups were established by the placement of breeding pairs of meadow voles or prairie voles into 1.3 x 1.3-m tablelike pens (see McGuire & Novak, 1984) on Day 17 of gestation. Once females gave birth, their pups were removed and in each case were replaced with four foster meadow vole young from the main breeding colony. Foster pups were approximately the same age as the young that were removed ( _ 24 h), and all fostering occurred within 24 _+ 12 h after birth. At 20 days of age the pups were removed and housed in groups of two or three animals of the same sex and rearing condition in aquaria (26 x 51 x 30 cm) that contained a peat-wood shaving substrate and a 10 to 20-cm-thick layer of hay. In both the pens and aquaria, sunflower seeds, oats, Purina rabbit chow, and water were provided ad libitum and animals were maintained on a 14L:10D photoperiod. Sixteen cross-fostered meadow voles (meadow vole young reared by foster prairie vole parents) and 16 in-fostered meadow voles (meadow vole young reared by foster meadow vole parents) were given a series of social preference tests between 50 and 60 days of age. Each group of 16 contained 8 males and 8 females. No more than 2 animals of the same sex were chosen from each litter for testing. The testing apparatus consisted of a four-armed Plexiglas maze. Each of the arms measured 40 x 10 x 10 cm and had a small chamber (20 x 20 x 16 cm) attached to the end. During a preference test, one chamber served
SOCIAL PREFERENCES IN MEADOW VOLES
169
as the start box, one chamber was empty, and the remaining two chambers contained stimulus voles. A fine wire mesh covered the entrances to the two stimulus chambers as well as the blank chamber. The maze was filled with a 3-cm-thick layer of hay to simulate the conditions of a runway system. The entire apparatus was placed on a 1.3 x 1.3-m Plexiglas surface which was set in a wooden frame approximately 70 cm from the floor. Observations were made from below through the use of a microprocessoroperated keyboard (MORE, Observational Systems, Seattle, WA). On the day prior to the start of a given testing period, each subject was given a 30-min habituation trial in the apparatus. All chambers were empty during the habituation trials. Actual preference testing lasted 9 days and each subject was tested once with each of the following combinations of stimulus animals: (1) meadow vole/prairie vole, (2) meadow vole/pine vole, and (3) prairie vole/pine vole. The locations of the four chambers as well as the order of the three trials were counterbalanced across subjects. Stimulus animals were of the opposite sex of the test animal and were selected from group-housed sexually inexperienced adults. No voles, whether test or stimulus animals, were used twice on the same day and no test subject went more than 3 days between trials. At the end of each 15-rain trial the apparatus was dismantled and cleaned, and the hay cover was discarded. During each test the cumulative time spent in each stimulus arm was recorded (stimulus arms were 40 cm in length and thus the duration data reflect the amount of time that test animals spent within 40 cm of stimulus animals). Data were analyzed using a two-way analysis of variance (ANOVA) with rearing condition and sex as between-subjects variables. However, because of the nonnormality of the data, all duration measures were transformed into proportions prior to performing the analysis of variance. For example, the amount of time spent next to a stimulus meadow vole in the meadow vole/prairie vole testing condition was changed into a proportion by dividing the amount of time spent next to a stimulus meadow vole by the total amount of time spent next to stimulus animals during that test. Proportion data were normally distributed. A t test also was used to determine whether preferences were present. The proportion of the total test time spent with stimulus animals did not vary as a function of rearing condition or sex in any of the three testing conditions. In general, subjects spent approximately 30% (range 22-46%) of the 15-min trial in the arms which contained stimulus animals (this is referred to as social time below); the remaining time (approximately 10 min) test animals spent in the start arm and blank arm. The proportion of social time spent with particular stimulus animals varied as a function of rearing condition in all three tests (Table 1). In the meadow vole/prairie vole testing condition, cross-fostered meadow voles spent a smaller proportion of their social time with stimulus meadow
170
MC GUIRE AND NOVAK
TABLE 1 C o m p a r i s o n s across Rearing Condition and Sex of the Proportion of Social Time Spent with Particular Stimulus A n i m a l s Rearing condition and sex In-fostered Test Meadow vole vs prairie vole Meadow vole vs pine vole Prairie vole vs pine vole
Measure M M M M PR PR
x + × + x +
Cross-fostered
Male
Female
Male
Female
100 a PR
58.5 _+ 9.8
68.9 +-- 8.3
32.5 _+ 6.3
49.9 --- 8.7
10~ b
71.2 -+ 5.8
65.5 -+ 6.9
51.9 _+ 9.1
47.1 -+ 4.7
PI I00 ~ PI
53.8 _+ 8.4
42.2 +- 4.9
80.3 + 4.9
65.0 -+ 4.5
Time with m e a d o w vole (s) × 100/time with m e a d o w vole and prairie vole (s). b Time with m e a d o w vole (s) × 100/time with m e a d o w vole and pine vole (s). c Time with prairie vole (s) × 100/time with prairie vole and pine vole (s).
voles than did in-fostered meadow voles (F(1, 28) = 7.22, P < .05). During this test, in-fostered males did not display a preference (i.e., scores did not differ significantly from 50%) while in-fostered females showed a tendency to prefer conspecific stimulus animals (t(7) = 2.29, P < .06). Furthermore, whereas cross-fostered males displayed a preference for prairie voles (t(7) -- 2.77, P < .03), cross-fostered females spent equivalent amounts of time with meadow vole and prairie vole stimulus animals. When provided with a choice of meadow voles or pine voles, crossfostered meadow voles spent a smaller proportion of their social time with meadow voles than did in-fostered controls (F(1, 28) = 7.65, P < .01). This result reflects, however, the general lack of preference by cross-fostered meadow voles as compared to the preference for conspecifics exhibited by in-fostered male (t(7) = 3.67, P < .01) and female (t(7) = 2.26, P < .06) meadow voles. During the prairie vole vs pine vole testing condition, cross-fostered meadow voles spent a larger proportion of their social'time than did infostered controls in the vicinity of the stimulus prairie vole (F(1, 28) = 17.55, P < .0005). In this case, cross-fostered males (t(7) = 6.14, P < .0005) and females (t(7) = 3.33, P < .02) both preferred the species which had reared them while in-fostered males and females did not display a preference. Finally, regardless of rearing condition, males tended to spend a larger proportion of their social time than females adjacent to stimulus prairie voles in the prairie vole vs pine vole tests (F(1, 28) = 5.25, P < .05). The development of species preference in meadow voles is subject to influence by characteristics of the early environment. However, the pro-
SOCIAL PREFERENCES IN MEADOW VOLES
171
pensity for spending time in the two maze arms that contained stimulus animals did not vary as a function of rearing condition, thus demonstrating that the preweaning environment influenced the direction of social responsiveness, rather than the tendency to approach stimulus animals. The large amount of time spent by both cross-fostered and in-fostered meadow voles in the neutral arms is consistent with the general lack of contact behavior noted for this species (Wilson, 1982). Despite the potential for novelty to influence preference testing results (Kirchhof-Glazier, 1979), previous cross-fostering studies with rodents have not controlled for this factor. In the present study, cross-fostered meadow voles were exposed to prairie vole parents from Day 1 to Day 20 postpartum. During the 30 days prior to preference testing, however, cross-fostered meadow voles were housed with meadow voles of the same sex and rearing condition. As a result of the more recent experience with meadow voles, preference for the parental species might reflect a response to novelty on the part of cross-fostered meadow voles rather than imprinting on parental characteristics during the postnatal period. This does not appear to be the case, as cross-fostered meadow voles clearly discriminated between prairie voles and pine voles in the parental species vs novel species testing condition while spending equal amounts of time with the two stimulus animals in the meadow vole vs pine vole tests. Although the estrous conditions of test and stimulus females were not monitored in the present study, several factors suggest that the results were unaffected by female reproductive state. First, both test and stimulus females were housed in all female groups from weaning at Day 20 until the completion of preference tests. This method of housing may have reduced the likelihood of females coming into estrus, as group-housed female voles exhibit delayed reproductive maturation when compared to isolated or male-exposed females (e.g., Pasley & McKinney, 1973; Richmond & Conaway, 1969). Second, previous cross-fostering experiments with microtine rodents consistently have shown that the preferences exhibited by control and cross-fostered animals are independent of the estrous conditions of test and stimulus females (Huck & Banks, 1980; McDonald & Forslund, 1978). Additionally, although studies not involving the manipulation of cross-fostering have demonstrated that the olfactory preferences of lemmings vary as a function of estrous condition, sexual experience is necessary for test males to exhibit a preference for estrous over diestrous stimulus females (Huck & Banks, 1984) and the preferences of estrous and diestrous test females have been shown only to differ in the context of choosing between dominant and subordinate conspecific males (Huck, Banks, & Wang, 1981). In the present study, all test males were sexually inexperienced and test females were choosing either between a conspecific and heterospecific male or between two heterospecific males.
172
MC GUIRE AND NOVAK
The development of preference for conspecifics may be particularly important in M. pennsylvanicus as this species occurs sympatrically with several closely related species. Because social contact among adult meadow voles typically consists of brief agonistic or mating encounters (e.g., Madison, 1980; McGuire & Novak, 1984), the formation of species attachments may occur primarily during the preweaning period. REFERENCES Huck, U. W., & Banks, E. M. (1980). The effects of cross-fostering on the behaviour of two species of North American lemmings, Dicrostonyx groenlandicus and Lemmus trimucronatus. I. Olfactory preferences. Animal Behaviour, 28, 1046-1052. Huck, U. W., & Banks, E. M. (1984). Social olfaction in male brown lemmings (Lemmus sibiricus = trimucronatus) and collared lemmings (Dicrostonyx groenlandicus). I. Discrimination of species, sex, and estrous condition. Journal of Comparative Psychology, 98, 54-59. Huck, U. W., Banks, E. M., & Wang, S. C. (1981). Olfactory discrimination of social status in.the brown lemming. Behavioral and Neural Biology, 33, 364-371. Hudgens, G. A., Denenberg, V. H., & Zarrow, M. X. (1968). Mice reared with rats: Effects of preweaning and postweaning social interactions upon adult behaviour. Behaviour, 30, 259-274. Kirchhof-Glazier, D. A. (1979). Absence of sexual imprinting in house mice cross-fostered to deer mice. Physiology and Behavior, 23, 1073-1080. Krebs, C. J., Keller, B. L., & Tamarin, R. H. (1969). Microtus population biology: Demographic changes in fluctuating populations of M. ochrogaster and M. pennsylvanicus in southern Indiana. Ecology, 50, 587-607. Lagerspetz, K., & Heino, T. (1970). Changes in social reactions resulting from early experience with another species. Psychological Reports, 27, 255-262. Madison, D. M. (1980). Space use and social structure in meadow voles, Microtus pennsylvanicus. Behavioral Ecology and Sociobiology, 7, 65-71. McCarty, R., & Southwick, C. H. (1977). Cross-species fostering: Effects on the olfactory preference of Onychomys torridus and Peromyscus leucopus. Behavioral Biology, 19, 255-260. McDonald, D. L., & Forslund, L. G. (1978). The development of social preference in the voles Microtus montanus and Microtus canicaudus: Effects of cross-fostering. Behavioral Biology, 22, 497-508. McGuire, B., & Novak, M. (1984). A comparison of maternal behaviour in the meadow vote (Microtus pennsylvanicus), prairie vole (M. ochrogaster), and pine vole (M. pinetorum). Animal Behaviour, 32, 1132-1141. Pasley, J. N., & McKinney, T. D. (1973). Grouping and ovulation in Microtus pennsylvanicus. Journal of Reproduction and Fertility, 34, 527-530. Richmond, M., & Conaway, C. H. (1969). Management, breeding, and reproductive performance of the vole, Microtus ochrogaster, in a laboratory colony. Laboratory Animal Care, 19, 80-87. Wilson, S. C. (1982). The development of social behaviour between siblings and nonsiblings of the voles Microtus ochrogaster and Microtus pennsylvanicus. Animal Behaviour, 30, 426-437.
The Effects of Cross-Fostering on the Development of Social Preferences in Meadow Voles (Microtus
pennsylvanicus) BETTY M c G U I R E AND MELINDA NOVAK 1
Department of Zoology, University of Massachusetts, Amherst, Massachusetts 01003 To determine the role of the parental environment in the formation of species attachments in the meadow vole (Microtus pennsylvanicus), young of this species were fostered to either meadow vole parents (in-fostered young) or to prairie vole parents (M. ochrogaster; cross-fostered young). At 50 to 60 days of age, subjects were tested for social preference in an evironment that simulated the conditions of a runway system. The testing design assessed the reaction of fostered animals to a novel species (M. pinetorum) as well as to the parental and biological species. The proportion of total test time spent near stimulus animals did not vary as a function of rearing condition. However, when compared to in-fostered controls, cross-fostered meadow voles displayed increased preference for prairie voles. Novelty did not appear to be a significant factor in test animal choice. These results suggest that in meadow voles, preference for the species of social partner is dependent on postnatal experience with parents. © 1987 AcademicPress, Inc.
Early experience dramatically influences the formation of species attachments in rodents (e.g., Huck & Banks, 1980; Hudgens, Denenberg, & Zarrow, 1968; Lagerspetz & Heino, 1970; McCarty & Southwick, 1977; McDonald & Forslund, 1978). This has been demonstrated using interspecific cross-fostering, a technique involving the exchange of young between mothers of different species. When tested in adulthood for social or olfactory preference, cross-fostered animals typically spend more time adjacent to either a stimulus animal of the foster species or the soiled bedding of the foster species than do control animals. In the present study we used the technique of cross-fostering to investigate the role of the parental environment in the formation of species 1 Correspondence and reprint requests should be addressed to Betty McGuire at the Department of Ecology, Ethology, and Evolution, Shelford Vivarium, University of Illinois, 606 E. Healey St., Champaign, 1L 61820. 167 0163-1047/87 $3.00 Copyright© 1987by AcademicPress, lnc. All rightsof reproductionin any form reserved.
168
MC GUIRE AND NOVAK
attachments in the meadow vole (Microtus pennsylvanicus). Meadow vole young were fostered to either meadow vole parents or to prairie vole parents (M. ochrogaster) and tested for social preference at 50 to 60 days of age. Meadow voles and prairie voles are closely related and occur sympatrically in parts of their geographic distribution (e.g., Krebs, Keller, & Tamarin, 1969). Thus, in areas of sympatry, neonatal learning of species characteristics may be important in the maintenance of sexual isolation. This study differs from previous investigations of social or olfactory preference in two ways. First, the testing design assessed the reaction of fostered animals to a novel species, the pine vole (M. pinetorum), as well as to the parental and biological species. Testing with a third species controls for novelty-based preferences and permits a relative ranking of parental vs biological vs novel species preference. Second, experiments were conducted in an environment that simulated the runway conditions under which social encounters might occur in the field. The testing apparatus was lined with a thick covering of hay and provided extensive neutral areas where test animals did not have to interact with stimulus animals. The animals used in this experiment were meadow voles (third or fourth generation offspring of voles trapped in Massachusetts), prairie voles (first generation offspring of voles trapped in Illinois), and pine voles (fourth generation offspring of voles trapped in Connecticut and New York). Experimental groups were established by the placement of breeding pairs of meadow voles or prairie voles into 1.3 x 1.3-m tablelike pens (see McGuire & Novak, 1984) on Day 17 of gestation. Once females gave birth, their pups were removed and in each case were replaced with four foster meadow vole young from the main breeding colony. Foster pups were approximately the same age as the young that were removed ( _ 24 h), and all fostering occurred within 24 _+ 12 h after birth. At 20 days of age the pups were removed and housed in groups of two or three animals of the same sex and rearing condition in aquaria (26 x 51 x 30 cm) that contained a peat-wood shaving substrate and a 10 to 20-cm-thick layer of hay. In both the pens and aquaria, sunflower seeds, oats, Purina rabbit chow, and water were provided ad libitum and animals were maintained on a 14L:10D photoperiod. Sixteen cross-fostered meadow voles (meadow vole young reared by foster prairie vole parents) and 16 in-fostered meadow voles (meadow vole young reared by foster meadow vole parents) were given a series of social preference tests between 50 and 60 days of age. Each group of 16 contained 8 males and 8 females. No more than 2 animals of the same sex were chosen from each litter for testing. The testing apparatus consisted of a four-armed Plexiglas maze. Each of the arms measured 40 x 10 x 10 cm and had a small chamber (20 x 20 x 16 cm) attached to the end. During a preference test, one chamber served
SOCIAL PREFERENCES IN MEADOW VOLES
169
as the start box, one chamber was empty, and the remaining two chambers contained stimulus voles. A fine wire mesh covered the entrances to the two stimulus chambers as well as the blank chamber. The maze was filled with a 3-cm-thick layer of hay to simulate the conditions of a runway system. The entire apparatus was placed on a 1.3 x 1.3-m Plexiglas surface which was set in a wooden frame approximately 70 cm from the floor. Observations were made from below through the use of a microprocessoroperated keyboard (MORE, Observational Systems, Seattle, WA). On the day prior to the start of a given testing period, each subject was given a 30-min habituation trial in the apparatus. All chambers were empty during the habituation trials. Actual preference testing lasted 9 days and each subject was tested once with each of the following combinations of stimulus animals: (1) meadow vole/prairie vole, (2) meadow vole/pine vole, and (3) prairie vole/pine vole. The locations of the four chambers as well as the order of the three trials were counterbalanced across subjects. Stimulus animals were of the opposite sex of the test animal and were selected from group-housed sexually inexperienced adults. No voles, whether test or stimulus animals, were used twice on the same day and no test subject went more than 3 days between trials. At the end of each 15-rain trial the apparatus was dismantled and cleaned, and the hay cover was discarded. During each test the cumulative time spent in each stimulus arm was recorded (stimulus arms were 40 cm in length and thus the duration data reflect the amount of time that test animals spent within 40 cm of stimulus animals). Data were analyzed using a two-way analysis of variance (ANOVA) with rearing condition and sex as between-subjects variables. However, because of the nonnormality of the data, all duration measures were transformed into proportions prior to performing the analysis of variance. For example, the amount of time spent next to a stimulus meadow vole in the meadow vole/prairie vole testing condition was changed into a proportion by dividing the amount of time spent next to a stimulus meadow vole by the total amount of time spent next to stimulus animals during that test. Proportion data were normally distributed. A t test also was used to determine whether preferences were present. The proportion of the total test time spent with stimulus animals did not vary as a function of rearing condition or sex in any of the three testing conditions. In general, subjects spent approximately 30% (range 22-46%) of the 15-min trial in the arms which contained stimulus animals (this is referred to as social time below); the remaining time (approximately 10 min) test animals spent in the start arm and blank arm. The proportion of social time spent with particular stimulus animals varied as a function of rearing condition in all three tests (Table 1). In the meadow vole/prairie vole testing condition, cross-fostered meadow voles spent a smaller proportion of their social time with stimulus meadow
170
MC GUIRE AND NOVAK
TABLE 1 C o m p a r i s o n s across Rearing Condition and Sex of the Proportion of Social Time Spent with Particular Stimulus A n i m a l s Rearing condition and sex In-fostered Test Meadow vole vs prairie vole Meadow vole vs pine vole Prairie vole vs pine vole
Measure M M M M PR PR
x + × + x +
Cross-fostered
Male
Female
Male
Female
100 a PR
58.5 _+ 9.8
68.9 +-- 8.3
32.5 _+ 6.3
49.9 --- 8.7
10~ b
71.2 -+ 5.8
65.5 -+ 6.9
51.9 _+ 9.1
47.1 -+ 4.7
PI I00 ~ PI
53.8 _+ 8.4
42.2 +- 4.9
80.3 + 4.9
65.0 -+ 4.5
Time with m e a d o w vole (s) × 100/time with m e a d o w vole and prairie vole (s). b Time with m e a d o w vole (s) × 100/time with m e a d o w vole and pine vole (s). c Time with prairie vole (s) × 100/time with prairie vole and pine vole (s).
voles than did in-fostered meadow voles (F(1, 28) = 7.22, P < .05). During this test, in-fostered males did not display a preference (i.e., scores did not differ significantly from 50%) while in-fostered females showed a tendency to prefer conspecific stimulus animals (t(7) = 2.29, P < .06). Furthermore, whereas cross-fostered males displayed a preference for prairie voles (t(7) -- 2.77, P < .03), cross-fostered females spent equivalent amounts of time with meadow vole and prairie vole stimulus animals. When provided with a choice of meadow voles or pine voles, crossfostered meadow voles spent a smaller proportion of their social time with meadow voles than did in-fostered controls (F(1, 28) = 7.65, P < .01). This result reflects, however, the general lack of preference by cross-fostered meadow voles as compared to the preference for conspecifics exhibited by in-fostered male (t(7) = 3.67, P < .01) and female (t(7) = 2.26, P < .06) meadow voles. During the prairie vole vs pine vole testing condition, cross-fostered meadow voles spent a larger proportion of their social'time than did infostered controls in the vicinity of the stimulus prairie vole (F(1, 28) = 17.55, P < .0005). In this case, cross-fostered males (t(7) = 6.14, P < .0005) and females (t(7) = 3.33, P < .02) both preferred the species which had reared them while in-fostered males and females did not display a preference. Finally, regardless of rearing condition, males tended to spend a larger proportion of their social time than females adjacent to stimulus prairie voles in the prairie vole vs pine vole tests (F(1, 28) = 5.25, P < .05). The development of species preference in meadow voles is subject to influence by characteristics of the early environment. However, the pro-
SOCIAL PREFERENCES IN MEADOW VOLES
171
pensity for spending time in the two maze arms that contained stimulus animals did not vary as a function of rearing condition, thus demonstrating that the preweaning environment influenced the direction of social responsiveness, rather than the tendency to approach stimulus animals. The large amount of time spent by both cross-fostered and in-fostered meadow voles in the neutral arms is consistent with the general lack of contact behavior noted for this species (Wilson, 1982). Despite the potential for novelty to influence preference testing results (Kirchhof-Glazier, 1979), previous cross-fostering studies with rodents have not controlled for this factor. In the present study, cross-fostered meadow voles were exposed to prairie vole parents from Day 1 to Day 20 postpartum. During the 30 days prior to preference testing, however, cross-fostered meadow voles were housed with meadow voles of the same sex and rearing condition. As a result of the more recent experience with meadow voles, preference for the parental species might reflect a response to novelty on the part of cross-fostered meadow voles rather than imprinting on parental characteristics during the postnatal period. This does not appear to be the case, as cross-fostered meadow voles clearly discriminated between prairie voles and pine voles in the parental species vs novel species testing condition while spending equal amounts of time with the two stimulus animals in the meadow vole vs pine vole tests. Although the estrous conditions of test and stimulus females were not monitored in the present study, several factors suggest that the results were unaffected by female reproductive state. First, both test and stimulus females were housed in all female groups from weaning at Day 20 until the completion of preference tests. This method of housing may have reduced the likelihood of females coming into estrus, as group-housed female voles exhibit delayed reproductive maturation when compared to isolated or male-exposed females (e.g., Pasley & McKinney, 1973; Richmond & Conaway, 1969). Second, previous cross-fostering experiments with microtine rodents consistently have shown that the preferences exhibited by control and cross-fostered animals are independent of the estrous conditions of test and stimulus females (Huck & Banks, 1980; McDonald & Forslund, 1978). Additionally, although studies not involving the manipulation of cross-fostering have demonstrated that the olfactory preferences of lemmings vary as a function of estrous condition, sexual experience is necessary for test males to exhibit a preference for estrous over diestrous stimulus females (Huck & Banks, 1984) and the preferences of estrous and diestrous test females have been shown only to differ in the context of choosing between dominant and subordinate conspecific males (Huck, Banks, & Wang, 1981). In the present study, all test males were sexually inexperienced and test females were choosing either between a conspecific and heterospecific male or between two heterospecific males.
172
MC GUIRE AND NOVAK
The development of preference for conspecifics may be particularly important in M. pennsylvanicus as this species occurs sympatrically with several closely related species. Because social contact among adult meadow voles typically consists of brief agonistic or mating encounters (e.g., Madison, 1980; McGuire & Novak, 1984), the formation of species attachments may occur primarily during the preweaning period. REFERENCES Huck, U. W., & Banks, E. M. (1980). The effects of cross-fostering on the behaviour of two species of North American lemmings, Dicrostonyx groenlandicus and Lemmus trimucronatus. I. Olfactory preferences. Animal Behaviour, 28, 1046-1052. Huck, U. W., & Banks, E. M. (1984). Social olfaction in male brown lemmings (Lemmus sibiricus = trimucronatus) and collared lemmings (Dicrostonyx groenlandicus). I. Discrimination of species, sex, and estrous condition. Journal of Comparative Psychology, 98, 54-59. Huck, U. W., Banks, E. M., & Wang, S. C. (1981). Olfactory discrimination of social status in.the brown lemming. Behavioral and Neural Biology, 33, 364-371. Hudgens, G. A., Denenberg, V. H., & Zarrow, M. X. (1968). Mice reared with rats: Effects of preweaning and postweaning social interactions upon adult behaviour. Behaviour, 30, 259-274. Kirchhof-Glazier, D. A. (1979). Absence of sexual imprinting in house mice cross-fostered to deer mice. Physiology and Behavior, 23, 1073-1080. Krebs, C. J., Keller, B. L., & Tamarin, R. H. (1969). Microtus population biology: Demographic changes in fluctuating populations of M. ochrogaster and M. pennsylvanicus in southern Indiana. Ecology, 50, 587-607. Lagerspetz, K., & Heino, T. (1970). Changes in social reactions resulting from early experience with another species. Psychological Reports, 27, 255-262. Madison, D. M. (1980). Space use and social structure in meadow voles, Microtus pennsylvanicus. Behavioral Ecology and Sociobiology, 7, 65-71. McCarty, R., & Southwick, C. H. (1977). Cross-species fostering: Effects on the olfactory preference of Onychomys torridus and Peromyscus leucopus. Behavioral Biology, 19, 255-260. McDonald, D. L., & Forslund, L. G. (1978). The development of social preference in the voles Microtus montanus and Microtus canicaudus: Effects of cross-fostering. Behavioral Biology, 22, 497-508. McGuire, B., & Novak, M. (1984). A comparison of maternal behaviour in the meadow vote (Microtus pennsylvanicus), prairie vole (M. ochrogaster), and pine vole (M. pinetorum). Animal Behaviour, 32, 1132-1141. Pasley, J. N., & McKinney, T. D. (1973). Grouping and ovulation in Microtus pennsylvanicus. Journal of Reproduction and Fertility, 34, 527-530. Richmond, M., & Conaway, C. H. (1969). Management, breeding, and reproductive performance of the vole, Microtus ochrogaster, in a laboratory colony. Laboratory Animal Care, 19, 80-87. Wilson, S. C. (1982). The development of social behaviour between siblings and nonsiblings of the voles Microtus ochrogaster and Microtus pennsylvanicus. Animal Behaviour, 30, 426-437.