Consistency in maternal behavior within families of free-ranging rhesus monkeys An extension of the concept of maternal style.код для вставкиСкачать
American Journal of Primatology 22:159-169 (1990) Consistency in Maternal Behavior Within Families of Free-Ranging Rhesus Monkeys: An Extension of the Concept of Maternal Style CAROL M. BERMAN Department of Anthropology, State University of New York at Buffalo, Buffalo,New York; Caribbean Primate Research Center, Box 906, Punta Santiago, Puerto Rico This paper assesses the extent to which free-ranging rhesus monkeys (Macaca mulatta) on Cay0 Santiago show consistent mother-infant interaction with individual infants over time and consistent maternal styles from one infant to another. Inter-pair differences in several measures of mother-infant interaction related to proximity and contact were correlated positively and highly significantly between 6-week age periods. Analysis of variance methods were used to show that, after the first 6 weeks, these measures of interaction were also more similar among pairs with the same mother (i.e., families) than among pairs with different mothers. Differences between families were also consistent between age periods, and appeared to be due primarily to differences between mothers rather than to differences between sets of infant siblings. The results suggest that the concept of maternal style can be used to describe variation between both individual Cay0 Santiago mother-infant pairs and between individual Cay0 Santiago mothers. Key words: Macaca mulatta, maternal behavior, maternal style, Cay0 Santiago INTRODUCTION Inter-individual differences in mother-infant interaction among nonhuman primates can often be described in terms of particular qualities of relationships (e.g., restrictive, protective) [Hinde & Simpson, 19751. The description of such qualities is generally based on the identification of differences in one or more measures of mother-infant interaction between individual mother-infant pairs which are consistent over time, and on consistent relationships between measures. The measures are typically those concerned with the regulation of contact and proximity between mothers and infants. Such consistency at the level of individual mother-infant pairs appears to be common among cercopithecines, e.g., rhesus [Hinde & Spencer-Booth, 1967; Berman, 1978, 1982; Simpson & Simpson, 19861, yellow baboons [Altmann, 19801, vervets [Fairbanks & McGuire, 1987,19881, but is not universal, e.g., bonnet macaques [Silk, in press]. Received for publication December 22, 1989; revision accepted May 25, 1990. Address reprint requests to Dr. Carol M. Berman, Department of Anthropology, State University of New York, Buffalo, NY 14261. 0 1990 Wiley-Liss, Inc. 160 I Berman The term maternal style has also been used to describe such consistency [Altmann, 19801. This refinement is justified when the primary responsibility for differences in interaction can be traced to differences between mothers rather than infants, as it has been for a range of species [e.g., Hinde & Spencer-Booth, 1967; Rosenblum & Youngstein, 1974; Altmann, 1980; Simpson & Simpson, 1986; Fairbanks & McGuire, 1987; see, however, Chalmers 119721. Because discussions of maternal style are usually based on analyses of individual mother-infant relationships within a population over a limited time period, it is not clear whether particular maternal styles are descriptive only of individual mother-infant relationships or of more enduring qualities of particular mothers. Some researchers have casually noted tendencies for mothers to display similar maternal styles with successive infants [Altmann, 1980; Goodall, 1986; see, however, Dolhinow, 1980, 19821, but so far there is little quantitative data available to assess the extent to which this sort of within-family consistency occurs. Berman [ 19901 recently used analysis of variance methods to show that free-ranging rhesus monkeys (Macaca mulatta) who have reared several infants both differ consistently from one another in the rates at which they reject infants, and display similar rejection rates with each of their infants. Fairbanks [19891 had similar results for percentage of time captive vervet mothers and infants spent in mother-infant contact with one another. However, the extent to which this is true for a variety of measures of mother-infant interaction remains unclear. This paper examines the extent to which the concept of maternal style can be used to describe variation in motherinfant relationships both among individual pairs and among mothers who have reared a number of infants. This is done by (1) reassessing the extent to which individual rhesus mother-infant pairs on Cay0 Santiago show consistent patterns of contact and proximity interaction across infant age periods, using a larger sample than previously [Berman, 1978, 19821; (2) extending analyses of within-family consistency to several measures of mother-infant interaction; and (3) assessing the extent to which differences between families are consistent across age periods; and (4) are due primarily to differences between mothers or between sets of infant siblings. Detailed descriptions of particular maternal styles displayed will be presented elsewhere [Berman et al., in preparation]. METHODS The study took place on Cay0 Santiago, Puerto Rico, between January 1974 and December 1986. Cay0 Santiago is a 15.2 ha island off the east coast of Puerto Rico that has supported a closed, free-ranging, provisioned, predator-free population of rhesus monkeys since 1938. The population has been censused daily since 1956, and hence is well habituated and made up of known individuals with known histories and maternal kinship relationships. See Rawlins & Kessler [19831 and Sade et al. [19851 for detailed descriptions of the history, terrain, population, and management. The study group, group I, was formed naturally by fissioning in 1961 and was organized in a manner typical of multimale, multifemale macaque groups [e.g., Lee, 19831. It consisted of about 64 individuals in 1974 and grew to a peak of about 320 in 1984, before fissioning into five smaller groups in 1985 and 1986. There were three matrilineages in group I, with members spanning as many as four generations. The sample included a total of 38 mothers and 124 infants born into all three lineages (9 in 1974,12 in 1975,15 in 1979,21 in 1983,22 in 1984, 21 in 1985, and 24 in 1986). The 88 infants born between 1983 and 1986 were the offspring of females observed previously as mothers and/or infants. Each mother had from 1to 7 infants in the sample (XtSD=3.3&1.3).Male infants represented Consistency in Maternal Behavior / 161 TABLE I. Measures of Mathe-Infant Interaction Time off: the percentage of time mothers and infants were neither in ventroventral contact or i n nipple contact was estimated from point time samples. Time on uentrum: the percentage of time mother-infant pairs spent in ventroventral contact was estimated from point time samples. Time on the nipple: the percentage of time infants spent on the mother’s nipple was estimated from point time samples. Time >5 m: the percentage of time off that was spent >5 m from the mother was estimated from point time samples. Relative rejections: the proportion of all attempted nipple contacts, including those initiated by the mother, that were prevented by the mother. Maternally initiated contact: the proportion of all nipple contacts which were initiated by the mother. Proximity index: a measure of the infant’s relative role in maintaining proximity to the mother; it is the percentage of approaches made by the infanC to the mother over a 60-cm limit minus the percentage of departures made by the infant from the mother over a 60-cm limit (Hinde & Atkinson 1970). 621124 or 50.0% of the sample. First-born infants represented about 25% of the sample. Observations began as soon as possible after birth and after the annual trapping period during January and February. In most cases they continued until the infant was at least 30 weeks of age. Exceptions included the infants born in 1979 (who were observed only during weeks 1-12), 4 infants that died before 30 weeks, and 5 infants that were born exceptionally late in the season. Focal-animal sampling methods [Altmann, 19741were used to record all social interactions of infants with their mothers, along with the identities of the initiator and the recipient of the interactions and the times they occurred. At two-minute intervals, point (instantaneous) time samples were taken identifying individuals in contact with the focal animal, within 60 cm and between 60 cm and 5 m. All observations were made between 700 hr and 1,200 hr local time, hours which included all major group activities: feeding a t the food bins, foraging, traveling, socializing, and resting. However, time spent near the feeder bins was not included in the observations. The sequence of infants observed was determined by a randomized order arranged at the beginning of each week. Each mother-infant pair was observed for 3 hr during each 2-week age period, generally in 3 half-hour sessions per week. This amount of time is comparable to that typically used in other provisioned colonies where observation conditions are excellent and where interaction rates tend to be high. These data were suffkient to show stable individual differences across age periods [Berman, 1978, 19821 (see aIso Table I11 below). Table I lists seven measures of mother-infant interaction chosen for analysis, along with their definitions. Four measures are basic measures of percentages of time mothers and infants spent in contact and in proximity to one another. Three additional measures deal with the dynamics of contact and proximity maintenance. These measures are particularly usefkl in determining the relative responsibility of mothers and infants for differences between relationships [Hinde, 19691. In addition, they provide a basis for direct comparisons between studies done in captivity [e.g., Hinde and Spencer-Booth, 19671and in the wild [e.g., Berman et al., in preparation]. Measures concerned with the dynamics of nipple contact are emphasized because differential patterns of nipple contact are thought to have par- 162 I Berman TABLE 11. Numbers of Infants in the Sample for Each Mother During Each 6-Week Age Period Age Period (weeks) Mother’s name 526 604 WE 530 673 FH 348 501 3R 233 446 262 454 497 234 44 1 317 498 676 470 443 AA cs 528 585 A43 584 674 A47 587 A41 678 633 A27 A26 A3 1 A25 C62 Total mothers: 38 Total infants: 124 1-6 2 2 1 3 1 2 3 2 1 1 2 3 1 2 1 3 1 2 1 2 2 1 3 1 1 1 26 44 7-12 1 1 2 3 1 - 4 3 2 4 4 2 4 1 6 3 4 2 1 3 1 2 2 2 2 4 2 1 3 2 1 3 2 1 2 2 1 36 84 13-18 2 1 3 3 2 3 3 2 5 3 2 4 1 6 2 2 1 2 3 1 2 1 3 3 3 3 2 3 3 2 3 2 1 2 3 1 36 88 19-24 25-30 3 3 1 1 4 3 2 1 3 3 2 6 3 3 5 1 5 2 2 3 4 2 1 3 4 2 6 3 3 4 5 2 2 1 3 3 1 2 2 3 3 3 3 2 4 3 2 5 3 1 3 3 1 37 100 1 3 3 1 2 3 3 3 3 3 3 4 3 3 5 2 2 3 3 35 101 Total hours per mother 81 36 126 135 63 27 144 126 90 216 135 99 162 45 225 9 99 90 45 90 135 45 90 81 99 99 135 117 72 126 99 81 171 90 45 90 108 36 - titularly important consequences for infant development and for subsequent maternal reproductive performance [e.g., Altmann, 1980; Nicolson, 1982, 1987; Lee, 1987; Stewart, 1988; Gomendio, 1989; Johnson et al., in preparation]. The data were analyzed within five 6-week infant age periods: (1-6 weeks, 7-12 weeks, 13-18 weeks, 19-24 weeks, and 25-30 weeks). Only mother-infant pairs whose data were complete for each 2-week period within a 6-week age period were included in the sample for that 6-week age period. Hence all mother-infant pairs included within an analysis for a given 6-week age period were observed for Consistency in Maternal Behavior I 163 TABLE 111. Consistency of Inter-pair Differences in Measures of Mother-Infant Interaction Over Time: Pearson Correlation Coefficients ~ ~ ~ ~~~~~ Between nonadjacent age periods Between adjacent age periods Time off N(m-i pairs)b Time on ventrum Nb Time on nipple Nb Time >5 m Nb Relative rejections Nb Maternally initiated contact Nb Proximity index Nb 1-6 wk vs . 7-12wk 7-12 wk vs. 13-18 wk 13-18 wk vs. 19-24 wk 19-24 wk vs. 25-30 wk -0.06 39 0.59*** 39 -0.10 39 0.33** 72 0.61*** 72 0.32** 71 0.60*** 72 0.76*** 61 0.53*** 0.41*** 88 0.66*** 88 0.41*** 88 0.61*** 88 0.69*** 86 0.39*** 0.36*** 95 0.68*** 95 0.35*** 95 0.63*** 95 0.57*** 95 0.29*** 60 0.54*** 72 81 0.61*** 88 90 0.72*** 95 - 0.60*** 28 0.65*** 28 - X w Proportion positive and significant" 0.26 0.08 216 0.64 0.46 616 0.19 0.04 116 0.61 0.48 313 0.66 0.50 616 0.47 0.29 516 0.62 0.38 313 aProportion of coefficients that are positive and significant a t least at a P<0.05 level bNumber of mother-infant pairs. *P<0.05. **P<O.Ol. ***P<O.OOl. the same amount of time (9 hr) during that age period. Table I1 lists the number of infants included in the sample for each mother during each 6-week age period, and the total amount of time each mother was observed. Since undisturbed observations could not begin until after the annual trapping period [Berman, 19891, a number of infants born before the end of trapping could not be observed in their early weeks. For this reason sample sizes increased with each 6-week age period. Arcsine transformation were used with all proportional measures, such that (T = arcsin where p was the proportional measure [Sokal & Rohlf, 19691. With such transformations (and in most cases without) the data met the assumptions for parametric correlation analysis and analysis of variance. All statistical tests were two-tailed. 6 RESULTS Consistency over Age Periods for Individual Dyads Pearson correlation tests were used to assess whether differences in interaction between individual mother-infant pairs were consistent across 6-week age periods. With only two exceptions, correlations for each measure of mother-infant interaction across adjacent age periods were positive and highly significant (Table 111). For each measure except Time on the Nipple and Time Off, most correlations across nonadjacent age periods were also positive and significant, although they tended to be less strong. Consistency over Age Periods for Families Pearson correlation tests were also used to assess whether differences in interaction between families were consistent across 6-week age periods. Mean mea- 164 I Berman TABLE IV. Consistency of Inter-familyDifferences in Measures of Mother-Infant Interaction over Time: Pearson Correlation Coefficients Between nonadjacent age periods Between adjacent age periods 1-6 wk 7-12 wk 13-18 wk 19-24 wk vs. vs. vs. vs. 7-12 wk 13-18 wk 19-24 wk 25-30 wk -0.09 Time off N (mothers) 24 Time on ventrum 0.47** N (mothers) 24 -0.07 Time on nipple N (mothers) 24 Time >5 m N (mothers) Relative rejections 0.61*** N (mothers) 22 Maternally initiated contact 0.58** N (mothers) 22 Proximity index N (mothers) 0.11 36 0.56*** 36 0.09 36 0.45** 36 0.71*** 34 0.57*** 34 0.44** 36 0.35* 36 0.61*** 36 0.34* 36 0.79*** 36 0.72*** 35 0.48** 35 0.68*** 36 0.45** 34 0.79*** 34 0.42** 34 0.79*** 34 0.70*** 33 0.27(*) 33 0.77*** 34 -X Proportion positive and significant" 0.20 0.11 116 0.61 0.35 316 0.19 0.06 2/6 0.68 0.42 313 0.68 0.51 616 0.48 0.38 4/6 0.63 0.44 3/3 "Proportion of coeficients that are positive and significant at least a t a P<0.05 level. '"P<O. 1. *P<0.05. **P<O.Ol. ***P<O.OOl. sures of mother-infant interaction were calculated for each family (i.e., for each mother with all her infants in the sample). Not surprisingly, the results were similar to those for individual mother-infant pairs (Table IV). With five exceptions, correlations across adjacent age periods were positive and significant. With the exceptions of Time on the Nipple, Time Off, and Time on the Ventrum, most correlations across nonadjacent age periods were also positive and significant, although less strong. Consistency of Mothers with Each of Their Infants Analysis of variance (ANOVA) methods were used to assess the extent to which mothers with two or more infants in the sample displayed similar measures of mother-infant interaction with each infant. A separate test was carried out for each measure and for each 6-week age period. In each test, a measure of interaction was entered as the dependent variable and mother's identity was entered as a factor. Because some measures varied with yearly changes in the size of the social group [Berman, 19881, group size was controlled by entering it as a covariate before the effect of the factor was assessed. Hence, each analysis was able to determine whether the mother's identity could account for a significant proportion of the variation after any variation attributable to group size was removed. With few exceptions after the first age period, mother's identity emerged as a significant or near significant source of variation between mother-infant pairs (Table V). In other words, after the first six weeks, scores for mother-infant pairs with the same mothers (i.e., families) tended to be more similar than those for pairs with different mothers. There was no tendency for infant siblings to be of the same sex (F ratio for infant sex ratio by mother's identity = 0.77, df = 27, NS). Although Consistency in Maternal Behavior / 165 TABLE V. Consistency of Measures of Mother-Infant Interaction within Families: F Ratios for Analysis of Variance of Mother-Infant Measures by Mother's Identity, Controlling for Group Size Age period (wk) 1-6 Time off Time on ventrum Time on nipple Time >5 m Relative rejections Maternally initiated contact Proximity index N (mothers) N (mother-infant pairs) 1.68 0.51 1.48 0.76 0.68 14 31 7-12 1.80*" 1.83*" 1.80*" 2.12*a 1.86*' 3.38***b 4.02***a 26 74" 66b 13-18 19-24 25-30 1.64'" 1.44 1.58"' 1.52"' 2.05* 1.53'*' 2.37** 29 82 1.63* 2.20** 1.83* 0.82 2.82*** 1.82* 2.63*** 30 94 2.20** 1.31 2.28** 1.17 2.54*** 2.49** 1.90* 31 98 '"P<O. 1. *P<0.05. *+P<O.Ol. ***P<O.OOl. infants within families tended to have similar numbers of older immature siblings (F ratio for number of immature siblings by mother's identity = 2.24, df=27, P<O.OOl), this factor could not account entirely for within-family consistency. When ANOVA tests for within-family consistency in measures of mother-infant interaction were repeated controlling for numbers of older immature siblings, mother's identity generally continued to emerge as a significant or near significant source of variation (Table VI). Responsibility for differences. Hinde's  methods were used to assess whether differences in measures of mother-infant interaction between families were due primarily to differences between mothers or between sets of infant siblings. Applying the method involved hypothesizing two simple kinds of differences between mothers or sets of infants: (1)mothers may have differed in their propensities to seek contact and proximity with infants, or (2) sets of infants may have differed in their propensities to seek contact and proximity with their mothers. Predictions were subsequently made about the directions with which the various measures of contact and proximity would differ for each hypothesized type of difference in prcpensities. For example, if differences between mothers were primarily responsible for differences between families, one would predict that relatively high levels of contact between mothers and their infants would be associated with relatively low levels of maternal rejections and with relatively high proportions of their contact initiated by the mother (i.e., negative correlations between time in contact and relative rejections, and positive correlations between time in contact and proportions of maternally initiated contact). Similarly one would predict that relatively low levels of time at a distance would be associated with infants taking relatively small roles in maintaining proximity to their mothers (i.e. positive correlations between Time >5 m and the Proximity Index). On the other hand, if differences between infants were primarily responsible for differences between families, one would predict correlations in the opposite directions. The hypothesis that simple kinds of differences were responsible for differences between families would be supported to the extent 166 / Berman TABLE VI. Consistency of Measures of Mother-Infant Interaction within Families: F Ratios for Analysis of Variance of Mother-Infant Measures by Mother's Identity, Controlling for Group Size and Numbers of Older Immature Siblings Ape Deriod (wk) Time off Time on ventrum Time on nipple Time >5 m Relative rejections Maternally initiated contact Proximity index N (mothers) N (mother-infant pairs) 1-6 7-12 1.59 7.25* 1.43 0.48 1.32 14 31 1.98'*'" 1.74"'" 1.94'*'" 1.82"'" 3.58**b 2.93**b 5.06***" 26 74" 66b 13-18 19-24 25-30 1.89* 1.81(*' 1.80'"' 1.92* 2.17* 1.37 2.19* 29 82 2.56** 2.34** 2.91** 1.02 2.53** 1.21 2.10* 30 94 3.17*** 1.69'*' 2.83** 0.87 2.82** 2.70** 1.68'" 31 98 '"P < 0.1. *P 10.05. **P <0.01. ***P <0.001. that marked correlations could be found between measures in the predicted directions. To apply the method, mean measures of mother-infant interaction were calculated for each family. Relationships between measures were assessed using Pearson correlation tests. The results (Table VII) were generally in harmony with the hypothesis that differences between mothers were primarily responsible for differences between families, at least after the first 6 weeks. DISCUSSION Both families and individual rhesus mother-infant pairs on Cay0 Santiago show significant degrees of consistency over successive infant age periods in several aspects of interaction related to proximity and contact. In addition, mothers who raise several offspring show a significant degree of consistency with each of their infants, at least after the first 6 weeks. This within-family consistency is not dependent on the sex of the infant or on the presence of older siblings, factors frequently cited as important sources of variation between mother-infant pairs [e.g., Hooley & Simpson, 1983; Berman, 1984; Simpson & Simpson, 1986; Silk, in press]. These findings, along with indications that in rhesus monkeys both inter-pair differences [e.g., Hinde & Spencer-Booth, 19671 and inter-family differences in measures appear to be due more immediately to differences between mothers than to differences between infants, are consistent with the concept of maternal style, and suggest that an extension of the concept to individual mothers may also be justified, at least for Cay0 Santiago rhesus. The extent to which such consistency applies to other environments or other species remains to be seen. Simpson & Howe [19861 found significant differences for several measures of mother-infant interaction between different groups of captive rhesus mothers, but not between mothers within each group. Fairbanks  found consistent differences in percentages of contact between 11 captive vervet mothers of 21 infants, and hypothesized that maternal consistency may be due to temperamental consistency, particularly with reference to confidence and timidity. Fairbanks suggested that this consistency is rooted in both genetics and the mother's experience as an infant. Consistency in Maternal Behavior / 167 TABLE VII. Responsibility for Differences in Mother-Infant Interaction Measures Between Families: Pearson Correlation Coefficients Between Mean Scores for Individual Mothers Aee ueriods (wk) Relative rejections with Time on ventrum Time on nipple Time off Maternally initiated contact with Time on ventrum Time on nipple Time off Proximity index with Time >5 m N (mothers) 1-6 7-12 13-18 19-24 25-30 -0.24 -0.10 0.15 -0.40**a -0.32*" 0.33*" -0.31* -0.26'*' 0.23 -0.46** -0.48*** 0.47*** -0.41** -0.28* 0.33* 0.66*** -0.20 0.08 0.15" 0.02" -0.02" 0.47** 0.27'" -0.30* 0.65*** 0.46** -0.42** 0.38* 0.09 -0.03 26 0.43**b 34" 0.43** 36 0.35* 36 0.18 34 36b '"P <0.1. *P <0.05. **P 10.01. ***P <0.001 However, later experience and relatively stable social circumstances during motherhood (e.g., rank) should also be considered [see also Berman, 19901. The findings that mothers are primarily responsible for differences between families do not necessarily imply that within-family consistency is due entirely to enduring maternal attributes. Consistent maternal behavior could also result from enduring aspects of the family's social environment. Nor do the results imply that differences between infants do not exist or do not influence mother interaction. Indeed, differences between infants could theoretically contribute to within-family consistency in maternal style. For example, captive rhesus infants show early and consistent differences in their responses to stressful situations [see review, Suomi, 19871. Such differences could influence mother-infant interaction directly or indirectly by influencing maternal behavior. That such early differences may be in part inherited [Suomi, 19871 could lead to similarities between mother-infant pairs with the same mothers. So far, however, cross-fostering studies suggest that such infant-based factors do not account for consistent maternal behavior [Suomi, 19871. Although the consequences of various maternal styles are not well understood, there is evidence that aspects of maternal style may influence both the infant's psychosocial development and the mother's subsequent reproductive performance, including the length of interbirth intervals [see reviews, Nicolson, 1982, 1987; Gomendio, 1989; Johnson et al., in preparation). An implication of the consistency of maternal style from infant to infant is that the long term consequences of particular styles may accumulate over a lifetime and have profound effects on life history variables. For example, t o the extent that maternal style influences the length of interbirth intervals, particular styles could have large effects on the total number of offspring a female produces in her lifetime. Similarly, to the extent that maternal style is associated with infant mortality, morbidity [Altmann, 19801, or the rate or quality of psychosocial development [e.g., Kaufman and Stynes, 1978; Dolhinow, 19801, consistent styles could lead to more consistent risks or advantages for offspring of particular mothers. 168 / Berman CONCLUSIONS 1. Individual rhesus mother-infant pairs on Cay0 Santiago, Puerto Rico, show consistent differences in measures of mother-infant interaction over successive infant age periods. 2. After the first 6 weeks, measures of interaction are more similar for mother-infant pairs with the same mothers (i.e., families) than for pairs with different mothers. 3. Differences in measures of mother-infant interaction between families are also consistent over successive age periods and are due primarily to differences between mothers rather than to differences between sets of infant siblings. 4. Maternal style is a term that can be used to describe both consistent differences in mother-infant relationships between individual pairs and between families of free-ranging rhesus monkeys. ACKNOWLEDGMENTS I thank the Donald Sade, Richard Rawlins, Matt Kessler, and Dell Collins for permission to carry out research on Cay0 Santiago; and L. Busaca, A. Homer, E. Kapsalis, N. Printer, L. Ritchie, and L. Smith for assistance in data collection. Cay0 Santiago was supported by NIH contracts DRR 71-2003 and RR-7-2115, and by NIH grant RR-01293 to the University of Puerto Rico School of Medicine. I received support from the Wenner-Gren Foundation for Anthropological Research, the Explorer’s Club, Sigma Xi Society, NIMH predoctoral fellowship MH 05195, NSF postdoctoral fellowship SPI-7815548, and NIMH grant MH38647. REFERENCES Altmann, J. Observational study of behavior: Sampling methods. BEHAVIOUR 49: 227-65, i97z. Altmann, J. BABOON MOTHERS AND INFANTS. Cambridge, Massachusetts, Harvard University Press, 1980. Berman, C.M. SOCIAL RELATIONSHIPS AMONG FREE-RANGING INFANT RHESUS MONKEYS. Ph.D. thesis, University of Cambridge, 1978. Berman, C.M. The social development of a n orphaned rhesus infant on Cay0 Santiago: Male care, foster mother-orphan interaction and peer interaction. AMERICAN JOURNAL OF PRIMATOLOGY 3:131141, 1982. Berman, C.M. Variation in mother-infant relationships: Traditional and nontraditional factors. Pp. 17-36 in FEMALE PRIMATES: STUDIES BY WOMEN PRIMATOLOGISTS. M.F. Small ed. New York, Alan R. Liss, 1984. Berman, C.M. Demography and mother-infant relationships: Imdications for ~ O U structure. Pp. 269-296in ECOLOGYIND BEHAVIOR OF FOOD-ENHANCED PRIMATE GROUPS. J. E. Fa; C. H. Southwick eds. New York, Alan R. Liss, 1988. Berman, C.M. Trapping activities and mother-infant relationships on Cay0 Santiago: A cautionary tale. PUERTO RICO HEALTH SCIENCES JOURNAL 8:73-78, 1989. Berman, C.M. Intergenerational transmission of maternal rejection rates among free-ranging rhesus monkeys. ANIMAL BEHAVIOUR 39:239-247, 1990. Berman, C.M.; Johnson, R.; Malik, I. Maternal style among rhesus monkeys: a comparison between populations, in preparation. Chalmers, N.R. Comparative aspects of early infant development in some captive cercopithecines. Pp. 63-82 in PRIMATE SOCIALIZATION. F. Poirier ed. New York, Random House, 1972. Dolhinow, P. An experimental study of mother loss in the Indian langur monkey (Presbytzs entellus). FOLIA PRIMATOLOGICA 33:77-128, 1980. Dolhinow, P.; DeMay, M.G. Adoption: the importance of infant choice. JOURNAL OF HUMAN EVOLUTION 11:391-420,1982, D Fairbanks, L.A. Early experience and crossgenerational continuity of mother-infant contact in vervet monkeys. DEVELOPMENTAL PSYCHOBIOLOGY 22:669681, 1989. Fairbanks, L.A.; McGuire, M.T. Mother-infant relationships in vervet monkeys: Response to new adult males. INTERNA- Consistency in Maternal Behavior i 169 TIONAL JOURNAL OF PRIMATOLOGY 8:351-366, 1987. Gomendio, M. Suckling behaviour and fertility in rhesus macaques (Macaca mulatta). JOURNAL OF ZOOLOGY, LOND. 217:449-467,1989. Goodall, J. THE CHIMPANZEES OF GOMBE: PATTERNS OF BEHAVIOUR. Cambridge, Massachusetts, Harvard University Press, 1986. Hinde, R.A. Analysing the roles of the partners in a behavioural interaction-motherinfant relations in rhesus macaques. ANNALS OF THE NEW YORK ACADEMY OF SCIENCES. 159:651-657, 1969. Hinde, R.A.; Atkinson, S. Assessing the roles of social partners in maintaining mutual proximity as exemplified by motherinfant relations in rhesus monkeys. ANIMAL BEHAVIOUR 18:169-176,1970. Hinde, R.A.; Simpson, M.J.A. Qualities of mother-infant relationships in monkeys. Pp. 39-68 in PARENT-OFFSPRING RELATIONSHIPS. CIBA Foundation Symposium 33. Amsterdam, Elsevier, 1975. Hinde, R.A.; Spencer-Booth, Y. The behaviour of socially-living rhesus monkeys in their first two and a half years. ANIMAL BEHAVIOUR 15:169-196,1967. Hooley, J.M.; Simpson, M.J.A. Influence of siblings on the infant’s relationships with the mother and others. Pp. 139-142 in PRIMATE SOCIAL RELATIONSHIPS: AN INTEGRATED APPROACH. R.A. Hinde ed. Oxford, Blackwell Scientific Publications, 1983. Johnson, R.; Berman, C.M.; Malik, I. Model of the lactational and environmental control of female mating behavior in a seasonally breeding primate. In preparation. Lee, P.C. Species, study sites and methods. Pp. 8-16 in PRIMATE SOCIAL RELATIONSHIPS: AN INTEGRATED APPROACH. R.A. Hinde ed. Oxford, Blackwell Scientific Publications, 1983. Lee, P.C. Nutrition, fertility and maternal investment in primates. JOURNAL OF ZOOLOGY, LOND. 213~409-422, 1987. Kaufman, I.C.; Stynes, A.J. Depression can be induced in a bonnet macaque infant. PSYCHOSOMATIC MEDICINE 4071-75, 1978. Nicolson, N. WEANING AND THE DEVELOPMENT OF INDEPENDENCE IN OL- IVE BABOONS. Ph.D. dissertation, Harvard University, 1982. Nicolson, N. Infants, mothers and other females. Pp.330-342 in PRIMATE SOCIETIES. B.B. Smuts; D.L. Cheney; R.M. Seyfarth; R.W. Wrangham; T.T. Struhsaker eds. Chicago, University of Chicago Press, 1987. Rawlins, R.G.; Kessler, M.J. Congenital and hereditary abnormalities in the rhesus monkeys of Cay0 Santiago. TERATOLOGY 28:169-174,1983. Rosenblum, L.A.; Youngstein, K. Developmental changes in compensatory dyadic responses in mother and infant monkeys. Pp. 141-162 in THE EFFECT OF THE INFANT ON ITS CAREGIVER. M. Lewis; L. Rosenblum eds. New York, Wiley Press, 1974. Sade, D.S.; Chepko-Sade, B.D.; Schneider, J.M.; Roberts, S.S.; Richtsmeier, J.T. BASIC DEMOGRAPHIC OBSERVATIONS ON FREE-RANGING RHESUS MONKEYS. New Haven, Connecticut, Human Relations Area File, 1985. Silk, J.E. Mother-infant relationships in bonnet macaques: Sources of variation in proximity. INTERNATIONAL JOURNAL OF PRIMATOLOGY, in press. Simpson, M.J.A.; Howe, S. Group and matriline differences in the behaviour of rhesus monkey infants. ANIMAL BEHAVIOUR 34:444-459, 1986. Simpson, M.J.A.; Simpson, A.E. The emergence and maintenance of interdyad differences in the mother-infant relationships of rhesus monkeys: A correlational study. INTERNATIONAL JOURNAL OF PRIMATOLOGY 7:379-399, 1986. Sokal, R.; Rohlf, F.J. BIOMETRY. San Francisco, Freeman, 1969. Stewart, K.J. Suckling and lactational anoestrus in wild gorillas (Gorilla gorilla). JOURNAL OF REPRODUCTION AND FERTILITY 83:627-634,1988. Suomi, S.J. Genetic and maternal contributions t o individual differences in rhesus monkey biobehavioral development. Pp. 397-420 in PERINATAL DEVELOPMENT: A PSYCHOBIOLOGICAL PERSPECTIVE. N. Krasnegor; E. Blass; M. Hofer; W. Smotherman eds. New York, Academic Press, 1987.