Data and code from: Discriminative reactions to behaviour and social attributes of non-mother individuals in wild infant Japanese macaques on Yakushima
Data files
Jul 13, 2026 version files 491.70 KB
-
analysis.R
28 KB
-
dominance.csv
12.20 KB
-
grooming.csv
15.11 KB
-
handling_reaction.csv
170.04 KB
-
id.csv
810 B
-
mother_infant_interaction.csv
235.34 KB
-
observation_time.csv
6.23 KB
-
proximity_mother_infant.csv
6.74 KB
-
README.md
6.58 KB
-
relatedness.csv
10.66 KB
Abstract
Mammalian infants exhibit context-dependent behavioural reactions in social interactions. Despite extensive research on these discriminative reactions, studies have lacked diversity in infant behaviours and have reported inconsistent findings. To address these gaps, we investigated how non-human primate infants respond to non-mother individuals, predicting fewer positive and more negative reactions to rough behaviours, and fewer negative reactions toward familiar individuals, such as close kin and maternal frequent grooming partners, as well as toward those ranked higher than their mothers. We tested these predictions using all-occurrence sampling of infant handling, recording infant reactions during the first 16 weeks of life in wild Japanese macaques (Macaca fuscata yakui). We found that infants exhibited fewer positive and more negative reactions, except for leaving reactions, toward rough handling. They exhibited more frequent crying when handled by close kin; showed more positive reactions and fewer leaving and crying reactions toward individuals who frequently groomed with their mothers; and displayed more twitching and crying reactions when roughly handled by individuals ranked higher than their mothers. Younger infants exhibited more twitching, and female infants displayed more crying. Higher scores of maternal rejectiveness were linked to fewer leaving reactions in infants. These findings indicate that non-human primates acquire behavioural flexibility early in development through active engagement in social contexts. Such behavioural development may contribute to ensuring safety, promoting social bonding, conserving energy during social interactions, and maintaining maternal relationships.
Description of each file and its variables
observation_time.csv: includes observation time for each study day
- session: unique identifier for each focal observation
- infant: unique identifier for focal infant
- mother: unique identifier for focal mother
- date: date for each focal observation
- time: duration of each focal observation
- time_I: duration of each focal observation of the focal infant
- time_M: duration of each focal observation of the focal mother
id.csv: includes subject information
- name: unique identifier for each individual
- age: age of each individual
- sex: sex of each individual
- birthday: date of birth of each infant. Empty cells occur only for non-infant individuals, which are excluded in the analysis code.
dominance.csv: includes dominance rank of adult females
- winner: unique identifier for the winner of the dominance–submission interaction
- loser: unique identifier for the loser of the dominance–submission interaction
relatedness.csv: includes relatedness between individuals
- subject: unique identifier for each mother
- partner: unique identifier for individual
- relatedness_sibling: relatedness between 'subject' and 'partner', assuming that the three females with uncertain kinship are siblings
- relatedness_nonkin: relatedness between 'subject' and 'partner', assuming that the three females with uncertain kinship are nonkin
grooming.csv: includes grooming events between adult females
- session: unique identifier for each focal observation
- subject: unique identifier for each mother
- partner: unique identifier for individual who groomed with 'subject'
- grooming: number of 1-minute interval records in which 'subject' and 'partner' were grooming
mother_infant_interaction.csv: includes mother-infant interaction to estimate maternal style
- session: unique identifier for each focal observation
- infant: unique identifier for each focal infant
- category: behavioral category of mother-infant interaction
- time: time of each mother-infant interaction
proximity_mother_infant.csv: includes the time spent in maternal contact and carrying
- session: unique identifier for each focal observation
- infant: unique identifier for each focal infant
- date: date for each focal observation
- MCA: number of 1-minute interval records in which 'mother' carried 'infant'
- BC: number of 1-minute interval records in which 'mother' and 'infant' were in physical contact
- time: duration of each focal observation
- time_I: duration of each focal observation of the focal infant
- time_M: duration of each focal observation of the focal mother
handling_reaction.csv: includes events of infant handling and subsequent reactions
- session: unique identifier for each focal observation
- date: date for each focal observation
- handler: unique identifier for each handler
- mother: unique identifier for each focal mother
- infant: unique identifier for each focal infant
- time: time of each handling event
- handling_start: start time of each handling event. Empty cells are treated as missing values and removed in the analysis code.
- handling_end: end time of each handling event. Empty cells are treated as missing values and removed in the analysis code.
- handling: behavioral form of each handling event
- positive: whether the infant showed a positive reaction following each handling event
- squirming: whether the infant showed a squirming reaction following each handling event
- twitching: whether the infant showed a twitching reaction following each handling event
- leaving: whether the infant showed a leaving reaction following each handling event
- crying: whether the infant showed a crying reaction following each handling event
- aggression: whether the infant showed an aggressive reaction following each handling event
For variables 10–15, empty cells indicate that the corresponding reaction did not occur. During the analysis, only cells with a value of 1 were used.
analysis.R: R script for analyses and figure generation
Before running the code in the "analysis.R" file, please download all CSV files and update the file path in the “read data files” section to match the location where the files are stored.
Descriptions of variables of GLMMs
infant properties: infant weekly age and sex
("I age in week" and "I sex" in GLMM)
maternal style ("M protectiveness" and "M rejectiveness" in GLMM):
This variable serves as an indicator of maternal style of each mother. Based on the frequencies or time proportions of maternal contact, restraint, rejection, and carrying behaviors, a principal component analysis (PCA) identified key dimensions of maternal protectiveness and rejectiveness. Each mother was assigned a protectiveness score and a rejectiveness score.
handling form ("H handing form" in GLMM):
This variable classifies handling types into rough and non-rough. Aggression, rejecting, crushing, lifting, shaking, dragging, pulling, and grasping are classified as rough, whereas all other handling types are classified as non-rough.
handler-mother relatedness ("HM relatedness" in GLMM):
This variable assigns a relatedness coefficient of 0.5 to the parent-offspring relationship and calculates the relatedness coefficient between each handler-mother dyad.
handler-mother relative rank ("HM relative rank" in GLMM):
Female ordinal ranks were determined using David's scores, calculated from win-loss records of dominance-submission interactions among females. This variable indicates whether, within each handler-mother dyad, the handler holds a higher or lower ordinal rank than the mother.
handler-mother grooming time proportions ("HM grooming time %" in GLMM):
This variable represents the proportion of the mean grooming time for each handler-mother dyad, calculated from instantaneous sampling data of grooming interactions among females recorded at one-minute intervals.
A warning message indicating a model convergence problem appears when glmmTMB is executed on lines 344-368 of the analysis.R file. This warning only occurs for the model analyzing aggression reactions displayed by infants toward handlers; other models run without issues.