Data from: Delineating connectivity and quality of peer–peer prepubescent rhesus macaque (Macaca mulatta) relationships, by examining coupled social behaviors
Data files
Oct 27, 2025 version files 458.26 KB
Abstract
Infant development is relatively extended in primates, with early life experiences having lifetime implications for individuals’ social competence and circumstances. Therefore, lifelong trajectories can be informed by a nuanced understanding of who individuals connect with (i.e., connectivity) and how invested they are in those connections (i.e., quality) during development. Though a simple premise, in principle, the practice of examining social connectivity and quality relies on a nuanced understanding of how prepubescent monkeys shift their behavioral repertoire over time. Such shifts involve the same putative partners that may recur over time. We focused on measuring peer-peer relationships through the first three years of life for two birth cohorts of 49 rhesus macaques (Macaca mulatta), while in their large outdoor housed mixed-sex home group. We recorded five social behaviors, allocated into monthly bins, and then built multiplex temporal networks to concatenate these bins into four temporal partitions. We then examined the auto- and cross-correlations of five social behaviors using a multivariate multiple response time series model to understand the behavioral dynamics of relationship connectivity and quality. We reinforced known principles of relationship formation by examining the effect of ego and alter traits, including demographics and assortativity. Coupled dynamics suggest that proximity broadly subsumes social dynamics (i.e., inclusive of aggression), while contact subsumes prosocial dynamics (i.e., exclusive of aggression). Directed behaviors did not show evidence of concordance with each other across time. Overall, these results highlight the dynamic nature of social development across multiple behaviors, underscoring how early social choices shape the formation, stability, flexibility, and long-term maintenance of relationships.
Alexander J. Pritchard1, 2, Tyler R. Bonnell3, Bidisha Chakraborty1, Rosemary A. Blersch1, 2, Brenda McCowan1, 2, Jessica J. Vandeleest1, 2
1 Neuroscience and Behavior Unit, California National Primate Research Center, University of California Davis, Davis, CA, USA
2 Department of Population Health & Reproduction, School of Veterinary Medicine, University of California Davis, Davis, CA, USA
3 Department of Computer Science, University of Calgary, Calgary, AB, Canada
Article DOI: https://doi.org/10.1098/rsos.250860
Pritchard AJ, Bonnell T, Chakraborty B, Blersch RA, McCowan B, Vandeleest JJ Delineating connectivity and quality of peer–peer prepubescent Rhesus macaque (Macaca mulatta) relationships, by examining coupled social behaviours. R. Soc. Open Sci. ;12: 250860.
Dataset DOI: 10.5061/dryad.ncjsxkt71
Description of the data and file structure
We focused on measuring peer-peer relationships through the first three years of life for two birth cohorts of 49 rhesus macaques (Macaca mulatta), while in their large outdoor housed mixed-sex home group. We recorded five social behaviors during focals of interactions with other enclosure members, allocated into monthly bins, and then built multiplex temporal networks to concatenate these bins into four temporal partitions. We then examined the auto- and cross-correlations of five social behaviors using a multivariate multiple response time series model to understand the behavioral dynamics of relationship connectivity and quality. We reinforced known principles of relationship formation by examining the effect of ego and alter traits, including demographics and assortativity.
Files and variables
File: Pritchard-et-al_2025_Analyze-Data.Rmd
Description: Code to execute the multivariate response model. Be aware that the model will take a decent amount of time to run and can be computationally intensive on older or smaller machines.
File: Pritchard-et-al_2025_GAMM-Plots.Rmd
Description: This code executes numerous gamm models that include random effects for the sole purpose of plotting developmental trends over time.
File: Pritchard-et-al_2025_Posterior-Network-Checks.Rmd
Description: This code must occur after obtain the model from the ...Analyze-Data.Rmd file (see above). This code compares actual degree/proportional investment measures using monthly data to predicted degree/versatility.
File: Pritchard-et-al_2025_Dataframe-For-Analysis.RDS
Description: This is dataset for the model and for posterior network checks. The data file includes the following columns in an R object dataframe:
CO: Versatility measures for contact-sitting for the relevant ego-alter dyad for the relevant ego-alter dyad and partition (year).
Cohort: A code for cohort "*1" or cohort "_*2", contingent on the year of birth for the ego focal.
Rank_Diff: A scaled continuous variable for the rank distance between the ego and their alter.
YOB_Diff: A scaled continuous variable for the year-of-birth distance between the ego and their alter.
Mat_Rank: A scaled continuous variable for the maternal rank of the ego.
BornYB_NY: A binary variable for whether the alter was born recently. (0 = No, 1 = Yes)
RemoveNY: A binary variable for whether the alter was removed recently. (0 = No, 1 = Yes)
SameSexNY: A binary variable for whether the ego and alter were the same sex. (0 = No, 1 = Yes)
Sex: Sex of the ego subject.
KinNMatSib: A factor coding whether the alter was unrelated (0), maternal kin (1), or a sibling (2) of the ego.
GR_lag: Scaled versatility measures for grooming during the ego-alter dyad's previous partition (year).
PX_lag: Scaled versatility measures for proximity during the ego-alter dyad's previous partition (year).
CO_lag: Scaled versatility measures for contact-sitting during the ego-alter dyad's previous partition (year).
AG_lag: Scaled versatility measures for aggression during the ego-alter dyad's previous partition (year).
DP_lag: Scaled versatility measures for play during the ego-alter dyad's previous partition (year).
Year: A code for the time partition of the relevant ego-alter dyad, with '2' signifying month 6-12, '3' signifying the second year of life, and '4' signifying the third year of life. '1' is excluded because there can be no lags prior to birth.
Focal: Animal codes for each of the ego subjects.
IDs: Animal codes for each of the alter subjects.
DP: Binary measures (0 = Absence, 1 = Presence) for (directed) play for the relevant ego-alter dyad and partition (year).
PX: Versatility measures for proximity for the relevant ego-alter dyad for the relevant ego-alter dyad and partition (year).
AG: Binary measures (0 = Absence, 1 = Presence) for aggression for the relevant ego-alter dyad and partition (year).
GR: Binary measures (0 = Absence, 1 = Presence) for grooming for the relevant ego-alter dyad and partition (year).
File: Pritchard-et-al_2025_Data-Rates-For-GAMM-and-PreNetChecks.RDS
Description: This is dataset for the GAMM figures and for posterior network checks. The data file includes the following columns in an R object dataframe:
Focal: An identifier for the relevant ego focal (animal subject)
Month: An identifier for the calendar month since the relevant ego focal's birth
Behav: A behavioral code for aggression (AG), contact-sitting (CO), grooming (GR), play (DP), or proximity (PX).
Cohort: A code for cohort 1 or cohort 2, contingent on the year of birth for the ego focal.
ID_Foc: The animal initiator for the relevant behavior (directionally important for AG, GR, and DP).
ID_Rec: The animal recipient for the relevant behavior (directionally important for AG, GR, and DP).
Count: Count of the animal behavior in the given time bin (month).
Focals: Count of the focal observations for the relevant month and ego.
Intervals: Count of focal sampling intervals for the relevant month and ego.
Rate: as the count of behavior divided by the count of intervals.
Code/software
The data and code are accessible using 'R'. The following packages are required:
brms, tidybayes, mgcv, tidybayes, dplyr, tidyverse, bayesplot, ggplot2, ggridges, ggpubr
