Dominance rank, facial morphology, and testes size in male white-faced capuchins: evidence for pre- and post-mating competition
Data files
Jun 30, 2025 version files 91.43 KB
-
Capuchin-Morphology.R
74.11 KB
-
photogrammetry_dataset_stable.csv
12.90 KB
-
README.md
4.42 KB
Abstract
Male reproductive success is determined by the interplay of female mate choice and male-male competition, often linked to dominance rank in social animals. Across taxa, elaborate ornaments, such as bright coloration or large antlers, often function as badges of status, signaling male competitive ability to rivals. In species where females mate with multiple males, post-mating sperm competition also plays an important role in male reproductive success and is associated with larger relative testes size. We investigate the relationship between morphological features and dominance rank in wild male white-faced capuchins. Using parallel-laser photogrammetry, we measured aspects of facial morphology, including facial width-to-height ratio, and testes size. We found that alpha males had significantly larger facial width-to-height ratios, wider faces, and wider scrota than subordinate males. These results suggest that facial traits potentially function as badges of status in male white-faced capuchins and may play a role in pre-mating competition and/or mate choice, while differences in scrotal size reflect adaptations for post-mating competition. This study highlights the underrecognized role of facial trait evolution in sexual selection among relatively gracile yet highly visually oriented mammals and highlights the potential variability of sexual traits in species characterized by strong reproductive skew among males.
Dataset DOI: 10.5061/dryad.rv15dv4kr
Description of the data and file structure
Files and variables
File: Capuchin-Morphology.R
Description: We first performed a Principal Component Analysis (PCA) using the principal() function from the ‘psych’ package (115) in R (v4.4.2; 116) and RStudio (v2024.9.1.394; 117) to assess the relationships among facial measurements (facial width, brow width, muzzle width, and facial height) and to check for multicollinearity. Prior to the PCA, we checked for sampling adequacy (Kaiser-Meyer-Olkin test: MSA = 0.66) and sphericity (Bartlett’s test: 𝜒2(6) = 49.75, p < 0.001) to ensure the appropriateness of our data for factor analysis. The data were standardized prior to the PCA using the scale() function, and the PCA analysis was conducted with a varimax rotation to enhance the interpretability of the principal components. The PCA identified four principal components, each accounting for approximately 25% of the total variance (Table S2). Each facial measurement loaded strongly onto a separate principal component, indicating that each measurement captured distinct aspects of morphological variation. Based on these findings, we analyzed each facial measurement independently.
Six separate linear mixed models (LMMs) were fit individually for scrotum width, facial width, muzzle width, brow width, facial height, and fWHR using the ‘lmer’ function from the ‘lme4’ package (118). Dominance rank (alpha/subordinate), body length (in mm), and age (in years) were included as fixed effects, while individual identity was included as a random effect to account for repeated measures from the same individuals. Each response variable followed a normal distribution (Shapiro-Wilks tests, p> 0.05). To further evaluate the significance of the fixed effects, we used the drop1() function to assess the contribution of each predictor to the model by comparing nested models. Conditional R2 values were estimated using the r.squaredGLMM() function from the ‘MuMIn’ package (119). Assumptions for each model were checked using the 'DHARMa’ package (120). Homogeneity of variance was assessed by examining residual vs. fitted plots, ensuring consistent variance across predictor levels. Residuals were tested for normality using visual inspection of Q-Q plots and confirmed by the Kolmogorov-Smirnov test, which indicated no deviations from normality. Additionally, no models exhibited signs of overdispersion (range across all models: 0.90 - 0.97), confirming the adequacy of the model fit.
File: photogrammetry_dataset_stable_(4).xlsx
Description: Measurements of individual photos (in mm) for each of the body parts/facial features of interest
Variables
All missing values in the dataset are represented as "na". These indicate instances where a measurement could not be taken — for example, due to poor visibility of the anatomical region in photogrammetry images or the position of the monkey. In some cases, such as facial photographs, measurements of the testicles were not possible. "na" may therefore mean "not available" or "not applicable," depending on context.
- obs: unique row index for each observation
- individual: Individual monkey ID code
- group: Social group identifier
- dominance_rank: Dominance status at time of photo (a = alpha, s = subordinate)
- date: Date the photograph was taken in YYYY.MM.DD format
- laser_type Laser: apparatus used (1 or 2) — differentiates between photogrammetry apparatuses
- Temp: Average ambient temperature in degrees Celsius
- age: Age in years (as a decimal) at time of photo
- known/estimate: Indicates whether birthdate was known or estimated
- brow_width: Width of the brow
- muzzle_width: Width of the muzzle
- facial_width: Overall facial width
- facial_length: Vertical length of the face
- scrotum_width: Width of the scrotum
- left_test_length: Length of the left testicle
- right_test_length: Length of the right testicle
- body_length: Length from the base of the neck to base of the tail
all measurements are in mm
Code/software
R software (https://www.r-project.org/)
Gimp GNU Image Manipulation Program (https://www.gimp.org)
We collected data on all males (6+ years old) residing in five habituated groups of white-faced capuchins (Cebus imitator) between January 2021 and May 2024 in the Santa Rosa Sector of the Área de Conservación Guanacaste, Costa Rica (SSR). We used parallel-laser photogrammetry (108,109) to noninvasively measure facial features, scrotal width, and body length, of male white-faced capuchins during periods of group stability. For each male, we collected multiple photographs (3 – 5) of their face, body length, and scrotum at least twice per month every three months over both dry and wet seasons. Selected photographs were uploaded in jpeg format to the GIMP GNU image manipulation program (version 2.10.34), a free and open-source image editing program (https://www.gimp.org). We applied GIMP’s standard high contrast filter to each photograph, which enhanced the visibility of the morphological features of interest due to the black and white pelage of the study species. Following Richardson et al. (2022), we used the calibrated inter-laser distance (30.00 or 45.00 mm depending on the laser used for the photograph) divided by the inter-laser distance from the image in pixels (calculated using the ‘measure tool’ in GIMP), to create a millimeter per pixel scale (112). We measured four facial features (Figure 1): a) brow-width – i.e., the lateral-medial width of the supraorbital torus, b) facial height – i.e., the height of the face from between the eyes to the middle of the upper lip, c) muzzle width – i.e., the lateral width of the upper snout, measured across the maxilla just above the upper lip and d) facial width – i.e., the width from the most lateral points of each side of the mandible, which is the widest visible portion of a capuchin male face. From these measurements, we calculated the fWHR by dividing facial width by facial height.