Data and code from: The functional significance of striping in rodents
Data files
May 08, 2026 version files 19.40 MB
-
all_predictor_variables.R
5.12 KB
-
Diurnal.R
2.96 KB
-
Grass.R
3 KB
-
Length.R
2.91 KB
-
mammaltrees100.nex
19.08 MB
-
Predators.R
8.98 KB
-
README.md
3.08 KB
-
RodentStripes-2025-06-30.csv
276.40 KB
-
Social.R
2.89 KB
-
SpOverlap-magnitude.R
2.98 KB
-
SpOverlap-presence.R
2.97 KB
-
stochastic_character_mapping.R
894 B
-
target_variables_only.R
4.36 KB
-
Terrestrial.R
3.08 KB
Abstract
Rodents constitute a large proportion of extant mammal species and have a uniform brown-gray coloration to avoid detection by predators. A minority, however, have longitudinal dorsal stripes, the function of which is unknown. Using a comparative approach, we explored whether striping in rodents is a form of background matching, an example of dazzle coloration, a social signal, or a signal to avoid hybridization with sympatric congeners. We found some evidence that striping is associated with small species, diurnality, and raptor and marginally with owl predation, suggesting it could be a form of dazzle coloration interfering with accurate interception by aerial predators approaching from above. There was no evidence that stripes are used in communication between conspecifics or for the avoidance of hybridization. Our study provides the beginning of a functional underpinning to developmental studies of pattern formation in mammals.
Data:
RodentStripes-2025-06-30.csv
Columns:
- Species: species scientific name
- PhyloName: corresponds to the full 5,991-species Upham et al. 2019 tree
- Family: family, according to the Upham et al. 2019 taxonomy
- Striped: whether the species is (1) or is not (0) striped
- Diurnal: whether the species shows diurnal activity, including diurnal-crepuscular and cathemeral species
- Species_Overlap: whether the species range overlaps with that of congeners
- Overlap_Magnitude: the number of congeners the species range overlaps
- Grassland_Savnnah: whether the species lives in grasslands or savannahs
- Social_Group_Size: social group size, with 1 = solitary, 2 = two to five individuals, 3 = six to 25 individuals, 4 = 26 to 100 individuals, 5 = >100 individuals
- Adult_Head_to_Body_Length_(mm): sex-averaged adult head-body length, in mm
- SnakeThreat: from Sheard et al. 2024, corresponding to the number of the Serpentes species whose ranges overlap with each species, weighted by degree of overlap
- OwlThreat: from Sheard et al. 2024, corresponding to the number of Strigiformes species known to eat rodents whose ranges overlap with each species, weighted by degree of overlap
- RaptorNonOwlThreat: from Sheard et al. 2024, corresponding to the number of raptor (Cathartiformes, Accipitriformes, and Falconiformes) species known to eat rodents whose ranges overlap with each species, weighted by degree of overlap
- Terrestriality-CS: from Sheard et al. 2024, corresponding to whether some or all of the species locomotory mode is terrestrial (i.e., including semiarboreal and semifossorial species as well as terrestrial specialists)
Sheard et al. 2024: https://www.journals.uchicago.edu/doi/full/10.1086/729751
Upham et al. 2019: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3000494
Tree:
mammaltrees100.nex
- 100 phylogenetic trees from Upham et al. 2019
Code:
The code for the 10 univariate models is included as separate files:
- Diurnal.R: diurnality
- Grass.R: grasslands
- Length.R: head-body length
- Predators.R: owl, diurnal raptor, and snake predation threat
- Social.R: sociality
- SpOverlap-magnitude.R: magnitude of congeneric overlap
- SpOverlap-presence.R: binary coding of congeneric overlap
- Terrestrial.R: terrestriality
The code 'all_predictor_variables.R' runs the model with all predictor variables, while 'target_variables_only.R' considers only the four variables that individually and significantly/marginally-significantly predict striping (body length, diurnality, diurnal raptor predation threat, and owl predation threat).
The code for stochastic character mapping is found in stochastic_character_mapping.R .
All code has been verified to run on R version 4.4.2, with packages ape 5.8, geiger 2.0.11, MCMCglmm 2.36, nlme 3.1-166, phangorn 2.12.1, and pictante 1.8.2.
