Skip to main content
Dryad

Survival data and code: Camouflage using 3D surface disruption

Cite this dataset

Kelley, Jennifer; Hemmi, Jan; King, Jemma (2023). Survival data and code: Camouflage using 3D surface disruption [Dataset]. Dryad. https://doi.org/10.5061/dryad.7wm37pvxr

Abstract

Disruptive markings are common in animal patterns and can provide camouflage benefits by concealing the body’s true edges and/or by breaking the surface of the body into multiple depth planes. Disruptive patterns that are accentuated by high contrast borders are most likely to provide false depth cues to enhance camouflage, but studies to date have used visual detection models or humans as predators. We presented 3D-printed moth-like targets to wild bird predators to determine whether: (1) 3D prey with disrupted body surfaces have higher survival than 3D prey with continuous surfaces, (2) 2D prey with disruptive patterns or enhanced edge markings have higher survival than non-patterned 2D prey. We found a survival benefit for 3D prey with disrupted surfaces, even after accounting for luminance differences among the treatments. There was no evidence that false depth cues provided the same protective benefits as physical surface disruption in 3D prey, perhaps because our treatments did not mimic the complexity of patterns found in natural animal markings. Our findings indicate that disruption of surface continuity is an important strategy for concealing a 3D body shape.

Methods

These data and supporting R code describe differences in survival among 3D-printed targets for six camouflage treatments: 3D non-patterned, 2D non-patterned, 2D disruptive, 2D enhanced edge and 3D physical surface disruption (two versions). Targets were pinned on trees (with a mealworm as bait) in April/May 2021 in southwest Western Australia. Targets were checked for signs of predation (mealworm missing) 3, 5 and 19 hours after placement. A subset of targets was photographed for avian visual modelling, allowing measurement of the luminance of targets in situ.

Usage notes

Analyses run on R version 4.1.0 (2021-05-18), macOS 12.6 and RStudio 2022.07.1, Build 554.

Funding

Australian Research Council, Award: FT180100491