Behavioural choice and background matching facilitate camouflage in the European garden spider
Data files
Aug 01, 2025 version files 110.45 KB
-
background_matching_data.xlsx
50.28 KB
-
colour_change_rnl_data.xlsx
31.86 KB
-
README.md
8.78 KB
-
script_spider.R
19.53 KB
Abstract
Visual camouflage via background matching involves a variety of adaptive traits to maintain crypsis, including intraspecific color variation, behavioral choice of substrates, and color change. These non-mutually exclusive solutions frequently act together to conceal and deceive prey and enemies. Here, we combine field observations, image analysis, and laboratory experiments to investigate which processes drive camouflage in the garden cross spider (Araneus diadematus), a species with body coloration that varies in shades of brown. We demonstrate that A. diadematus does not change color significantly, at least within the same instar, when retained on substrates of different coloration. However, there is strong behavioral selection through active substrate choice across spiders for color-matching substrates (dead brown leaves over green leaves) when offered a choice under laboratory conditions. Similar background selection also apparently occurs in nature, where spiders were often observed on brown leaves, even though they are less common than green ones. In general, vision modeling shows that there is a high overlap in the diversity of brown shades (from pale to dark) between spider bodies and dead leaves available in the environment. Image analyses also revealed that spiders fine-tune their camouflage on an individual level by matching the tones (from pale to dark) of their host leaves. Therefore, we demonstrate how behavior coupled with variation in color phenotypes facilitates camouflage at different scales.
https://doi.org/10.5061/dryad.6wwpzgn9j
Authors: Yuri Fanchini Messas; Rafael Campos Duarte; João Vasconcellos-Neto; and Martin Stevens
Contact: Yuri Fanchini Messas, yurimessas@gmail.com
Abstract
Here, we used image analysis and visual modelling of an avian predator (the Eurasian Blue Tit Cyanistes caeruleus) to assess the color variation and camouflage of European garden spiders (Araneus diadematus) inhabiting shrub vegetation in southwest UK. We predicted that spiders would be more effectively camouflaged on their chosen resting substrates (brown host leaves) compared to other substrates available in the surrounding environment. Additionally, we hypothesize that A. diadematus background matching relies on phenotypic plasticity in spiders’ body coloration, with a fine-scale individual resemblance to the color of the host leaf, or of behavioral selection of backgrounds resembling an individual’s own appearance. Herein, we tested whether individual spider camouflage in the field better matches the chosen host leaf than the leaves chosen by sympatric conspecifics. To understand how camouflage is achieved, we tested in the laboratory whether background matching of A. diadematus is mediated by one or a combination of habitat choice by active behavioral selection of brown leaves on shrub vegetation and color change according to substrate coloration. We demonstrate that A. diadematus does not change color significantly, at least within the same instar. However, there is strong behavioral selection through active substrate choice across spiders for color-matching substrates. In general, our results evidenced that there is a high overlap in the diversity of brown shades (from pale to dark) between spider bodies and dead leaves available in the environment to predator vision. Spiders can also fine-tune their camouflage on an individual level by matching the tones (from pale to dark) of their host leaves.
Date of data collection
March 2022 - August 2022
Location
A coastal trail between Castle (50°08'54" N, 05°03'26" W) and Maenporth (50°7'28" N, 05°05'36" W) beaches in Falmouth, Cornwall, Southwest UK.
Responsible for field collections
Yuri Fanchini Messas
Responsible for writing the code
Rafael Campos Duarte
List of files
- scripts: script_spider.R;
- datasets: background_matching_data.xlsx; colour_change_rnl_data.xlsx
The script contains the code for all analyses and figures of the study.
Session info
R version 4.4.0 (2024-06-14 ucrt) -- Platform: x86_64-w64-mingw32/x64 (64-bit)
Packages installed and their versions in R 4.4.0
colorspace v.2.1-1
emmeans v.1.10.5
lme4 v.1.1-35.5
lmerTest v.3.1-3
munsell v.0.5.1
patchwork v.1.3.0
readxl v.1.4.3
rstatix v.0.7.2
tidyverse v.2.0.0
General information about the scripts and datasets
The script and datasets refer to the analyses of the (i) color variation of European garden spiders (Araneus diadematus), and fresh green/dead brown leaves of their host plants; (ii) color variation of brown shades (from pale to reddish and dark) of spiders and brown leaves available in nature; (iii) camouflage of spiders against their brown host leaves, and against the mean of brown host leaves of conspecific individuals; (iv) behavioral choices by spiders between green and brown leaves of their host plants; and (v) colour change of spiders against brown and green substrates. The repository consists of three files, including one code script and two Excel files. The script contains the code to run both the statistical analyses and the graphics of the manuscript.
Description of the datasets and variables
DATASETS
background_matching_data.xlsx
This dataset contains two spreadsheets. The first spreadsheet describes the luminance, colour, and saturation using the RNL chromaticity space of spiders (Araneus diadematus) and the green and brown leaves of their host plants, as well as brown leaves of surrounding vegetation, as viewed by the Blue Tit (Cyanistes caeruleus) in the receptor noise-limited (RNL) chromaticity color space. The first spreadsheet also shows the spider coloration in the human-based RGB color space. The second spreadsheet contains the Blue Tit visual discrimination (as color and luminance JNDs) of spiders against their brown host leaves and the mean of brown host leaves of conspecific individuals.
Column names and description:
rnl_data
image: the code used to identify each photograph
spider_id: the identification code of each spider individual or leaf from each photograph, empty cells correspond to brown leaves of the surrounding vegetation
batch: the ID of the subset of analyzed images
roi: the region of interest (ROI) consisted of a fresh green leaf (g), a brown dry leaf (b), or a spider (a)
lum_mean: mean luminance of ROI (relative units, range 0–1)
lum_dev: standard deviation of luminance (relative units, range 0–1)
X_mean: mean x-axis chromaticity (RNL color space; green-red variation)
X_dev: standard deviation of x-axis chromaticity (RNL color space)
Y_mean: mean y-axis chromaticity (RNL color space; yellow-blue variation)
Y_dev: standard deviation of y-axis chromaticity (RNL color space)
saturation_mean: mean saturation
saturation_dev: standard deviation of saturation
R: red channel (RGB values, 0–255)
G: green channel (RGB values, 0–255)
B: blue channel (RGB values, 0–255)
jnd
spider: the code used to identify each spider
site: the spider substrate; 'local' is the brown host leaf where we found the individual, 'general' is the mean of brown host leaves of conspecific individuals
lum: Blue Tit luminance discrimination (JND units)
colour: Blue Tit color discrimination (JND units)
colour_change_data.xlsx
This dataset contains two spreadsheets. The first describes the luminance and colour change using the RNL chromaticity space of brown European garden spiders (Araneus diadematus) maintained for 14 days in green or brown substrates. The second spreadsheet contains the calculated color JND values of spiders against green or brown substrates over the experimental time.
Column names and description:
colour_change
image: the photo ID
spider_id: the identification code of each individual
day: days since the start of the experiment (days)
substrate: the color of the substrate
roi: the region of interest (ROI) consisted of a fresh green leaf (g), a brown dry leaf (b), or a spider (a)
lum_mean: mean luminance of ROI (relative units, range 0–1)
lum_dev: standard deviation of luminance (relative units, range 0–1)
X_mean: mean x-axis chromaticity (RNL color space)
X_dev: standard deviation of x-axis chromaticity (RNL color space)
Y_mean: mean y-axis chromaticity (RNL color space)
Y_dev: standard deviation of y-axis chromaticity (RNL color space)
saturation_mean: mean saturation
saturation_dev: standard deviation of saturation
jnd_change
spider_id: the identification code of each individual
day: days since the start of the experiment (0, 7, or 14)
substrate: the color of the substrate (brown or green)
jnd: Blue Tit color discrimination (JND units)
Note:
All luminance values are expressed in relative units normalized between 0 and 1.
Chromaticity (X, Y) and saturation are unitless values in the RNL color space.
JND values are unitless and represent visual discriminability for Blue Tit vision.
RGB values range from 0 to 255.
Time is measured in days.
SCRIPT
script_spider.R
Contains the analyses and the codes to create: (i) Figure 2 showing (A) the color variation of European garden spiders and brown/green shrub vegetation as viewed by a Blue tit predator in the RNL color space, (B) the variation in the color of spiders and brown leaves that are available in the wild as viewed by a Blue tit predator in the RNL color space; (ii) Figure 3 showing Blue tit chromatic and achromatic visual discrimination (as JNDs) of spiders against their host leaves (local) and the host leaves of conspecifics (general); (iii) Figure 4 showing the behavioral choice of spiders between green and brown leaves of their host plants, considering both the foraging habitat (where the spider anchored its web hub) and the host leaf (where the spider remains hidden); (iv) Figure 5 showing the color and camouflage (as color JND) change of spiders maintained for 14 days on green and brown leaf substrates.
Funding Sources
Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)
Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)
Fundação para a Ciência e Tecnologia (FCT), Ministério da Ciência, Tecnologia e Ensino Superior (MCTES)
We collected garden cross spiders (Araneus diadematus) and plant leaves (green and brown from host plants; brown from sympatric species) in the field for use in laboratory experiments. In the field study and colour change experiment, we quantified spider colour and camouflage from digital photographs using methods of image analysis and visual modelling. In the behavioural-choice experiment, we recorded spider preferences between green and brown leaves in binomial trials.