Eyespot peek-a-boo: Leaf rolls enhance the antipredator effect of insect eyespots
Data files
Dec 16, 2024 version files 920.58 KB
-
LRpred.csv
250.84 KB
-
LRspectra.csv
614.38 KB
-
LRsurvey.csv
46.69 KB
-
README.md
8.67 KB
Abstract
Animal color patterns are often accompanied by specific, synergistic behaviors to most effectively defend against visual predators. Given the inherent context-dependence of color perception, understanding how these color-behavior synergies function in a species’ natural environment is crucial.
For example, refuge-building species create a unique visual environment where most (or all) of the body is obscured unless closely inspected. How these built environments affect the perception of defensive color patterns by predators is not well understood.
Using artificial caterpillars that resemble a refuge-building species with conspicuous markings (Papilio troilus; Lepidoptera: Papilionidae), I tested the hypothesis that leaf rolls amplify the antipredator effect of this species’ eyespots. I compared wild avian predation rates on 659 artificial swallowtail-like caterpillars from four treatment groups: eyespotted and non-eyespotted, and presented in leaf rolls or on open leaves of live host plants.
In support of my hypothesis, eyespots only reduced predation for larvae in leaf rolls. On open leaves, eyespots had no antipredator effect. I also found that leaf rolls reduced predation in general for both eyespotted and non-eyespotted prey.
These results highlight the importance of considering relevant behaviors in studies of animal coloration whenever possible, including behaviors that influence color perception indirectly (e.g. through habitat use or modification).
README: Eyespot peek-a-boo: Leaf rolls enhance the antipredator effect of insect eyespots
Dates of study:
- predation trials (07/03/2021-07/30/2021)
- leaf roll surveys (08/23/2022-08/27/2022)
Contributors: Elizabeth G. Postema
Data files:
LRpred.csv: a csv file containing data from the artificial prey field experiment.
LRspectra.csv: a csv file containing spectral measurements for live/artificial swallowtails and host plant leaves.
LRsurvey.csv: a csv file containing data from surveys of naturally-occurring leaf rolls.
LRpredation.html: full analysis of artificial prey field experiment in R.
LRspectra.html: visual modeling of live/artificial swallowtails and host plant leaves in R using package pavo.
LRsurvey.html: visualizing data from roll surveys in R.
JAE_2024_00285_supp_materials.docx: a word document with supplemental tables and figures associated with the manuscript.
Description of data columns (LRpred.csv):
id: unique identity of each plant used in the study.
treatment: treatment group of a specific artificial prey, based on whether prey was in a leaf roll/on an open leaf, had eyespots/no eyespots, and was oriented up or down. OE = open with eyespots, RE = rolled with eyespots, RNE = rolled with no eyespots, ONE = open with no eyespots, RE-Up = rolled with eyespots oriented up, RE-Down = rolled with eyespots oriented down.
trial: which of the two predation trials (1 or 2) the data in this row were collected from.
rolled: whether artificial prey were in leaf rolls (rolled) or presented on open leaves (open).
color: whether artificial prey had eyespots (eyespots) or no eyespots (none).
orientation: whether artificial prey was oriented up (up) or down (down). NAs represent artificial prey where orientation was not specifically manipulated.
roll.opened: whether artificial leaf rolls was opened by a predator. Only applies to rolled prey. 0 = not opened, 1 = opened.
attacked: whether artificial prey was visibly attacked by a predator. 0 = not attacked, 1 = attacked.
pred.id: taxonomic identity of the predator that attacked the artificial prey, based on attack mark morphology. b = bird, m = mammal, u = unknown, NA = prey not attacked.
found: whether the artificial prey was recoverable. y = prey recovered at end of trial, n = prey not recovered at end of trial.
coords: geographic coordinates of the unique host plant, in decimal degrees (latitude, longitude).
lat: latitude of the unique host plant, in decimal degrees.
long: longitude of the unique host plant, in decimal degrees.
deployed.date: day that the artificial prey was deployed (day-month format).
collect.date: day that the artificial prey was collected (day-month format).
hour.deployed: hour that the artificial prey was deployed, in standard time (HourMinute format).
hour.collect: hour that the artificial prey was collected, in standard time (HourMinute format).
exposed.hours: total number of hours that the artificial prey spent exposed to predators.
exposed.days: total number of days that the artificial prey spent exposed to predators.
location: general site where the unique plant was located; Bird Hills Nature Area (Bird Hills) or Nichols Arboretum (Nichols Arboretum).
sublocation: more specific site where the unique plant was located. Within Bird Hills = mixed gully, sassafras patch, tulip ridge. Within Nichols Arboretum = hawthorn valley, heathdale.
host.species: species identity of unique plant. Either tulip tree (Liriodendron tulipifera) or sassafras (Sassafras albidum).
shape.leaf: shape of the leaf that the artificial prey was presented on/in. Either lobed (lobed) or ovate (unlobed).
length.leaf: length of the presentation leaf, in centimeters.
width.leaf: width of the presentation leaf, in centimeters.
roll.length: for artificial prey in leaf rolls, the length of the roll, in centimeters.
roll.prop: for artificial prey in leaf rolls, the length of the leaf roll divided by the total length of the presentation leaf.
lengthwidth.leaf: the length of the presentation leaf multiplied by the width of the presentation leaf.
area.leaf: the area of the presentation leaf, in centimeters^2.
height.plant: the height of the unique plant (from ground to top), in centimeters.
prey.height: the height of the artificial prey from the ground, in centimeters.
height.prop: the relative height of the artificial prey (height.prop = prey.height/plant.height).
diameter.1: diameter of the unique plant, in centimeters. Perpendicular to diameter.2.
diameter.2: diameter of the unique plant, in centimeters. Perpendicular to diameter.1.
radius.1: diameter.1 divided by two (in centimeters).
radius.2: diameter.2 divided by two (in centimeters).
plant.area: area, expressed as an oval, encompassing the unique plant in centimeters^2.
open.prop: proportion of the canopy above the unique plant that is open sky.
covered.prop: proportion of the canopy above the unique plant that is covered in vegetation(1 ? open.prop).
canopy.status: whether canopy openness was successfully calculated (complete), missing (missing), or un-measurable due to lens flare (lens flare).
data.complete: whether columns describing host plant and presentation leaf measurements are complete or incomplete.
inclusion: whether to include the data in the analysis or not (include, exclude).
prev.treatment: documents the previous artificial prey treatment group (OE, RE, RNE, ONE) for artificial prey in trial 2 that re-use a unique host plant.
notes: any notes and/or data corrections.
Description of data columns (LRspectra.csv):
wavelength.nm: light wavelength, in nanometers.
modelgreen1 - modelgreen3: % reflectance of artificial caterpillar body for a given wavelength, 3 measurements.
realgreen1 - realgreen3: % reflectance of live caterpillar body for a given wavelength, 3 measurements.
topleaf1a - topleaf10c : % reflectance of 10 sassafras leaves (top) for a given wavelength, 3 samples per leaf.
bottomleaf1 - bottomleaf3: % reflectance of sassafras leaf (bottom) for a given wavelength, 3 measurements.
modelyellow1 - modelyellow3: % reflectance of artificial caterpillar eyespot for a given wavelength, 3 measurements.
realyellow1 - realyellow3: % reflectance of live caterpillar eyespot for a given wavelength, 3 measurements.
Description of data columns (LRsurvey.csv):
plant.num: unique numerical identity of a plant being surveyed.
transect: which transect (1-5) the plant was found on.
sublocation: specific location of the plant (mixed gully, sassafras patch, tulip ridge, hawthorn valley, heathdale).
date: date the plant was surveyed on (month-day-year format).
roll.num: the individual leaf roll being surveyed (1, 2, or 3). NR = no rolls present on the surveyed plant.
status.code: 0 = leaf roll is not occupied by something living, 1 = leaf roll is occupied by something living. NR = no rolls present on the surveyed plant.
status: unoccupied = leaf roll is not occupied by something living, occupied = leaf roll is occupied by something living. NR = no rolls present on the surveyed plant.
contents: the organism or type of debris present in the leaf roll. Nothing in the roll = empty. NR = no rolls present on the surveyed plant.
plant.type: whether the surveyed plant is a host plant of Papilio troilus (host plant) or not (non-host plant).
flight.behavior: whether the contents of the leaf roll jump, drop, or do nothing (none) in response to the roll being disturbed/opened.
notes: any notes and/or corrections.
Description of Supplemental Matierals (JAE_2024_00285_supp_materials.docx):
Table S1: Visual/aural surveys of bird species present at the study sites (7/14/2021 & 7/30/2021)
Table S2: Observations and measurements of live spicebush swallowtail larvae and their host plants at the study sites, with dates for all observations.
Table S3: Notes on the behavior of live spicebush swallowtail larvae in response to disturbance, observed throughout the main predation experiment.
Figure S1: A map of the locations used in the field predation trials, with symbols for different host plant species and the locations of live spicebush swallowtail larvae. Generated with Google My Maps and LRpred.csv.
Figure S2: A map of the locations of observed predator (avian, mammalian, and unknown) attacks on artificial prey. Generated with Google My Maps and LRpred.csv.
Figure S3: Proportions of artificial prey attacked by mammal and unknown predators. Figures created in R (LRpredation.html).
Figure S4: Method and results of surveys of naturally-occurring leaf rolls at the study sites. Figures created in R (LRsurvey.html)
Methods
This project includes 3 datasets:
- LRpred.csv: data from an artificial prey field experiment using hand-painted clay caterpillars (2021).
- Beyond treatment (color and roll vs no roll), this dataset also includes a considerable number of plant- and habitat-related variables.
- LRspectra.csv: spectral measurements for live and artificial swallowtails and host plant (Sassafras) leaves.
- Spectral data were collected with an Ocean Optics Flame Miniature (FLAME-S-UV-VIS-ES) spectrometer with Ocean Optics PX-2 Pulsed Xenon light source, calibrated with a 99% Labsphere reflectance standard at the University of Michigan Museum of Zoology.
- LRsurvey.csv: data from surveys of naturally-occurring leaf rolls along transects at the same site as the field experiment (2022).
All data were originally collected in google spreadsheets and downloaded as CSVs. I then cleaned, organized, and analyzed the data in R. All corresponding R scripts are available in the repository.
The repository also contains supplemental materials associated with the manuscript (JAE_2024_00285_supp_materials.docx), including 3 supplementary tables of observational data, 3 supplemental figures generated from LRpred.csv, and 1 supplemental figure generated from LRsurvey.csv.