Frog Colouration

You are a visual ecologist studying warning colouration and mimicry in poison dart frogs. Your study system is the toxic Ameerega bilinguis and the non-toxic Allobates zaparo, a pair of sympatric terrestrial frog species native to the Ecuadorian Amazon. Ameerega bilinguis utilizes a multi-component warning signal, with a red dorsum, bright yellow limb-pit spots, and a bright blue belly. Allobates zaparo has evolved to mimic this colouration, but exhibits ‘imperfect mimicry’ – the exact quantitative properties of the color components are not perfectly matched. You collected 20 adult individuals of each species from the field and took color-calibrated photographs of each of their body regions. You then used the micaToolbox in ImageJ to simulate avian vision and compute the strength of the chromatic (i.e. hue) and achromatic (i.e. brightness) contrast of each of four color components (Front Spots, Back Spots, Dorsum, Venter) against a natural leaf-litter background, for each species. Visual contrast values are presented in the unit of ‘JND’, or ‘Just Noticeable Difference’, where higher values indicate that the color patch contrasts more strongly with the background. In other words, higher values mean that signal component is more conspicuous.

1. Load the data. Create histograms for chromatic and achromatic contrast each component (Front Spot, Back Spot, Dorsum, Venter) with both species’ data laid on the same panel (2 total figures; 4 panels per figure, 16 total histograms).
   • Use the facet_wrap() layer in ggplot2 to create a separate panel for each colour region in the same figure
   • Make the model Am. bilinguis’ fill turquoise, and the mimic Al. zaparo’s fill red.
   • Make both species’ fills semi-transparent so any potential distribution overlap is apparent
   • Label the x-axis “Color Contrast (JND)” and “Luminance Contrast (JND)” respectively

2. Using the ddply() function (package: “plyr”) to create a summary dataframe for each of color and luminance contrast values, with columns of Species, Component, and n, as well as mean JND, sd, se, Ci.lwr, and Ci.upr values for both chromatic and achromatic contrast

3. Using the summary dataframe, create a figure visualizing the mean chromatic (x-axis) and achromatic (y-axis) contrasts of each species’ signal components
   • Use the ggplot package to create a scatter plot
   • Color the dots by species, with the model in cyan and mimic in red
   • Separate the colour regions by dot shape
   • Create a vertical dashed line and a horizontal dashed line, each with intercept = 3
   • Make the x and y scale identical, to show relationship between achromatic and chromatic contrast values
   • Save the plot as “AvianBackgroundContrasts”

4. Perform a pair of factorial anovas (1 for each chromatic and achromatic contrast) to determine whether background contrast is affected by signal component, species, or an interaction between the two.
   • Using the histogram figures as guides, compute follow-up LSD t-tests between model and mimic for regions that appear to differ in their background contrast by species. Report directional differences, i.e. directional imperfect mimicry by zaparo.

Files to download:

To download, right-click and press “Save File As” or “Download Linked File”

1. FrogJNDs.csv

Laboratory and Institution or PI

Behavioural and Sensory Ecology Lab; Dr. James Barnett @ Trinity College Dublin

https://www.jbbarnett.co.uk/

References and Further Reading

McEwen, B.L., Yeager, J.D., Kinley, I., Anderson, H.M., Barnett, J.B.B. (Under Review). Body posture and viewing angle modulate detectability and mimic fidelity in a poison frog system.

Darst, C. R., & Cummings, M. E. (2006). Predator learning favours mimicry of a less-toxic model in poison frogs. Nature, 440(7081), 208-211.

Stevens M, Párraga CA, Cuthill IC, Partridge JC, Troscianko TS. (2007). Using digital photography to study animal coloration. Biol. J. Linn. 90, 211-237