Supplementary data for article: Camouflage accuracy in Sahara-Sahel desert rodents

1 Supplementary data for Nokelainen et al. 2020 Journal of Animal Ecology Title: Camouflage accuracy in Sahara-Sahel desert rodents Authors: Ossi Nokelainen1*, José Carlos Brito2, Nicholas E. Scott-Samuel3, Janne K. Valkonen1 & Zbyszek Boratyński2 1Department of Biological and Environmental Science, University of Jyväskylä, P.O. Box 35, 40014 University of Jyväskylä, Finland 2CIBIO-InBIO Associate Laboratory, Research Center in Biodiversity and Genetic Resources, University of Porto, Vairão, 4485-661, Portugal 3School of Psychological Science, University of Bristol, Bristol, UK * Authors for correspondence ([email protected], [email protected]) Copyright: All authors listed above and University of Jyväskylä (Finland). Published by the British Ecological Society. Licence: Creative Commons Attribution 4.0 International (CC-BY 4.0) Contents: Page Supplementary tables Table S1 2 Table S2 3 Table S3 4 Supplementary figures Figure S1 6 Figure S2 7 Supplementary references 8 2 Table S1: Background match estimation using Euclidian distances To illustrate the spread of the colour values in animals and backgrounds, using another way than nonparametric colour map approach reported in the main text, we converted normalised camera responses (i.e. raw RGB-values) to two-dimensional XY-colour space. The conversion was done following Kelber et al. 2003. After conversion of RGB values to XY-space distance determination was calculated using D = √((x animal – x background)2 + (y animal – y background)2); as in Nokelainen et al. 2018). To compare background matching at different scales using XY-coordinates, animal-to-background distances were tested against different spatial scales. (McLachlan and Ladle 2011, Todd et al. 2012, Nokelainen et al. 2017). Average distances were plotted in relation to the original habitat from where the animal was found (i.e. focal), appearance of predominant background types (i.e. the predominant habitat types where animals were found were subjectively classified as: clay (n = 44), sand (n = 78), gravel (n = 18) or rocks (n = 8)), and to an average of colour variation across all background types (i.e. global). Animal-to-background distances using Euclidian distances were generally very low (range = 0.01 – 0.19, mean = 0.03, s.d. = 0.03) indicating very accurate background match (ESM Figure 1). Overall, match to predominant background types and global average were better than match to focal habitats. This was especially so in Gerbillus gerbillus (but also in Meriones libycus). Interestingly, there were species-background interaction (F2,30 = 4.23, P < 0.001, see further details from ESM Table S3). The results are in line with the ‘similarity-to-background analysis’, an extension of the multispectral imaging tool box in image J (Troscianko and Stevens 2015): Quantitative Colour and Pattern Analysis framework (Van Den Berg et al. 2019). 3 Table S2: Overall correlation analysis of animal colour and pattern compared to background habitat. Subscript refers to region of interest: FUR = dorsal fur coloration, BG = background against which animal was found from. Sample size in each comparison is 163 including all individuals from all species sampled. Colour LightnessFUR LWFUR MWFUR SWFUR LightnessBG 0.256** . . . LWBG 0.299** 0.333** . . MWBG 0.250** 0.257** 0.244** . SWBG 0.131 0.114 0.133 0.136 Pattern MarkingSizeFUR DominanceFUR DiversityFUR ContrastFUR MarkingSizeBG 0.090 . . . DominanceBG 0.143 0.344** . . DiversityBG 0.068 0.066 -0.026 . ContrastBG 0.091 0.205** 0.089 0.255** Pearson correlation is significance levels are flagged at the 0.05 (*) and at the 0.01 (**). 4 Table S3: Linear mixed effects analyses (LMER) testing whether desert rodents are found from predominant subjective background types where they minimise the animal- to-background distance in XY colour space (i.e. maximise the match). Specifically, LMER tests animal-to-background colour distances in relation to rodent microhabitat from it was found from (focal), mesohabitat matching to average habitat appearance of different backgrounds (local), macrohabitat matching to overall background appearance (global) as well as matching of different species and the respective interactions. Intercept includes individual ID as random variable. Subject Estimate s.e. DF t-value P (Intercept) 2.518e-02 4.179e-03 3.270e+02 6.025 <0.001*** Gravel 3.595e-03 3.172e-03 7.781e+02 1.133 0.257 Rocks 4.120e-03 3.194e-03 7.772e+02 1.290 0.197 Sand 3.905e-03 3.172e-03 7.781e+02 1.231 0.218 Global 1.513e-03 3.194e-03 7.772e+02 0.474 0.635 Focal 4.004e-03 3.172e-03 7.781e+02 1.262 0.207 Gerbillus gerbillus -1.635e-04 6.053e-03 2.834e+02 -0.027 0.978 Gerbillus tarabuli -4.960e-03 5.858e-03 2.859e+02 -0.847 0.397 Jaculus hirtipes 3.271e-02 5.058e-03 3.914e+02 6.468 <0.001*** Meriones libycus 1.216e-02 1.424e-02 2.556e+02 0.854 0.393 Pachyuromys duprasi 8.562e-03 9.326e-03 2.634e+02 0.918 0.359 Psammomys obesus -5.987e-03 9.326e-03 2.634e+02 -0.642 0.521 G. gerbillus * Gravel -3.347e-03 4.501e-03 7.777e+02 -0.744 0.457 G. gerbillus * Rocks -5.087e-03 4.516e-03 7.772e+02 -1.126 0.260 G. gerbillus * Sand -7.284e-03 4.501e-03 7.777e+02 -1.618 0.106 G. gerbillus * Global -2.892e-03 4.516e-03 7.772e+02 -0.640 0.522 G. gerbillus * Focal 1.790e-02 4.501e-03 7.777e+02 3.977 <0.001*** 5 G. tarabuli * Gravel -2.259e-03 4.362e-03 7.777e+02 -0.518 0.604 G. tarabuli * Rocks -2.644e-03 4.378e-03 7.772e+02 -0.604 0.546 G. tarabuli * Sand -4.537e-03 4.362e-03 7.777e+02 -1.040 0.298 G. tarabuli * Global -1.606e-03 4.378e-03 7.772e+02 -0.367 0.713 G. tarabuli * Focal -6.723e-04 4.362e-03 7.777e+02 -0.154 0.877 J. hirtipes * Gravel 1.377e-03 3.966e-03 7.777e+02 0.347 0.728 J. hirtipes * Rocks 4.839e-03 3.972e-03 7.772e+02 1.218 0.223 J. hirtipes * Sand 6.495e-03 3.966e-03 7.777e+02 1.638 0.101 J. hirtipes * Global 3.782e-03 3.972e-03 7.772e+02 0.952 0.341 J. hirtipes * Focal -5.171e-03 3.966e-03 7.777e+02 -1.304 0.192 M. libycus * Gravel 1.618e-03 1.042e-02 7.773e+02 0.155 0.876 M. libycus * Rocks 5.293e-03 1.043e-02 7.772e+02 0.507 0.612 M. libycus * Sand 6.543e-03 1.042e-02 7.773e+02 0.628 0.530 M. libycus * Global 4.055e-03 1.043e-02 7.772e+02 0.389 0.697 M. libycus * Focal 2.743e-02 1.042e-02 7.773e+02 2.631 0.008** P. du p r a s i * Gravel 2.593e-03 6.858e-03 7.774e+02 0.378 0.705 P. duprasi * Rocks 6.127e-03 6.868e-03 7.772e+02 0.892 0.372 P. du p r a s i * Sand 6.806e-03 6.858e-03 7.774e+02 0.992 0.321 P. du p r a s i * Global 4.602e-03 6.868e-03 7.772e+02 0.670 0.503 P. du p r a s i * Focal -7.082e-04 6.858e-03 7.774e+02 -0.103 0.917 P. obesus * Gravel 3.562e-03 6.858e-03 7.774e+02 0.519 0.603 P. ob e s u s * Rocks 6.853e-03 6.868e-03 7.772e+02 0.998 0.318 P. ob e s u s * Sand 6.961e-03 6.858e-03 7.774e+02 1.015 0.310 P. ob e s u s * Global 5.032e-03 6.868e-03 7.772e+02 0.733 0.464 P. obesus * Focal 9.506e-04 6.858e-03 7.774e+02 1.386 0.166 °Intercept includes factor level: background [clay] & species [Gerbillus amoneus]. 6 Figure S1: Animal-to-background colour distances. First panel (A) shows the colour variation between animal (red circles) and background (grey circles) plotted in XY- colour space. In second panel (B) average distances are plotted in relation to the original habitat from where the animal was found (i.e. focal), appearance of predominant types of backgrounds (i.e. clay, gravel, rocks, sand) and general background appearance (i.e. global). 7 Figure S2: Predominant habitat classes plotted in plotted in XY-colour space. On x-axis the X colour coordinates, on Y-axis the Y colour coordinates. In panels, predominant types of backgrounds (i.e. clay, gravel, rocks, sand). 8 References: Van Den Berg, C. P., J. Troscianko, J. A. Endler, N. J. Marshall, and K. L. Cheney. 2019. Quantitative Colour Pattern Analysis (QCPA): A Comprehensive Framework for the Analysis of Colour Patterns in Nature. Methods in Ecology and Evolution 2019:1–17. Kelber, A., M. Vorobyev, and D. Osorio. 2003. Animal colour vision – behavioural tests and physiological concepts. Biological Reviews of the Cambridge Philosophical Society 78:81–118. McLachlan, A. J., and R. J. Ladle. 2011. Barriers to adaptive reasoning in community ecology. Biological Reviews 86:543–548. Nokelainen, O., N. Hubbard, A. E. Lown, L. E. Wood, and M. Stevens. 2017. Through predators’ eyes: phenotype–environment associations in shore crab coloration at different spatial scales. Biological Journal of the Linnean Society 122:738–751. Nokelainen, O., M. Stevens, and T. Caro. 2018. Colour polymorphism in the coconut crab (Birgus latro). Evolutionary Ecology 32:75–88. Todd, P. A., J. Oh, L. H. L. Loke, and R. J. Ladle. 2012. Multi-scale phenotypesubstrate matching: Evidence from shore crabs (Carcinus maenas L.). Ecological Complexity 12:58–62. Troscianko, J., and M. Stevens. 2015. Image Calibration and Analysis Toolbox - a free software suite for objectively measuring reflectance, colour and pattern. Methods in Ecology and Evolution 6:1320–1331.