期刊
BEHAVIORAL ECOLOGY
卷 26, 期 2, 页码 367-375出版社
OXFORD UNIV PRESS INC
DOI: 10.1093/beheco/aru198
关键词
Anas platyrhynchos; colour; Cyanistes caeruleus; just noticeable differences; JND; PCA; tetrachromatic; vision; Vorobyev-Osorio model
资金
- Australian Research Council [DE120102323, FT110100505]
- Monash University Research Accelerator Program
- Max Planck Society
- Australian Research Council [DE120102323, FT110100505] Funding Source: Australian Research Council
To study biologically relevant variation in visual signals, these need to be assessed in relation to the sensory abilities of receivers. For the study of colors, reflectance spectrometry has been the method of choice, but analyses of reflectance spectra present challenges that hamper our understanding of color variation. Among these are computing meaningful color variables and interpreting their biological relevance. Here, we suggest how to overcome the limitations of commonly used approaches. We describe how to use psychophysical visual models to assess chromatic variation in the visual space of animals. This approach consists of 1) obtaining cone quantum catches from reflectance spectra, 2) transforming these into visual space coordinates where Euclidean distances reflect perceptual distances, 3) summarizing variation in visual space using principal component analysis (PCA) maintaining original perceptual units, and 4) interpreting the axes of chromatic variation (PC) based on their loadings and relative and absolute levels of chromatic variation. We illustrate this approach by comparing it to traditional color indices (hue and saturation) and PCA computed directly on reflectance spectra, using 2 examples: 1) determining the biological relevance of correlations between bill coloration and male quality in mallards and 2) assessing the success of experimental color manipulations in blue tits. In both cases, re-analyzing the data suggests different interpretations. This approach provides a simple way of objectively summarizing chromatic variation and interpreting the magnitude of biologically relevant effects. We provide R scripts to carry out computations and recommendations on how to report results to make data comparable between studies.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据