School of Life Sciences, University of Sussex, Brighton, United Kingdom.
Am Nat. 2011 May;177(5):681-90. doi: 10.1086/659626.
It might seem obvious that a camouflaged animal must generally match its background whereas to be conspicuous an organism must differ from the background. However, the image parameters (or statistics) that evaluate the conspicuousness of patterns and textures are seldom well defined, and animal coloration patterns are rarely compared quantitatively with their respective backgrounds. Here we examine this issue in the Australian giant cuttlefish Sepia apama. We confine our analysis to the best-known and simplest image statistic, the correlation in intensity between neighboring pixels. Sepia apama can rapidly change their body patterns from assumed conspicuous signaling to assumed camouflage, thus providing an excellent and unique opportunity to investigate how such patterns differ in a single visual habitat. We describe the intensity variance and spatial frequency power spectra of these differing body patterns and compare these patterns with the backgrounds against which they are viewed. The measured image statistics of camouflaged animals closely resemble their backgrounds, while signaling animals differ significantly from their backgrounds. Our findings may provide the basis for a set of general rules for crypsis and signals. Furthermore, our methods may be widely applicable to the quantitative study of animal coloration.
似乎显而易见的是,伪装的动物通常必须与其背景相匹配,而要引人注目,生物体必须与背景有所不同。然而,评估图案和纹理的显著性的图像参数(或统计数据)很少得到很好的定义,并且动物的颜色图案很少与各自的背景进行定量比较。在这里,我们在澳大利亚巨型乌贼 Sepia apama 中研究了这个问题。我们将分析仅限于最知名和最简单的图像统计量,即相邻像素之间的强度相关性。Sepia apama 可以迅速将其身体图案从假定的显眼信号转变为假定的伪装,从而为研究这些图案在单一视觉栖息地中的差异提供了绝佳而独特的机会。我们描述了这些不同身体图案的强度方差和空间频率功率谱,并将这些图案与它们所观察到的背景进行了比较。伪装动物的测量图像统计数据与它们的背景非常相似,而信号动物则与它们的背景有明显的不同。我们的发现可能为隐身和信号提供了一套通用规则的基础。此外,我们的方法可能广泛适用于动物颜色的定量研究。