Macro-ecological patterns in the colours of animals and plants

Colour is one of nature’s most remarkable signal features, and plays crucial roles in numerous biological interactions. This thesis uses modern comparative methods to quantify trends in colouration across macroecological gradients. To advance the study of biogeographical patterns in colour, I needed to establish the best methodological approach for quantitative cross-taxa analyses. Analysis of bird and flower colours revealed that to quantify a species’ colour to within 5% of the true mean, only one sample is really necessary. Brightness is more variable, requiring four samples to achieve this precision. The idea that species in tropical regions are more colourful than those at higher latitudes has endured for more than a century. I provide the first taxonomically and spatially broad, quantitative test of the colourful tropics hypothesis. I measured the colours of 570 bird species, 424 butterfly species and flowers of 339 Angiosperm species using reflectance spectrometry and wave-band limited photography. I calculated colour brightness, saturation, diversity, hue disparity and maximum contrast for each species. Phylogenetic and cross-species analyses showed that birds, butterflies and flowers in tropical regions are not on average more colourful than species in higher latitudes, and formally reject the long-standing hypothesis. Next, I determined which biotic and abiotic factors were most influential in shaping macroecological patterns in colouration, and evaluated the generality of several current hypotheses regarding environmental drivers of colour. Relationships between flower colouration and environmental variables were often opposite to those of bird and butterfly colours. For example, birds and butterflies tend to be less colour-saturated under high solar radiation, while flowers tend to be more colour-saturated. Macroecological gradients in flower colour were best predicted by rainfall and the diversity of plants and of pollinating insects, while bird and butterfly colours were best predicted by bird diversity, solar radiation and temperature. Birds, butterflies and flowers show similar latitudinal gradients in colour, but these are driven by different environmental variables. My thesis substantially advances comparative analysis of colouration. As well as finally quantifying the latitudinal gradients in colour, this research will inform future study design, and enable development of novel hypotheses in colour ecology and evolution