Fig. 4. Comparison of adult male satin bowerbird reflectance spectra, either measured using a spectrometer (solid lines) or predicted from thin-film optical models (broken and dotted lines). Predicted spectra are modeled based on measurements of barbule nanostructure from the individual whose measured spectrum is shown. Model 2 (dotted line) closely predicted measured spectra (solid line). By contrast, Models 1 (dashes and dots), 3 (short dashes), and 4 (long dashes) were poor predictors of measured spectra.