부산대학교 도서관

Roses ( Rosa sp.) are an important ornamental crop worldwide. Their colorful fl owers mainly refl ect an accumulation of anthocyanins and carotenoids. Developing a reliable method to classify rose petal color and identifying relationships between pigment contents and color space values may off er better evaluation criteria for rose varieties. In this study, we classifi ed 60 rose varieties into three groups based on their color parameters, corresponding to red varieties, white and yellow varie- ties, and pink and dark pink varieties. We measured the total pigment contents and identifi ed the underlying anthocyanins and carotenoids using both UV spectrophotometry and ultraperformance convergence chromatography coupled to mass spectrometry. Flower petals of white roses contained the lowest pigment levels, while those of yellow roses contained only carotenoids (40.65–244.42 μg/g) and mainly in the form of β-carotene and violaxanthin. The petals of pink and dark pink roses only accumulated anthocyanins (91.72–1703.93 μg/g) and mainly as cyanidin 3,5-diglucoside and cyanidin 3- O-glu- coside. The petals of red roses contained both large amounts of anthocyanins (1484.8–3806.22 μg/g) and small amounts of carotenoids (1.81–18.77 μg/g). We divided the 60 rose varieties tested here into fi ve color groups based on optical spectrum and pigment content analyses. We also explored the relationships between anthocyanin contents, carotenoid contents, and fl ower color space values using principal component analysis, Pearson’s correlations, and non-linear models. In addition to providing a more accurate system of rose petal color classifi cation, our results can be used to predict pigment contents based on color parameters.