Influence of geographical conditions on carotenoid content of Portuguese cabbage

Rationale and objectives. Carotenoid content of food plants often varies not only with species and variety but also with natural factors such as light exposition, soil and temperature. The extent of this variation is not established since many non-controllable factors are involved for a long period of time and these factors may affect different plants in different manners. The objective of this work is to study the carotenoid content of the variety Portuguese cabbage in three landraces of three country regions and to contribute to the definition of the respective sampling plan. Materials and methods. Three different samples of Portuguese cabbage (Brassica oleracea L. var. costata D.C.) landraces, Valhascos, glória de Portugal and penca respectively from Ribatejo (centre), Beira Alta (inland north) and Minho (north) were studied in relation to their carotenoid content which was quantified, after extraction, by a reversed phase HPLC method. Analytical measurement uncertainty was estimated based on data from the in-house method validation. The between samples variance was compared with the measurement uncertainty through an F-test. Results. Lutein and all-trans- -carotene content in Portuguese cabbage samples were, in mg/100 g, respectively 4.7 and 3.6 for Valhascos, 0.52 and 0.46 for glória de Portugal, and 3.3 and 2.8 for penca. Relative analytical measurement uncertainty was 0.19 and 0.21, respectively, for lutein and all-trans- - carotene. At a significance level of 5%, Valhascos and penca landraces did not present statistically significant differences. However, glória de Portugal landrace was statistically different from the last two. Conclusions. The analysed Portuguese cabbages are very good sources of lutein and -carotene presenting diet advantages. The results show that carotenoid content of Portuguese cabbage varies with the geographical region of production easily overshadowing contributions from the analytical process. That factor should be addressed in the production of data for Food Composition Data Bases and based on this study, for the group of Valhascos and penca, and for lutein, 5 primary samples are necessary to estimate the population’s mean value, with a 95% confidence and 10% accuracy. Keywords. Carotenoid, Uncertainty evaluation, Sampling, HPLC, Food