Influência dos Defeitos dos Grãos de Café na Percepção das Características Sensoriais da Bebida Pelo Consumidor

Abstract

The coffee segment has important role on the Brazilian agribusiness. The coffee quality depends on the strong control of several factors such as production and roasting, which affect hundreds of components that yield the characteristic aroma and flavour. Defects as black bean, immature bean, sour bean, stones, stick, husk, bean fragment, shell or ear, broken bean are obtained after mechanical sorting and represent about 20% of the production, reaching around 2.6 million of bags/year. The addition of defects into the good quality beans is common in Brazil, mainly for the internal market products in order to minimise losses. However, this addition may have a negative impact on the internally consumed beverage quality, affecting the consumer hedonic responses. This study aimed at evaluating the effect of the defects on the sensory quality of the arabica coffee beverage (Coffea arábica L.) by identifying the point at which coffee consumers would begin to perceive the difference between the control sample (good quality one) and those added of defects (detection threshold), as well as the addition limit at which consumers would begin to reject the beverage containing defects (rejection threshold). Two hundred regular coffee drinkers of at least one cup a day were recruited from the University of Veiga de Almeida to take part in this study. The paired preference test was used to investigate the rejection threshold, where 100 consumers evaluated the control sample vs. five different levels of defects (5, 10, 20, 30 e 40%), and indicated the preferred sample in each pair. Samples were sweetened ad libitum with artificial sweetener or sugar (or none, according to each participant habit). Triangle test was applied to determine the detection threshold for coffee defects where five different levels of defects (1, 3, 5, 10 e 20%) were compared to the control sample. Similarly to the previous study, the beverages were sweetened ad libitum. Considering the Paired Preference Test, it was observed no difference among the control sample and those added of defects, revealing that participants liked both the control beverage and those with defects. The results from the Triangle Tests have shown that consumers were able to identify the odd sample at the 20% level (p<0.05). The detection threshold was estimated as 16%, i.e., the point from which consumers were able to perceive difference between the control and the beverage added of defects. This study will contribute to the knowledge about the role of the defects on the coffee beverage, which is a crucial factor on the orientation regarding the use of the referred beans in the commercial blends, without compromise the quality perceived by the consumer.