Geosmina em Tambaqui, Colossoma macropomum, (Cuvier, 1818), de pisciculturas em diferentes ecossistemas de Roraima

Abstract

In fish farming, many factors can cause the appearance of off-flavor phenomenon, which means undesirable tastes and odors and is one of the main obstacles that affect the sale of fish in the world. The occurrence of off-flavor in the fish flesh is due to the existence of certain substances which are present in the water, the geosmin (GEO) is the most important compound. This study aimed to determine and quantify the presence of GEO in the flesh of tambaqui (Colossoma macropomum, Cuvier, 1818) grown in fish farms located in different ecosystems of Roraima. The study was conducted in the Brazilian Amazon region, state of Roraima, during the months of December 2013 to April 2014 in 4 fish farms located in two types of ecosystems, 2 in savanna areas and 2 in forest, being analyzed meat 60 fish and water tanks 12, distributed in a completely randomized design (CRD) with two treatments and two replications and determining the difference between the treatment means was carried out using Student's t test for independent observations when variances are not equal. We evaluated the quality of water, the analytical method for the detection of geosmin by QuEChERS and Liquid Chromatography of High Efficiency and the correlation of two situations, one of the analytical measures of GEO and the scores of the sensory panel and one between the physicochemical characteristics chemical water and analytical measures of GEO. There were statistically significant differences between the mean concentration of geosmin (t = 3.050, p = 0.04) in different types of ecosystems. Only the pH characteristics and the temperature maintained appropriate values in all fish farms studied. The strongest correlations between physico-chemical properties of water and geosmin concentrations obtained by HPLC was dissolved oxygen (r = 0.58; p <0.001), pH (r = 0.41; p <0.001) and nitrite (r = 0.38; p <0.001). As for validation, was obtained excellent linearity (r = 0.9998), accuracy (98% and 87%) and precision (CV = 8.7%). Low correlation coefficients (r = -0.001 -0.363 = air) were observed when the correlation between the variables sensory panel and analytical measures of geosmin.