34343

DISCRIMINATION ACCORDING TO THE AGRONOMIC CONDITION OF AÇAI (EUTERPE OLERACEA) SAMPLES USING HAND-HELD NEAR INFRARED SPECTROMETER

Risaldo Amaral Jr 1 ; João Paulo Silva 1 ; Vincent Baeten 2 ; Hervé Rogez 1
Favorite this paper

INTRODUCTION
Açai tree (Euterpe oleracea Mart.) is a native and abundant palm tree in the Amazon estuary. It is cultivated in several states of Brazil as well as in South and Central America countries. The composition of the fruit pulp varies with period of the year, agronomic conditions (floodplains and lands) and geographic origin. It influences greatly the quality of the final product whatever it is juice or concentrated extract. In 2011, domestic production of açai fruits 215,000 tons, for only 121,000 tons in 2000. This evolution reflects the increase of interest in this fruit at national and international levels. Economic sector and researchers are very attracted by açai juice due to its high-energetic value and functional properties, in direct relation with the high lipid and phenolic compound concentration, respectively. Phenolic compounds represent 1 to 3% of the dry matter and anthocyanins are largely predominant. Phenolic compounds are known to be excellent antioxidants able to slow down the aging process by neutralizing free radicals in the human body. To raise the national and international competitiveness and high quality standards, it is necessary to develop technological tools to guide the food industry in the post-harvest monitoring and quality control. Therefore, a hand-held Near Infrared (NIR) Spectrometer (Phazir, Polychromix) coupled with chemometrics has been tested. This instrumentation showed to be a reliable, fast and non-invasive technique suitable for post-harvest monitoring. This study aimed to discriminate samples of açai fruit from two agronomic conditions (i.e. acai coming from flooplains or lands) based on NIR spectroscopy combined with chemometrics.

EXPERIMENTAL
In this study 96 samples of açai fruits were collected from different regions of the states of Pará and Maranhão. The samples have been classified according to the type of agronomic conditions (floodplains and lands). Several parameters have been determined for the samples: juice yield after pulping, average fruit weight, total lipids content and total anthocyanins content. For each sample, mean NIR spectra of 25 individual fruits have been collected (4 NIR analysis/fruit) and averaged. Averaged spectra have been pre-treated using MSC. Principal Component Analysis (PCA) and Partial Least Squares Discriminant Analysis (PLSDA) have been used to discriminate the sample according to the agronomic condition.

RESULTS AND DISCUSSION
Samples of floodplains açai present a higher (P<0.05) yield of dry matter and lipid content than land samples. However, the mean fruit weight and the content of anthocyanins did not vary between both types of soils (P>0.05). The Principal Component Analysis (PCA) applied on the average NIR spectra show according to the two first PCs cluster related to the agronomic conditions. PC1 and PC2 explain 70% and 22% of the total variance, respectively. PLSDA was also applied and allows discriminating açai samples. 100% of the samples were correctly classified according to their agronomic growing conditions. The results of the study demonstrated that NIR spectroscopy combined with chemometrics can be a suitable technique for the quality control in the açai industry. Moreover, the methodology could be also proposed for traceability and authenticity purposes.