| Summary: | 碩士 === 國立臺灣海洋大學 === 食品科學系 === 92 === Live eels (Anguilla japonica) were purchased from the package plant, which were preliminarily divided into three grades (good, off-color and sick eel) by packagers. The non-destructive instruments including near-infrared reflectance spectroscopy (NIRS), color difference meter and electromagnetic scanning (EM-Scan) meter were used to evaluate the quality of eels. In addition, chemical compositions of three kinds of eels were also detected.
Sick eels had the lower condition factor and higher hepatosomatic index than good and off-color eels. In addition, the liver color of good eel was darker than that of sick eel. The percentage not meeting the size criteria (4-6 fish/kg) for good, off-color and sick eels was 15.4, 51.8 and 39.0％, respectively. The difference in proximate composition of good and sick eels was not apparent, but sick eels had a higher level of moisture and lower levels of protein and ash. The fat content showed no good correlation with body weight; however, it had a negative correlation (R2=0.7672) with moisture. Moisture could be used to predict the fat content (Fat＝-0.9952 x Moisture + 79.4352).
Good eel meat had higher levels of free amino acid among three kinds of eels. Good eel meat also had higher levels of carnosine (Car) and lower levels of histidine (His). Car was not detected in the plasma and liver, and a small amount of histidine was found only in the liver of good eel. Anserine (Ans), however, was not detected in any tissues and plasma of three kinds of eels. Good eel had higher carnosinase activities in plasma and tissues than off-color and sick eel. The activity in order as liver＞gill＞vertebra＞meat. The activity was not found in the meat of three kinds of eels.
Sick eels had a higher L value, but off-color eels had a higher b value. The b value “+7.00” was an objective color discriminator between good and off-color eels. If the b value was greater than +7.00, the eel was assigned to the group of off-color eel.
Off-color eel had higher total body electrical conductivity (TOBEC) values than good and sick eel measured by EM-Scan meter. The correlation between TOBEC value and the total moisture content was found. The correlation of proximate composition of eel meat using NIRS spectrum through Vision Analysis System was also investigated. The results indicated that either moisture (R2 = 0.7870) or fat (R2 = 0.8828) had a good correlation between predicted and measured values, but protein (R2 = 0.3305) and ash (R2 = 0.1827) showed no good correlation. In addition, the a value (R2 = 0.7477) also had a good correlation. NIRS spectrum of good, off-color and sick eels showed a significant distribution in 3D diagram via statistical analysis. NIRS could be used not only to predict the fat and moisture content of eel, but to discriminate the quality of eels.
|
|---|
