Abstract:

Electronic tongue (E-tongue) technology was applied in this study to detect the taste profiles of the abdomen,
claw, leg meat and gonad part of female Eriocheir sinensis farmed in Yangcheng Lake, Songjiang and Chongming regions
from different quality grades. The principal component analysis (PCA) of E-tongue data showed that the taste profiles
of four edible parts of Yangcheng-Lake crabs in three grades (premium, first and second) could be distinguished well.
Soft independent modeling of class analogy (SIMCA) method was applied to establish the habitat identification models
of Yangcheng-Lake crabs based either on one single part or on multiple parts with a rejection ratio of 100% for non-
Yangcheng-Lake crabs. In addition, partial least squares-discriminant analysis (PLS-DA) was also used in this study to build
three grade evaluation models for Eriocheir sinensis of different grades. The correlation coefficients of the PLS-DA models
were all more than 0.90. Also, based on the PLS-DA models, the grades of 16 unknown crab samples could be correctly
identified with 100% accuracy.

Key words: Chinese mitten crab (Eriocheir sinensis), taste, habitat identification, grade evaluation, electronic tongue