Statistical Approach to Finding Probable Feeding Attractants for Young Herbivorous Abalone Haliotis discus

Abstract

The attraction index a, which is the coefficient a of a logistic curve y=g/{1+exp[-r(j-a)]} in the remaining number time-courses, seemed not to be correlated positively with any changes of the contents of the following nine general chemical constituents, volatile organic acids, crude lipids, aldehydes, crude proteins N, crude non-proteins N, amino acids, volatile bases N, betaines, and sugars, which were determined in the four kinds of representative food seaweeds during storage at low temperature. This led us to seek statistically feeding attractants by the significance test of correlation between the index a and the contents of each of the nine constituents or the combination of two or three of them. The significance test of multiple and partial regression coefficients, food seaweed dummies and storage days showed that amongst the constituents the combination of three, the proteins and the two constituents amongst the amino acids, the organic acids, the aldehydes, the lipids and the betaines, was significant at high level of 0.01 or low one of 0.05. From this, it was suggested that the proteins as main component possessed the most probable attractive capacity for young herbivorous abalone.