Abstract:
During the seasons of 1986 and 1987 research was done on the influence on keeping quality of the characteristics firmness, colour, russeting, goldspot, damage/bruising, EC values of the fruits and temperature and radiation levels before harvest, to develop a model for predicting keeping quality at the point of delivery at the auction.
Firmness proved to be the most important characteristic for predicting the keeping quality.
Colour, russeting and goldspot were of some but minor importance.
EC value of the fruit and weather before harvest proved to be of very little importance especially when the influence of all the factors combined was examined.
By using stepwise multiple regression analysis, models were made to predict the keeping quality.
The best fitting model had a correlation coefficient of 0.67 and an explained part of 0.44, which is not sufficient for practical use.
By using stepwise discriminant analysis another model was made for selecting tomatoes with a keeping quality less than 5 days.
This model, which was completely based on firmness at the point of delivery at the auction, proved to be more usable.
|