Statistical analyses. For packaging experiments,
destructive quality data were subjected
to analysis of variance using the general linear
models (GLM) procedure (SAS Institute, 2002)
for a randomized complete block design with
four blocks for each treatment. A block consisted
of fruit harvested from an individual
tree in an orchard. Data for nondestructive
measurements were analyzed for repeated
measures over time using the GLM procedure.
Data from the packaging treatments stored
under different temperature regimes were analyzed
using the GLM procedure for a split-plot
design, with temperature regime as the main
plot and package treatment as the subplot.
The experiment was repeated once, with four
replications per treatment. Data for each experiment
were analyzed separately. After checking
for homogeneity of variance, a combined
analysis was performed for those variables
where differences between experiments (harvest
dates)were not significant and interactions