260 P. BISWAS ET AL.
protein and enzyme functions directly. Therefore, it is possible that
at a given chilling temperature, enzyme activities related to different
processes that lead to injury symptoms will be affected differently. For
example, it is possible that the activation energy of enzymes associated
with lycopene synthesis or natural defense mechanisms are different
and that results in a differential extent of damage at different chilling
temperature.
Since at a given time, different chilling temperature may result
in different symptoms or at a given temperature, different processes
(organelle structure, enzyme activity, or protein functions) may be ham-
pered sequentially, it is possible that each symptom has an independent
time–temperature threshold to appear and this may explain why some
symptoms appear earlier than others.
A model is presented to show different chilling symptoms of mature-
green tomatoes as influenced by temperature and time of exposure
to different chilling thresholds (Fig. 5.1). The observed CI symptoms
are divided into seven categories—flavor loss, uneven/blotchy red col-
oration, complete failure to ripen, excessive softening, pitting/sunken
patches, decay, and ion leakage. The first appearance of symptoms
described in each study was plotted on a time–temperature plane. Each
symptom was then grouped by identifying a line of the first time of
appearance of a symptom at a temperature indicating the potential sen-
sitivity of each symptom to chilling. The resulting area under each
respective line represents where there is an expectation for that chilling
injury to be induced by the storage condition described by the time and
temperature axis.
Across the published literature, tomatoes differ in genotype and grow-
ing location, and while this introduces error, for this meta-analysis
method it was assumed that generic effects would apply to all toma-
toes. The data used for the meta-analysis is also limited since recording
of symptoms in postharvest experiments tends to occur after a “pre-
decided” time (determined during experimental planning). A number
of studies do not involve earlier examination of fruit (e.g., Artes and ́
Escriche 1994). Hence, a number of data points may represent a longer
time than required to establish the described chilling injury. As a result,
the model presents conditions that are known to cause symptoms, as
opposed to a conservative model, which would identify regions where
there would be a risk of symptoms occurring. The resulting model
(Fig. 5.1) suggests that there is a relatively constant sequence for ini-
tiation of chilling symptoms. Some symptoms appear very soon dur-
ing exposure to low temperatures, some appear rapidly but take a little