Rj. Dufault, DETERMINING HEAT UNIT REQUIREMENTS FOR BROCCOLI HARVEST IN COASTAL SOUTH-CAROLINA, Journal of the American Society for Horticultural Science, 122(2), 1997, pp. 169-174
The objective of this research was to determine the least variable met
hod to predict the dates of the first and last broccoli (Brassica oler
acea L. var Italica) harvests based on heat unit summation using coeff
icients of variation (Cv). The method with the lowest Cv for predictin
g first harvest was to sum, over days from planting to harvest, the di
fference between the growing season mean (GSM) temperature and a base
temperature of 7.2 degrees C. If the GSM maximum (max) temperature, ho
wever, was >26.7 degrees C, an adjusted max temperature was calculated
by first subtracting 26.7 degrees C from the GSM max temperature and
then subtracting the GSM mean temperature, Then the growing degree day
s (GDDs) were summed by subtracting the base temperature of 7.2 degree
s C from the average of the GSM minimum (min) and adjusted max tempera
tures, This method produced a Cv of 3.96 compared to 4.13 for the stan
dard method of summing over the entire growing season, the mean temper
ature minus the base temperature of 4.4 degrees C, The method with the
lowest cv for predicting last harvest was to sum, over days from plan
ting to harvest, the difference between the GSM max temperature and a
base temperature of 7.2 degrees C, If the GSM max temperature, however
, was >29.4 degrees C, the base temperature was subtracted from 29.4 d
egrees C and not the actual GSM max temperature. This method produced
a cv of 3.71 compared to 4.10 for the standard method of summing over
the growing season, the mean temperature minus the base of 4.4 degrees
C.