Ir. Brooking, EFFECT OF TEMPERATURE ON KERNEL GROWTH-RATE OF MAIZE GROWN IN A TEMPERATE MARITIME ENVIRONMENT, Field crops research, 35(2), 1993, pp. 135-145
Maize (Zea mays L.) kernel growth rates were estimated during the line
ar period of grain-filling for 10 held sowings of cultivar Pioneer P39
01 and 12 field sowings of cultivar Pioneer P3709. The objective of th
e study was to determine the relationship between kernel growth rate a
nd air temperature in a cool climate, and to compare this relationship
with published controlled-environment studies and with the quadratic
function used to describe kernel growth in the CERES-maize crop simula
tion model. The crops were grown in a temperate maritime environment a
t latitude 40.3 degrees S, and mean air temperatures during the period
s for which kernel growth rates were measured ranged from 12.5 to 19.3
degrees C. Measured kemel growth rates ranged from 3.6 to 9.2 mg kern
el(-1) day(-1). Between 13.5 and 19.3 degrees C, the rates were linear
ly related to temperature, increasing by 0.59 and 0.44 mg kernel(-1) d
ay(-1) degrees C-1 for P3901 and P3709 respectively. Kernel growth rat
es were markedly reduced at mean air temperatures below 13.5 degrees C
, indicating that there was either a non-linear response of the potent
ial kernel growth rate to temperature over the 12 to 19 degrees C temp
erature range, or that assimilate supply limited the actual kernel gro
wth rate in the field at low temperatures. Non-linear regression proce
dures were used to fit kernel growth rates from four sowings in one se
ason to linear and quadratic thermal-time models using hourly mean air
temperatures. There were no apparent assimilate limitations during ke
rnel growth in this season, and hourly mean air temperatures ranged fr
om 0 to 27 degrees C. The results supported the use of a non-linear fu
nction to describe the overall response of kernel growth rate to tempe
rature, which in turn cautions against the use of mean daily temperatu
res to derive such a relationship. The quadratic model provided the be
st fit to the data, and gave a higher base temperature than that used
in the CERES-maize model (10 degrees C vs. 6 degrees C).