Abstract
Sri Lanka requires nearly 385, 000 tons ot sugar per
Only 10% is produced locally from 6, 300 ha Of
annum.
irrigated and 5, 400 ha of rainfed cultivation in the
The rainfed yield is low when compared to
dry zone .
irrigated cultivation due to erratic rainfall. However the
acute shortage of water forced the sugar industry to expand
cultivation sugarcane under rainfall.
Traditional field experiments used to analyse growth
and yield of sugarcane in the past were confined to few
environmental interactions. The results obtained were Site
and season specific. Moreover, the long growth cycle of
sugarcane, limited the number of trials possible within a
In this proj a model on growth of
time period.
sugarcane under rainfed and irrigated condi t ions is
undertaken, to be used as a tool for analyzing growth and
yield to help planning and management of sugarcane
production in Sri Lanka.
A mathematical model was developed to simulate the
environmental and physiological process involved in the
This consists
growth of rainfed and irrigated sugarcane.
Of a growth, and soil water balance sub model to keep track
Of the soil moisture.
The crop model uses daily maximum
and minimum temperature (degree days) to calculate the phenological development (emergence, tillering and canopy
development, grand period of growth and maturity) .
Dry
mat eer produced from the canopy by the interception of
daily photosynthetic radiation after accounting for
respiration, is partitioned between leaves, roots, stalks
and sugar according to the phenological development .
water stress factor obtained from the ratio Ot actual to
potential cranspiration is taken to account for moisture
stress. The daily biomass is proportionally reduced to the
stress .
The output of the model was compared with data
obtained from Sugarcane Research Institute of Sri Lanka
The resulLs showed that the simulated
from 1981 to 1990.
dry matter is in close agreement with the measured dry
matter. The simulated results are within a accuracy.
Regression analysis of combined irrigated and rained
2
actual and simulated biomass gave a high correlation (r
0.83).
Reasons for variation in yield and possible additional
improvements to the model are discussed in this report .