MODELING AND FORECASTING YIELD OF RICE FOR KANGRA DISTRICT OF HIMACHAL PRADESH
Loading...
Date
2019-01-30
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
CSKHPKV. Palampur
Abstract
A field experiment was conducted during Khraif 2017 and 2018 at Experimental Farm of Rice
and Wheat Research Centre Malan with five dates of transplanting (15th June, 25th June, 5th July,
15th July and 25th July) and three varieties (HPR 1068, HPR 2880 and HPR 2143) in factorial RBD
with three replications. The soil of the experimental field was silty clay loam in texture, acidic in
reaction, medium in available N, high in available P and medium in available K. The growth and
yield attributes viz., plant height, number of tillers m-2, leaf area index, number of grains panicle-1
and 1000-grain weight were significantly higher in 15th June transplanted crop during both the
years though it was at par with 25th June transplanted crop. There was reduction in grain yield with
delay in transplanting after 15th June. The decrease averaged over two years was 7 % in 25th June,
16 % in 5th July, 32 % in 15th July and 45 % in 25th July transplanted crop over 15th June
transplanted crop. The early transplanted crop took more days for tillering, panicle initiation, 50 %
flowering and maturity. Net returns and B: C ratio was significantly higher in 15th June
transplanted crop compared to subsequent dates of transplanting. The agrometeorological indices
indicated more values for 15th June to 5th July transplanted crops and lowest values in late
transplanted crop. Amongst varieties, HPR 2143 gave significantly higher values of growth and
yield attributes during both the years. Likewise, grain yield, biological yield, net returns and B: C
ratio was significantly higher in variety HPR 2143 followed by HPR 1068 and HPR 2880. The
DSSAT CERES- Rice model performed well, [high correlation coefficient (r), low root mean
square error (RMSE) and low mean absolute percentage error (MAPE)] in simulating the
physiological maturity (r= 0.89 and 0.95; RMSE= 3.8 and 2.9 days; MAPE= -1.4 and -2.1 %), leaf
area index (r= 0.97 and 0.96; RMSE= 0.08 and 0.06; MAPE= 0.96 and 0.82 %), grain yield (r=
0.98 and 0.96; RMSE= 372 and 333 kg ha-1; MAPE= -5.0 and - 9.0 %), biological yield (r= 0.96
and 0.98, RMSE= 656 and 824 kg ha-1; MAPE= 3.2 and 7.2 %) during 2017 and 2018,
respectively. The simulated days to physiology maturity, LAI, grain and biological yield matched
closely with observed values for all transplanting environments. The model performance was
somewhere under estimated or overestimated but found within acceptable limits. The predicted
yields of rice using validated CERES-Rice model indicated -6.4 to 6.2, -10.8 to1.2 and -5.2 to 7.6
% at F1stage during 2015, 2016 and 2017, respectively and -4.2 to 7.2, -8.7 to 2.2 and -3.1 to 8.6
% errors during 2015, 2016 and 2017 at F2 stage. The yield was more reliable in F2 stage of
district Kangra. The model predicted ±10 percent deviation in rice yield compared to actual
productivity of district under study and can be used for the yield forecast in the district.