Statistical analyses were used to develop predictive
models of rice blast and to relate the favorability of
environment to disease incidence and severity on different
rice cultivars at five sites in Asia. The WINDOW PANE
program was used to search for weather factors highly
correlated with blast. Stepwise and r-square linear
regression procedures were then applied to generate the
predictive models at each site. Models developed at Icheon,
South Korea included relative humidity and rainfall factors
as the most important predictors of disease. Temperature,
rainfall, wind speed, and relative humidity factors were
components of models at Cavinti and the IRRI blast nursery
in the Philippines. Rainfall, temperature, and solar
radiation factors were important at Gunung Medan and
Sitiung, Indonesia. Model validation was done to verify
accuracy of models for predictions. Model predictions were
also used to determine the effects on blast of sowing time,
nitrogen amount, and increase in temperature. Limitations of
the models are discussed.
Path coefficient analysis was used to identify direct
and indirect influences exerted by weather factors on blast.
The largest direct influence on disease was exerted by
humidity factors at Icheon; temperature, rainfall, and wind
speed factors at Cavinti; temperature and humidity factors
at IRRI; rainfall factors at Gunung Medan; and temperature
factors at Sitiung. Although path coefficient values (Py)
were estimated from the decomposition of correlation
coefficients, factors that had a high correlation with
disease parameters did not always give high Py.
Multivariate analysis was used to determine the effects
of sowing times on proneness of tropical rice to blast.
Cluster analysis of 24 hypothetical sowing months at
Cavinti, the IRRI blast nursery, and Sitiung sites revealed
three blast proneness groups. Principal component analysis
showed that IR50 cultivar would be susceptible at Cavinti at
any time of the year. Sowing C22 cultivar at Cavinti in
Group I and III months would make it prone to panicle and
leaf blast, respectively. At the IRRI blast nursery, leaf
and panicle infections on IR50 would be probable only in
Group I and II months. This trend was also observed for C22
at Sitiung, although some months in Group III at this site
had moderate to high degree of proneness to leaf blast. / Graduation date: 1994
Identifer | oai:union.ndltd.org:ORGSU/oai:ir.library.oregonstate.edu:1957/35832 |
Date | 26 April 1994 |
Creators | Calvero, Silvino B. |
Contributors | Coakley, Stella M. |
Source Sets | Oregon State University |
Language | en_US |
Detected Language | English |
Type | Thesis/Dissertation |
Page generated in 0.0016 seconds