Estimates

The coefficients and significance levels for trait variable coefficients are reported in Table 10.2 for both data sets. Two specifications are reported for the district data. In the first, all trait variables are included. In this specification three trait coefficients had marginally significant negative coefficients, possibly reflecting susceptibility. In the second specification, these three variables are dropped. In the state regressions, no varieties with resistance to bacterial leaf stripe are included in the data set.

The estimates actually tend to be reinforcing. In the case of the first data set,

Table 10.2. Trait value estimates.

Trait

District 1

District 2

State

Blast

-0.134*

0.184**

Bacterial leaf blight

-0.077

-0.069

-0.134**

Bacterial leaf stripe

-0.065

0.173

Rice tungro virus

-0.146*

0.068

Sheath blight

1.48**

1.49**

0.108

Brown plant hopper

-0.151*

0.033

Green leaf hopper

0.037

0.052

0.123**

White-backed plant hopper

0.309**

0.309**

0.377**

Gall midge

0.091

0.102*

0.174**

Stem borer

0.155*

0.029

0.141*

sheath blight resistance and white-backed plant hopper resistance contributed positively to yield increases in both the specifications. Stem borer resistance contributed positively to yield in the first specification, and gall midge resistance had positive effects on yield in the second specification. In the state data set, bacterial leaf blight resistance depressed rice yield, contrary to normal expectations. Resistance to blast contributed positively to yield. Resistance to other diseases did not yield any significant impacts. Resistance to insect pests such as green leaf hopper, white-backed plant hopper, gall midge, and stem borer yielded significant and positive impacts on yield. Only the resistance to brown plant hopper did not yield a significant effect.

To sum up the results, varieties with insect resistance show better performance in the field in both data sets, although neither showed that resistance to brown plant hopper is important. The estimates for disease resistance, on the other hand, are much weaker. Both data sets show yield effects for sheath blight resistance and the state data set shows a blast resistance effect and a positive but non-significant rice tungro virus effect.

Was this article helpful?

0 0

Post a comment