K.L. Heong, M.M. Escalada, N.H. Huan, H.V. Chien, L.V. Thiet and N H Quynh
International Rice Research Institute, Los Baños, Philippines
Visayas State University, Leyte, Philippines
Plant Protection Department, Ho Chi Minh City, Vietnam
Southern Regional Plant Protection Center, Long Dinh, Tien Giang, Vietnam
Can Tho Department of Agriculture, Can Tho, Vietnam
An earlier post presented recent data from China, Thailand and Vietnam to show that the yields obtained by farmers were unrelated to the number of insecticide sprays or spending. We explored the data collected in the new millennium from 2002 to 2007 in a series of on-farm surveys. Details of these surveys are found in Escalada et al 2009.
We analyzed the relationships between farmers’ yields and the number of insecticide sprays they had applied, by comparing the mean yields of cases with different numbers of sprays using one way ANOVA. The spray categories with less than 5 cases were not used in the analyses. In the regression analyses, we excluded the farmers who did not use any insecticides. A total of 7071 cases (1724 or 24% did not use any insecticides) from 9 data sets used.
Table 1: Mean yields of farmers with different number of insecticide sprays in 9 data sets.
[table id=5 /]
Table 1 shows the mean yields of farms with no insecticide spray and those with 1 to 10 sprays. Of the 9 surveys, there was no significant difference in yields obtained in different number of sprays in 4 surveys. In two surveys the differences were just marginally significant, while 3 surveys had significant differences in yields. In 2006 and 2007 Can Tho surveys the farms with high sprays had higher yields. However in the Mekong wide survey yields of farms sprayed 10 times were the lowest, about 2.4 tons or 45% less than farms that had no spray.
In the regression analyses, 5 of the 9 data sets had no significance and 1 had marginal significance but the regression coefficient was very low at +0.10 (see Table 2) The 3 surveys conducted in 2006 and 2007 had significant regressions, but R2 values were very low. The data set with the highest R2 value and seem to give a reasonable regression was that of the Mekong wide survey in 2006 and the regression coefficient was negative, – 0.36. This strongly implied that farms that had high number of sprays yielded lower.
Table 2: Regression analyses of yield and number of insecticide spray relationships in 9 surveys. Cases with no insecticide spray were excluded.
|Location||Year||Cases with no spray||Cases who sprayed||Regression
|Can Tho||2002||390||453||0.1 ns|
|Tien Giang||2003||47||538||1.9 ns|
|Can Tho||2004||430||477||3.4 ns|
|Tien Giang||2004||115||522||2.8 ns|
|Vinh Long||2004||128||472||1.1 ns|
|Vinh Long||2005||198||411||4.4 *||+ 0.10||0.01|
|Can Tho||2006||196||404||7.7 **||+ 0.14||0.02|
|Mekong wide||2006||98||1692||242.7 **||– 0.36||0.12|
|Can Tho||2007||122||478||20.0 **||+ 0.20||0.04|
Based on the 9 farm data sets, we conclude that farmers’ insecticide sprays were either unrelated or negatively related with their yields in 7 out of the 9 data sets. In 2 data sets, the relationships were positive but weak. In the regression analyses variations were high and only accounted for less than 15% of the data sets’ variations. Thus we continue to suspect that farmers’ insecticide use would not contribute towards their yields. Farmers are better off not using insecticides and we support the conclusion of Way and Heong (1994) that pest management “should be based on the contention that insecticides are not needed rather than that they are, and that ‘pests’ should now be critically re-assessed and proven guilty before insecticide use is contemplated”.
Way, M.J. and Heong, K.L. 1994. The role of biodiversity in the dynamics and management of insect pests of tropical irrigated rice – A review. Bulletin of Entomological Research, 84, 567-587.