Published online by Cambridge University Press: 10 July 2009
Field experiments conducted on cotton in southern Sri Lanka, where the main pests were Heliothis armigera (Hb.), Earias vittella (F.) and Pectinophora gossypiella (Saund.), indicated that a damage action threshold of five squares and six bolls per 30 plants could be used for forecasting and early warning in pest management. A parallel sampling study indicated that a sample size of 30 plants would reduce the sampling error to acceptable levels as well as being a more convenient number in the field. An expression that corrects experimentally established thresholds for changes in the cost of control and the market price of the yield is given.