A three-stick sampling technique for data collection in direct-seeded rice experiments

B. H. Yap , M. Azmi and P. M. Chang

Abstract

An experiment to evaluate the innovative use of a three-stick sampling procedure to determine the locations of sampling units was conducted. The final sampling units from which data were obtained, comprised quadrats of sizes 25 cm x 25 cm and 15 cm x 15 cm. Four sampling plans and four seed rates were evaluated. The data collected were for the variables total paddy/weed seedling counts (early stage), number of paddy plants, total number of green leafhoppers and white-back planthoppers, total number of predators and weed counts. An analysis of variance was done on each of the variables for each quadrat size. The findings revealed that the systematic and random sampling plans incorporating the use of the three-stick sampling procedure and a 5 m x 5 m wooden frame were practical and would ensure a more objective approach to data collection. The main plot, sub-plot and sampling variances for the 15 cm x 15 cm quadrats were much smaller than the values for 25 cm x 25 cm quadrats, except insect counts. Ways to simplify the sampling procedure using a triquadrat and a four-sided 5 m pole instead of the 5 m x 5 m wooden frame are suggested. The paddy and weed seedlings in the plots were found to be clumped or clustered as opposed to the random or uniform patterns desired. This discrepancy was mainly due to the unsatisfactory land preparations carried out by the farmers.

Full Text ( 40 KB )