A Statistical Study on Agricultural Factors Related to Paddy Cultivation in Sri Lanka

Abstract

Rice has been the staple food of Sri Lanka for centuries. The continuously increasing nation's population increases the demand for rice. Recent statistical evidence shows that in year 2004 nation's paddy yield has declined compared to year 2003. Therefore, increasing production of paddy is vital for the social-economic development of the nation. I have undertaken this research on the yield of paddy production as my dissertation for my M.Sc Degree in Applied Statistics where I attempt to suggest ways and means of increasing the yield of paddy production by changing the factors, such as type of irrigation, methods of land preparation,. . etc, used in paddy production. In order to actualize the objectives described above, I have obtained the data from the Department of Census and Statistics. The data belong to the Maha and Yala season from years 1976 to 2004 were used for identifying trend in the paddy production. The other set of data belong to year 2002/2003 Maha season and 2003 Yala season were used to identify the influential factors in the paddy production. One crucial aspect treated in this dissertation is forecasting yearly yield in terms of Maha and Yala season. In order to achieve this objective a statistical analysis was carried out, using Iterative Time Series Modelling (ITSM) computer software package where the minimum AICC technique was applied in the ARIMA model. The forecasting yields of paddy production obtained from the study were checked with years 2003 - 2004 data. These outcomes were within the 95% confidence interval. Finally, the forecasting value derived for year 2005, with respect to both Maha and Yala seasons were 78.71 bushels per acre, 77.51 bushels per acre; 79.85 bushels per acre. 78.75 bushels per acre for year 2006; 81.03 bushels per acre and 79.97 bushels per acre for year 2007. It is clear that the above production per acre in the Maha and Yala season is not enough to feed to the entire nation. So, I decided to identify the influential factors to increase yield of paddy production Therefore, the second set of data belong to year 2002/2003 Maha season and 2003 Yala season were used in the Gamma model for the purpose of determining influential factors of the paddy production. I have found some influential factors in this analytical study. These influential factors are major irrigation, minor irrigation, use of tractor, use of buffaloes, hand weeding and impounding water, use of weedecides, insecticides, chemical fertilizer, organic fertilizer, season, broadcast sowing method and districts. Out of these influential factors broadcast sowing method and certain districts are negative influential factors. The above statistical analysis indicates that yield could be increased subject to variation of the controllable factors in paddy cultivation process. The controllable factors are method of land preparation, application of chemical fertilizer, application of organic fertilizer, application of insecticides and use of weedicides.