Abstract.Data from 15 rice‐fish experiments, conducted at the Freshwater Aquaculture Center of Central Luzon State University. Philippines during the years 1976‐1983. were analysed using the multiple linear regression technique with the objective of explaining the variation in fish production from input and climate data. The experiments dealt with the effects of various management practices on the production of Nile tilapia,Oreochromis niloticus(L.), in concurrent culture with several varieties of lowland irrigated rice,Oryzasativa(L.). Climate data were obtained from the nearest weather station.Significant models (P<0.001) were derived for gross fish yield (kg/ha), net fish yield (kg/ha), fish recovery (), fish growth rate (g/day) and rice yield (kg/ha). For gross fish yield, 66 of the variation could be explained by seven independent variables: length of the culture period, fish stocking weight and density, nitrogen and phosphorous application rates; insecticide application and air temperature.The interpretation of the regression results is discussed. Despite the problems with linking regression coefficients to ‘real life’ processes, this kind of exploratory data analysis, applying modem statistical techniques, can be a useful tool in summarizing the results of previous research and obtaining more information from existi
展开▼
