Optimum plot size, shape and number of replications for field experiment on wheat (Triticum aestivum L.) 

Abstract

Field experiment on wheat crop was conducted during rabi-2012 at Junagadh. The crop yield data of 1296 basic units (Basic unit = 0.90 m⨯ 1.00 m) were recorded and analyzed to work out optimum size, shape and number of replications of the plot using Fairfield Smith's , maximum curvature methods and Cochran & Cox equation respectively. The per unit decrease in C.V. % for different plot sizes estimated from the differences between C.V.% of consecutive points indicated that the effective decrease in C.V.% due to increase in plot size was up to 12 units sized plot. The graphical presentation of C.V.% also revealed similar trend, therefore, 12 units size i.e. 10.80 sq. m. size could be considered as optimum plot size. The Smith's model was considered as a measure of correlation between adjacent units with "b" = 0 indicating perfect correlation between adjacent plots. The medium range of "b" obtained in this study indicated the use of smaller plots in field experiments. The equation obtained was, Y = 11.08 ⨯^-0.0384. The coefficient of determination (R²) between plot size and C.V.% was computed as 0.36. The results revealed that the smaller plots with fairly large number of replications were observed to achieve 5 per cent accuracy in any of the block size. The increase in plot size from 0.90 m² to 10.80 m² resulted in reduction of 14 to 2 or 3 to 2 replications, respectively, to obtain the same precision. For controlling soil variation, a block of 3 ⨯ 3 plots was required with minimum 2 replications to achive 5 per cent accuracy for field experiments in wheat crop in Saurashtra region.