Optimum sampling size for straw and grain yields and plant height in experimental plots of wheat

Abstract

Selection of a representative sampling area for accurate and reliable yield evaluation of wheat ( Triticum aestivum L.) is important for optimizing straw retention in soils and its removal as biofuel feedstock. A field experiment was conducted at four wheat fields in the North China in 2020. Wheat straw and grain yields varied from 359.2-682.8 and 338.5-640.4 g m -2 , respectively, and plant height varied from 51.3-59.7 cm across all plots in the four sites. Variation in either relative deviation (RD) or standard error (SE) of straw and grain yield and plant height estimates decreased with the increase in random sampling square (RSS) (one square = approximately 1 m²) and random sampling plant (RSP) numbers, respectively. Minimum RSS numbers of 3-10 and 1-10 m 2 were needed to satisfy RD less than 5% in two-third of the plots for straw and grain yield estimates, respectively. This suggests that 10 m 2 could be recommended as the minimum RSS number per plot. However, the incidence frequency of RD was 63.75% and 60.00% within the RD interval of 0%-5% for straw and grain yield estimates, respectively, from the RSS number of 10 m 2 , indicating that yields from RSSs in field trials are prone to large variations. Therefore, it is strongly recommended to design a large plot as possible and to harvest the whole plot for estimating yields. The threshold RSP number ranged from 14-18 to satisfy RD less than 1.5% and a minimum RSP of 20 plants (including mains and tillers) per plot could be recommended for wheat field experiments.