Influence of plot size and row width on experimental error with lucerne
1991
Uzik, M. | Zofajova, A. (Vyskumny Ustav Rastlinnej Vyroby, Piestany (CSFR))
Four independent experiments with 4 experimental lucerne populations were carried out in different row width, row length and number of replications and their influence on the experimental error values was studied. The green matter weight was selected as a parameter in an analysis by plots and rows. The minimal significant deviation (MSD) has shown that the preferred combination in tests of identical numbers of genotypes carried out with identical test plots was smaller plots and higher number of replications compared to the opposite approach. Experimental error (EE) in the average percentage depended on the plot size whereas MSD on the number of replications; hence MSD was found to be the same in large plots and low replication numbers, and in small plots with a high repetition frequency. Comparisons with the experimental error for other parameters have shown that the smallest plots were suitable for high stands, with the EE only 3 percents. As for the number of stems per unit of area, errors in the range from 18 to 25 percent were found and they increased from 45 to 61 percent due to blank spaces in rows. Row-by-row analysis showed that in case when only limited quantity of seeds is available, the experiment should be organized in wide rows and long, or even single-row test plots. The wider rows and longer experimental plots were used, the lower was the ratio of variance in plots to experimental error values.
显示更多 [+] 显示较少 [-]