INCREASING PRECISION OF FIELD EXPERIMENTS IN MAIZE USING UNIFORMITY TRIALS | ||||
| Journal of Plant Production | ||||
| Article 4, Volume 29, Issue 3, March 2004, Page 1065-1076 PDF (185.25 K) | ||||
| Document Type: Original Article | ||||
| DOI: 10.21608/jpp.2004.238515 | ||||
 View on SCiNiTO | ||||
| Author | ||||
| F. Ashmawy, | ||||
| Cent. Lab. For Design & Stat. Analysis Research, ARC. | ||||
| Abstract | ||||
| The present study aimed to find out optimum plot size. shape and the proper number of replications in maize experiments. Two uniformity trials were carried out at the Agricultural Research Station of Sids, Beni-Sowef Governorate during the successive growing seasons of 2002 and 2003 using maize cultivar Three Way Cross 310 (T.WC 310). The actual area of each field trial was divided into 8 strips, each consisted of 100 rows of 3.0 m long and 70 cm apart. Each row was considered as a basic unit i.e 2.1 m2. Consequently, a total of 800 basic units was utilized for each trial in the two seasons. Optimum plot size was estimated using Smith method and maximum curvature procedure. Increasing plot size decreased variance per basic unit and C.V. However, the reduction was not in proportion with the increase in plot size. Index of soil heterogeneity was 0.5430 and 0.6679 for the first and second seasons, reppectively, indicating moderate variation in the soil. Increasing plot size from one basic unit to 400 basic units reduced C.V from 20.187 to 4.925 respectively. in the first season and correspondingly from 26.731 for one basic unit plot size to 3.582 for a plot size of 400 basic units in the second season. The exponential relationshi~s between C.v and plot size (X) were C.V = 20.187X.o·m and C.v = 26.731x.o3 5 for the first and second seasons, repectively. Optimum plot size was one and two basic units (112000 and 111000 fed) for both seasons, respectively, using Smith method while it was 3 and 5 basic units (1/667 and 11400 fed) using maximum curvature method. Plot shape had no clear effect in all cases. However, long and narrow plots were more efficient as it decreased variance per basic unit and C.V. Increasing plot size generally decreased number of replications required to detect differences of 15 % and 20 % of the mean. | ||||
|
Statistics Article View: 36 PDF Download: 143 |
||||
