Abstract

Experimental recognostic trial was carried out in spring barley agrophytocenoses, evaluating data standard error (SE) and mean accuracy (MA%) depending on arranged elementary trial plots, treatment and replication number, their model size. The received experimental data are applicable carrying out all agronomical and biological field trials. Trial field was divided into 222 elementary plots of 13.44 m². The trial objective was agrophytocenoses of spring barely Roland. Observation number increasing from 2 to 6, trial data accuracy demonstrably increases, while increasing it from 6 to 10 accuracy increases more slowly, and increasing the number of observations from 10 to 30, accuracy almost does not change. Enlarging plot size from 13.44 to 134.4 m², the accuracy of experimental data increased. However, experimental field was homogeneous enough. Increasing replication number, trial SE regularly decreases independently of the arranged treatment number in the trial. Modelling trial with 40 m² size plots with 2 to 3 replications, data accuracy reaches 10.2 to 6.2%, so trials can be acknowledged as inaccurate. Increasing replication number from 4 to 6, data accuracy decreases from 4.7 to 2.8%. Increasing replication number to 10, trial accuracy increases to 2.0%. Further increasing replication number from 10 to 37, data accuracy progressively increases from 1.5 to 1.6%. Therefore, carrying out field trials in moderate plots relating to 40 m², experiments should not have less than 3 to 4 replications. Optimal number of treatments, according to SE and accuracy evaluation, is between 4 and 7. Then the highest accuracy of experimental data is reached.

Key words: Elementary plot, number of observation, replication, treatment, recognostic field trial, standard error, data accuracy.