Is it Necessary to Leave Border Effects in Field Experiments?

Abstract

The field experiment technique and statistical science determine the rules for conducting and evaluating agricultural research. Good planning, observation and measurement are essential to obtain rational results. In field trials, work is done in small areas but the results are recommended for large areas. The units where each transaction to be examined in the research are placed are called parcels. In the studies to be carried out on the cultivation technique, gaps are left between the parcels so that the processes do not interfere with and affect each other. In this case, the edge effect emerges, based on the possibility that the plants on the edge give different figures due to more water, lighting, and nutrient intake than the middle ones. The parts that are thrown from all over the parcels, excluded from the evaluation, that is, the parts where the number is not obtained, are called "Border Effect". In our country, there were studies on the need to leave an edge effect, but it was noticed that this issue was not emphasized in the following years. For this reason, in this article, a statistical comparison of the effects of the plants in the margins on some field crops and the plant observations in the parcel net areas was made. In the study, 7 different cultural practices were tried in 3 blocks; plant height, number of branches, fresh fruit yield, number of fruits in fresh harvest, number of pods in the plant in dry harvest, dry grain yield and 100 seed weight were observed in the pod trial, which was sowing in 5 rows in each parcel, which was left with 1 m between the parcels and 1,5 m between the blocks. Except for the weight of one hundred grains, it was determined that the data in the edge effect were statistically higher (P<0,01). In a study in which 7 different applications of 2 microelements were tried in the wheat plant, spike length, number of grains in the spike, spike weight, and seed weight observations were made in 10 plants from the middle and edge rows of the plot. It was determined that the edge effect on the length and weight of the spike had a statistical effect of being the net area (P<0,01), but these values were higher in the net area. Whether higher or lower data are obtained, the necessity of leaving margins in trials can also be seen from these data. The new genotypes developed and the change in the techniques applied to show that more information is needed about edge effects in different plants.