The effect of variety, planting date, spacing, and seed treatment on cotton yields and stands

Five-year average results of a multiple-factor field experiment with cotton cultural factors including varieties, planting dates, spacings, and seed treatments at the Georgia Agricultural Experiment Station are reported in this paper. This type of experiment was found to have a number of advantages over the common single-variable field experiments where only one factor, such as variety or stand, is varied. Of the five varieties tested, Half and Half and Stoneville 2 were the highest yielders when all planting dates were averaged. The varieties responded differently to the different dates of planting, indicating that the date of planting of varietal experiments may have a marked effect on the relative yields of the varieties and the value of the results for making cotton varietal recommendations. The varieties showed some significant differences in their ability to produce satisfactory stands, especially when planted early or during late March at Experiment, Ga. Of the three planting dates studied, the early (late March) planting and the medium (late April) planting produced the same 5-year average annual yields of lint. The late (late May) cotton yielded only about one-half as much lint as was harvested from the earlier plantings. Stands were poorest on the earlier planted cotton and thickest where the seed was sown in May. Three-year average results are reported with the Stoneville 2 variety spaced as follows in 3 1/2 foot rows: (1) Not thinned, hills 1 foot apart; (2) not thinned, hills 3 feet apart; and (3) thinned to two plants, hills 1 foot apart. Two plants 1 foot apart produced the highest yields, except for the medium planting date, when the unthinned cotton in hills 1 foot apart made as much cotton as the thinned cotton spaced 1 foot apart. Contrary to common belief, the unthinned cotton was relatively poorer when late planting was done than for the other dates. The number of hills or the distribution of the plants rather than the total number of plants per acre is most important in spacing experiments. The results show that planting date may affect the ratings of different spacings in cotton spacing experiments. Treating the Stoneville 2 cotton seed with Ceresan had no significant effects on stands or yields under the conditions of this experiment. Correlation coefficients are presented to show the relationships between annual yields of lint of the five varieties and the total number of plants and hills per acre. Where the cotton was planted early, although the stands were somewhat irregular during some seasons, there was no significant correlation between the yield per acre and the total number of plants per acre and only a slightly significant correlation between yield and number of hills per acre. The results for late planted cotton, where stands were good every year, were similar to those for early planting. The medium planting showed a highly significant, positive correlation between lint yield and both number of plants and hills per acre. The number of hills on the variety plats like those on the spacing plats was more closely and positively correlated with the number of pounds of lint cotton produced than with the number of plants per acre. These results indicate that in studying cotton stands the number of hills may be more important than the number of plants per plat or acre.