• Boll Sampling to Predict Lint Yield in Upland and Pima Cotton

      Unrah, Bryan L.; Norton, E. R.; Silvertooth, J. C.; Silvertooth, Jeff (College of Agriculture, University of Arizona (Tucson, AZ), 1994-03)
      Giving a cotton (Gossypium spp.) producer a method to predict lint yield, would be a useful management tool. The objective of this study was to determine if relatively simple measurements could be made near cut -out which could be used to adequately estimate lint yield for Upland (G. hirsutum L.) and Pima (G. barbadense L.) cotton. Data and samples were collected from the nitrogen (N) management study at Maricopa Ag. Center from two N treatments which were imposed on both Upland (var. DPL 5415) and Pima (var. S-7) cotton. The treatments were no added N and N added on an as- needed basis. Twenty hard -green bolls from the first or second fruiting positions were collected from each plot on 19 August 1993. The number of bolls expected to reach maturity prior to crop termination were then determined from five randomly selected plants in each plot. Measurements on each boll collected included fresh weight, diameter, number of locks, number of seeds, and dry seed cotton weight. Plant population was determined from early season stand counts. Seed cotton per boll was most highly correlated to boll weight for DPL 5415 and for Pima S-7 it was most highly correlated with boll diameter. These respective parameters were then used in linear regression to predict seed cotton /boll. Lint yield calculated from the regression models (using boll weight or diameter) and yield calculated from means of the data collected agreed quit well. Predicted yields from regression analysis overestimated the actual Upland yield by about 730 lb lint /acre and under estimated Pima yields to within about 150 lb lint /acre. It appears that this procedure has the potential to estimate lint yields to within about 150 lb lint /acre. However the sampling scheme will he refined especially in regard to estimation of plants /acre and bolls /plant which should improve yield estimate accuracy.