ABSTRACT

A John Deere Pro 12 spindle picker row unit (In-line head) with modifications to cut one row and crowd it into a standing row located 15 inches to the left and passing directly through the picking unit was mounted on a single row picker chassis (John Deere Model 122, one row cotton harvester mounted on a JD 4020 tractor). The harvester was operated in cotton produced in eight different row configurations including 15, 30, and 38-inch row solid cotton, 15-inch row, 2 x 1 skip row, 15-inch row, 2x2 skip row, 30-inch rows with a 1 x 1 skip-row (cotton in 60-inch rows), 30-inch row, 2 x 1 skip row, and 38-inch row, 2 x 1 skip row. These row configurations were planted at the North Mississippi Research and Extension Center, Verona, MS, and at a private farm in Falkner, MS. Harvester operation was observed, yield and losses measured, samples ginned and lint sampled for AFIS and HVI quality determinations.

Wet soil condition through most of May delayed planting of the plots until 5/28/03. This resulted in unusually late and green plots with immature bolls immediately prior to defoliation. The wet spring conditions also resulted in irregular plant spacing within the row. Coupled with a cool September and early October and field moisture keeping the plants green; many of the bolls were not fully fluffed and dried at the time of harvest. The picking unit performed very well; however, the combination of these conditions and adjustments to the picking row unit resulted in the cut row stalks bunching at intervals, thus slug feeding the picking unit to a minor degree. Only one choke-up to the picking unit was noted while harvesting the plots and that occurred in the 30 inch, 2 X 1 skip row plot where no cut stalks were being harvested. That choke-up did not appear to be a result of the picking unit modifications. The picking unit handled a wider range of plant conditions that would have impeded the operation of a finger stripper.

The harvest simulation model, XLCOTSIM, was used to predict the impact of row spacing and machine performance on net revenue after harvest costs were deducted. Yield, fiber quality, harvest losses, and estimates of machinery costs were used in the model for each row pattern. The trend for both locations was for wider rows and lower yielding plots to have greater harvest losses. No statistically significant fiber quality differences were found between row patterns. Micronaire was lower at the Falkner location due to an early application of defoliation with few open bolls and was lower for wider row patterns with a greater number of immature bolls. This was the only variable in the quality measurements that impacted the value of the lint; causing a slightly lower value due to a discount for the later maturing wider row treatments. Trash and nep levels were consistent with spindle harvested seed cotton levels and would not be expected to create spinning performance difficulties for textile mills.