A three-year study was initiated in 1991 at the University of Arkansas Fayetteville Main Experiment Station to determine the extent by which allelopathic sunflower residues can augment or reduce herbicide use in no-till cotton and to determine how sunflower can be incorporated into a no-till cotton production system. A split-plot design was used, with sunflower residue management as the mainplot and weed control treatments as the subplot. Main plots used were four sunflower management levels: (1) fall-planted sunflower, (2) spring-planted sunflower, mechanically-cut or herbicide-desiccated, (3) companion crop, and (4) no sunflower. Subplots used were four weed control systems: (1) weed free by hoeing, (2) standard intensive herbicide program including glyphosate preplant, fluometuron + metolachlor preemergence, and post-directed applications or methazole or cyanazine plus MSMA as needed, (3) a total post, as-needed program including fluazifop-P for grass weeds and DPX-PE350 (Staple) for broadleaf weeds early in the season, with cyanazine optional at lay-by, and (4) a weedy check with no weed control. The experiments were conducted with the same treatments on the same plots for two years. Two weeks before cotton was planted, the experimental area was sprayed with glyphosate to kill existing vegetation. Data collected were cotton stand count, plant height and yield, weed counts and weed weights. Weed samples were taken during early- and mid-season using two 50 cm x 50 cm quadrats per plot.

In both years, no significant interaction effect was observed between sunflower residue management and weed control treatments. In 1992, significant reductions in total weed counts occurred during the first two to six weeks of cotton growth when residues from sunflower grown in the spring of the same year were used. Residues from sunflower planted in the fall of the previous year did not result in significant reductions in weed counts. Due to weather limitation, fall-planted sunflower was not established in 1992. There was no weed control advantage in planting sunflower as a companion crop and this treatment was not included in 1993. In 1993, sunflower residues showed no significant inhibitory effects on weed counts and weed weights due to intense weed pressure. In both years, the use of herbicides on a complete or as-needed regimen significantly reduced total weed density and total weed weight with or without sunflower residues. Both herbicide programs produced comparable seed cotton yields. This indicates some potential for limiting herbicide use on an as-needed basis under a no-till cotton production system.