Two long-term continuous cotton (Gossypium hirsutum L.) field experiments were initiated in 1990 at Chickasha (dryland), and Altus, OK (irrigated). Seven weed species were either transplanted or seeded in uniform densities in 1990. Weed and N fertilizer management levels were initiated in 1991. Weed management systems included low (no herbicides); medium (Treflan (trifluralin), preplant incorporated and Caparol (prometryn), banded preemergence); and high (same as medium except other treatments including postemergence herbicides (Fusilade 2000 (fluazifop), hand hoeing, and Pix (mepiquat chloride)). Routine weed control measures within the entire experiment were provided by cultivations. Nitrogen fertilizer management included 40 lb N/acre applied as ammonium nitrate (AN, 33-0-0), urea ammonium nitrate solution (UAN, 28-0-0), or urea ammonium nitrate plus 1% w/w dicyandiamide (a nitrification inhibitor) (UAN+DCD, 28-0-0). The AN was broadcast preplant and incorporated, while UAN and UAN+DCD treatments were knifed into the seedbed 6 inches to the side of the row prior to planting. A stripper-type cotton cultivar, 'Paymaster HS-26', was planted at both sites.

Lint yields were affected by weed management levels at both sites in all years. Chickasha results indicate higher lint yields with increasing weed management levels in all three years. Increases in lint yields at Chickasha at the high weed management level (compared to medium level) are attributed mainly to reduction in competition from devil's-claw [Proboscidea louisianica (Mill.) Thellung] by hand hoeing. High weed management did not increase lint yields above medium management at Altus in the first two years of the experiment, however, increased lint yields were observed in 1993, with greater yields under high weed management resulting mainly from reduced morningglory (Ipomoea spp.) competition by hand hoeing.

When averaged over weed management levels, 40 lb N/acre increased lint yields only at Chickasha in 1992, with no significant differences attributed to source/placement effects. Greater increases in lint yields were obtained under high (compared to medium) weed management as a significant weed management by N management interaction was observed at Chickasha in 1992. Nitrogen fertilizer rate responses for lint yield were noted under high weed management at Chickasha in 1992 and 1993 and at Altus in 1993. In short-season cotton production environments, such as the Rolling Plains resource area, N fertilizer response is seasonally dependent and may only be realized if weed competition is kept to a minimum.