Three annual applications of glyphosate [N-(phosphonomethyl)glycine] with three ropewick applicators were compared to hoeing and cultivation for the control of johnsongrass [Sorghum halepense (L.) Pers. ♯3 SORHA] in cotton (Gossypium hirsutum L. ‘Acala SJ-2′) in 1980 and 1981. The three applicators provided about equal control of johnsongrass. Averaged over treatments, years, and dates of application, plots treated with glyphosate contained 83% fewer live johnsongrass shoots (14/m2) than cultivated control plots (84/m2) 2 to 3 weeks after each application. Glyphosate-treated plots yielded an average of 81% more seed cotton (2010 kg/ha) than cultivated control plots (1110 kg/ha). However, yields of treated plots averaged 33% less than plots maintained weed-free by hoeing. Although applications of glyphosate reduced johnsongrass populations 83%, the competitive effect of johnsongrass on cotton prior to the initiation of treatments, plus the competition from escape plants between treatments, prevented cotton from yielding normally.
