RESPONSE OF NO-TILLAGE AND CONVENTIONAL-TILLAGE COTTON TO STARTER FERTILIZATION ON LOESS SOILS

Starter fertilizers have been utilized to improve cotton (Gossypium hi rsutum L.) yields, but yield increases have differed with soil type, a pplication method, application rates, and tillage. Starter fertilizer tests were conducted from 1991 through 1993 on a Gigger silt loam (Typ ic Fragiudalf) in Louisiana and on a Loring silt loam soil (Typic Frag iudalf) in Tennessee to evaluate methods and application rates of 11-3 7-0 liquid fertilizer for cotton. Treatments were evaluated under conv entional-tillage (CT) and no-tillage (NT) production systems. Applicat ion methods included in-furrow application at planting (IF), spraying a 4-inch wide surface band behind the planter (SE), and banding fertil izer two inches to the side and two inches below the seed at planting (2x2). The IF treatments were applied at 1.5, 3.0, and 4.5 gal/A. The SE and 2x2 treatments were applied at 7.5 gal/A. Starter fertilizer tr eatments were supplemented with broadcast granular fertilizers to achi eve a total fertilization rate of 80-40-60 (N-P2O5-K2O lb/acre). Start er fertilizer treatments were compared to broadcasting 80-40-60 and 80 -0-60. Soil test levels for phosphorus (P) were high on both soil type s. In-furrow applications of 3.0 and 4.5 gal/acre usually reduced plan t population for both tillage systems on both soil types. The 1.5 gal/ acre IF treatment was less detrimental to stand establishment than the higher IP: rates. The effect of starters on plant height varied from year to year. Starter fertilizers usually did not affect early-season plant height relative to broadcast treatments. In several experiments starter fertilization increased plant height compared to one, but not both, of the broadcast fertilization treatments. High IF rates (3.0 an d 4.5 gal/acre) reduced plant height in one experiment. The 2x2 starte r fertilizer treatment increased leaf area per plant relative to the b roadcast fertilization treatments in two of six experiments on the Gig ger soil. Responses for NT and CT studies were similar. bra one CT exp eriment on the Loring soil, the 1.5 gal/acre IF treatment increased le af area relative to the broadcast treatments. Lint yield responses to starter fertilization were inconsistent. Starter fertilization increas ed lint yield in one of six experiments on the Gigger soil and in two of six experiments on the Loring soil. Increased yields from starters varied with year and application method, however, the 2x2 and SE treat ments tended to provide better responses than IF treatments. Starter f ertilizer responses for NT and CT tests were generally similar.