Interest in starter fertilizers has increased recently due to reports of increased cotton yields in some studies. Responses to starter fertilizers, however, have been inconsistent due to differences in soils, environmental conditions, methods of application and other factors. Research is needed to identify the soil and environmental conditions that affect starter fertilizer responses and to determine the optimum methods and application rates for different conditions.

Two studies were initiated on a Gigger silt loam soil at the Macon Ridge Branch Research Station in 1991 to evaluate application rates and methods of ammonium polyphosphate (11-37-0) as a starter fertilizer for conventional-till (CT) and no-till (NT) cotton. The soil test P levels were high in both tests. Starter fertilizer application methods were: 1) in-furrow sprays of 11-37-0 directed into the seed furrow, 2) banded application of 11-37-0 two inches to the side and two inches below the seed (2 x 2 placement), and 3) spraying a four-inch wide surface band of 11-37-0 over the row behind the planter. Rates for the in-furrow treatments were 1.5, 3.0, and 4.5 gallons/A of 11-37-0. These treatments were diluted with water and applied in five gallons/A total volume. The 2 x 2 and surface band treatments were applied at 7.5 gallons/A of undiluted 11-37-0. All starter fertilizer treatments were supplemented with preplant broadcast applications of ammonium nitrate, triple superphosphate, and potassium chloride to provide total fertilization rates of 80-40-60 (lb/A of N-P₂O₅-K₂O). Two additional treatments were included for comparison, preplant broadcast fertilization at rates of 80-0-60 and 80-40-60, respectively. Broadcast fertilizers were incorporated with a bed conditioner prior to planting in the CT test.

No significant differences in growth, yield, or earliness were noted between the no P and broadcast P treatments in the CT and NT tests. In both tests, plant growth measurements taken 5 to 7 weeks after planting indicated significant increases in leaf area/plant, shoot weight/plant and plant height with the 2 x 2 treatment compared to broadcast fertilization. Other treatments generally had no effect on early plant growth. However, the surface band treatment was usually intermediate between the 2 x 2 and broadcast treatments. Stands were significantly higher in the 2 x 2 starter treatment compared with broadcast fertilization under both tillage systems. All in-furrow starter treatments reduced stands significantly with CT, but not with NT. No significant yield differences occurred among treatments in the CT test. In the NT test the surface band and 2 x 2 starter treatments increased lint yields by 81 and 146 lb/A compared to no P fertilization. Furthermore, the surface band treatment significantly outyielded the broadcast P fertilization treatment by 85 lb/A or 7.8%. Preliminary results indicate that surface banding or 2 x 2 placement of N + P starter fertilizer may be useful in improving early growth and yield of cotton on soils testing high in P, especially with NT. This study will be continued in 1992 and 1993.