|Frederick, J - CLEMSON UNIV.|
Submitted to: Soil & Tillage Research
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: June 5, 2001
Publication Date: N/A
Interpretive Summary: Hard soils can restrict root growth and reduce crop yields in the US southeastern Coastal Plains. Root restrictions can be overcome when producers use specialized, expensive deep tillage to loosen the soil. Producers can make well informed decisions on whether to deep till or not if we can link knowledge of the condition of soil hardness at planting to potential yield losses. Deep tilling incurs the expense of 2.5 gal of fuel per acre and 0.25 hr of labor per acre while requiring less need for large 200 hp tractors. Not deep tilling permits soil strength to build up to the point that it can restrict root growth. This research (for corn) and other recent research (for soybean) used a combination of soil strength measurements at the beginning of the growing season and rainfall amounts at critical times of plant growth to predict yield loss for intensively managed corn and soybean. The research included treatments that had not been deep tilled and that had been deep tilled from one day to three years before planting. Treatments where deep tillage was more recent had softer soils and higher yields. For the coastal plain soil used in this experiment, predicted yield losses were 20 to 50 bu/a of corn and 16 to 26 bu/a soybean for each 10 atm increase in soil strength.
Technical Abstract: In many soils throughout the US, subsurface hard layers reduce yield by limiting root exploration of the profile. We evaluated the impact of reduced frequency of deep tillage (and thus increased penetration resistance) and timing of rain on corn (Zea mays L.) yield for a 38-cm row width management system. Treatments were either disked or not disked; treatments were also deep tilled from zero to three years before planting on a Goldsboro loamy sand (fine loamy, siliceous, thermic Aquic Kandiudult). Cone indices showed that disked treatments resulted in a pan at the 10- to 30-cm depth. Cone indices for disked treatments were greater than for non-disked treatments. Whether caused by disking or by reduced tillage frequency, each megapascal of mean profile cone index increase reduced corn grain yields by 1.1 to 2.4 Mg/ha. Cone index vs. grain yield linear regression relationships differed among years. Regressions for the three years could be combined into a single relationship by including rainfall during 42-56 days after planting (vegetative growth) and 70 to 98 days after planting (silking) to the relationship (R**2= 0.87). The same procedure was then applied to soybean (Glycine max L. Merr.) grown in the same plots for three previous years giving similar results (R**2= 0.73). For each year, grain yield decreased with increasing mean profile cone index. For multiple years, yield variations were successfully described by mean profile soil cone index and rainfall from selected times of the growing season.