LONG-TERM TILLAGE EFFECTS ON SEED BANKS IN 3 OHIO SOILS

Soils from long-term tillage plots at three locations in Ohio were sampled to determine composition and size of weed seed banks following 25 yr of continuous no-tillage, minimum-tillage, or conventional-tillage corn production. The same herbicide was applied across tillage treatments within each year and an untreated permanent grass sod was sampled for comparison. Seed numbers to a 15-cm depth were highest in the no-tillage treatment in the Crosby silt loam (77 800 m-2) and Wooster silt loam (8400 m-2) soils and in the grass sod (7400 m-2) in a Hoytville silty clay loam soil. Lowest seed numbers were found in conventional-tillage plots in the Wooster soil (400 m-2) and in minimum-tillage plots in the Crosby (2200 m-2) and Hoytville (400 m-2) soils. Concentration of seeds decreased with depth but the effect of tillage on seed depth was not consistent among soil types. Number of weed species was highest in permanent grass sod (10 to 18) and decreased as soil disturbance increased; weed populations were lowest in conventional tillage in the Hoytville soil. Common lambsquarters, pigweeds, and fall panicum were the most commonly found seeds in all soils. Diversity indices indicated that increased soil disturbance resulted in a decrease in species diversity. Weed populations the summer following soil sampling included common lambsquarters, pigweeds, fall panicum, and several species not detected in the seed bank.