LONG-TERM EFFECTS OF DEEP TILLAGE C. L. Hooks, Southern Illinois University, Carbondale, Illinois R. E. Dunker and R. G. Darmody, University of Illinois at Urbana/Champaign Abstract The effects of seven tillage treatments ranging in depth from 9 to 48 inches applied to a reconstructed surface mine soil were evaluated over a ten year period beginning in 1988. The southern Illinois mine soil consisted of 8 inches of scraper-placed topsoil over 40 inches of scraper-placed rooting media. The pre-tillage physical condition of this mine soil is described as compact and massive. A nearby tract of Cisne silt loam (fine, montmorillonitic, mesic Mollic Albaqualf) was used as an unmined comparison. Significant differences in corn and soybean yield, soil strength, and net water extraction were observed among tillage treatments. Depth of tillage needed on the mine soil to achieve productivity comparable to permit target yields were found to be affected by initial levels of soil strength. Soil strength and depth of tillage were highly correlated to long-term yields. Introduction Poor soil physical condition has proven to be the most severe and difficult limiting factor in the reclamation of many prime farmland soils (Fehrenbacher et al., 1982). Newly constructed soils commonly lack a continuous macropore network necessary for water movement, aeration, and root system extension. Also, plant root growth is often severely inhibited by excessively high soil strength (Thompson, et al., 1987; Meyer, 1983). There are two sources of the physical condition problem in man-made soils. One is the use of severely compacted, high strength soil materials that are naturally present in the lower horizons of many southern Illinois soils. If this is not adequately disrupted in the excavation process, the soil may maintain high strength. This "transportation" of compaction is generally associated with scraper placed subsoils. In that process, large monoliths of intact subsoil are sheared out and folded into the scraper pan. The resulting subsoil is largely massive with interfaces between the monoliths and where they were broken and folded together. Mottling and other characteristics of the original soil remain detectable with varying degrees of distortion. Secondly, and more common with all placement methods, is compaction induced by earth moving equipment in the process of moving, placing, and grading the soil material. In natural soils, a physical condition problem can be improved by growing forage legumes for an extended period or at least within a crop rotation. Illinois has completed two experiments over the last ten years to evaluate its efficacy in solving the deep compaction problem of reconstructed soils. The practice, though having some merit, has proven inadequate. Soil strengths are commonly just too high to allow diffuse distribution of even alfalfa root systems. The roots tend to form mats in desiccation cracks and leave much of the soil volume largely unaffected. Physical improvement is slow, if detectable, especially in the lower horizon. Perhaps that should not be surprising, as severely compacted glacial till layers in some natural soils have also remained intact, even after one or two centuries of agriculture. A logical approach would be to reduce compaction by limiting the moving of soil materials to periods when they are dry. This approach has some merit, but is also inadequate. The reality is that the mines simply do not have that option. Experience has also shown that, even though moving materials dry does help substantially, the finished product still has excessive soil strength and bulk density. Research should continue to be directed towards finding soil construction methods that will prevent the problem, but meanwhile, means for amelioration of deeply compacted soils must be investigated. There are many tillage options that have been proven effective to 12 to 15 inches depth for ameliorating wheel traffic effects of farm machinery on undisturbed soils. Standard agricultural tillage equipment cannot reach the depths of the compaction problem in reconstructed soils. A deep ripper, the Kaeble Gmeinder TLG-12, which has an effective depth of 32 inches, has been tested in preliminary studies in southern Illinois (Hooks, et al. ,1987) and western Illinois (Dunker, et al., 1989). Results from both studies were very encouraging with significantly increased 87