Conservation-tillage systems are primarily used to address concerns about soil erosion, soil quality, and water availability (Blevins et al., 1971; Reeves, 1994; Reeves, 1997; Kaspar et al., 2001). Peanut (Arachis hypogaea L.) hectarage in conservation-tillage systems is estimated to be 45, 30, 25, 20, 10% respectively, in Alabama, Florida, Georgia, North Carolina, and Virginia (personal communications, Dallas Hartsog, Barry Brecke, Eric Prostco, David Jordan, and Joel Faircloth). Approximately 98% of peanut grown in Alabama, Florida, Georgia, North Carolina and Texas in 2004 received herbicides (Anonymous 2005). Practical alternatives to the intensive use of herbicides for peanut weed control offers economical as well as environmental benefits.

The use of cover crops in conservation tillage has advantages, one of which is weed suppression through physical and chemical allelopathic effects (Nagabhushana et al. 2001; Phatak 1998). Cereal rye and soft red winter wheat are common winter cover crops recommended for peanut production in the U.S. (Wright et al., 2002). These cover crops contain allelopathic compounds that inhibit weed growth (Akemo et al., 2000; Chase et al., 1991; Perez and Ormeno-Nunez, 1991; Yenish et al., 1996).

In southern Brazil, black oat is the predominate cover crop on millions of hectares of conservation-tilled soybean [Glycine max (L.) Merr.] due in part to its weed suppressive capabilities (Derpsch et al., 1991). Black oat has recently been introduced in the southeastern U.S. through a joint release between Auburn University and The Institute of Agronomy of Paraná, Brazil, and is currently marketed as “SoilSaver black oat” (Bauer and Reeves, 1999). In a greenhouse study, allelopathic compounds released from black oat inhibited cotton root elongation 16% compared to rye when residue was mixed with soil (Bauer and Reeves, 1999). However, in a field study where residue remained on the soil surface, cotton stand establishment was not affected by black oat, rye, or wheat winter cover crops and cotton lint yield was higher in plots containing black oat residue compared to rye (Bauer and Reeves, 1999). No other published research has been conducted evaluating black oat as a winter cover crop preceding row crop establishment in the U.S.

Typically, cooperative extension service recommendations in the southeastern U.S. encourage growers to terminate cereal winter cover crops relatively early prior to row crop production, citing concerns for excessive residue interfering with planting operations or excessive moisture depletion (Jost et al., 2006). Cooperative extension service recommend waiting approximately 2 wk after desiccating cereal winter cover crops before cash crop planting to avoid allelopathic effects on the following crop (Reeves, 1994).

The Brazilian conservation-tillage system is based on terminating cover crops during early reproductive growth, by treating with glyphosate and mechanically rolling the cover crop to form a dense mat of residue on the soil surface into which crop seeds are planted (Derpsch et al., 1991; Reeves, 2003). Technological advancements in cover crop residue management are allowing producers to maximize cover crop biomass in conservation agriculture systems located where water utilization by winter cover crops is not usually a concern (Kornecki et al., 2005; Raper et al., 2004). In the southeastern U.S., winter cereal cover crops reach anthesis and can be terminated in a timely fashion prior to the recommended planting window for peanut. Ashford and Reeves (2003) evaluated a mechanical roller-crimper as an alternative method for termination of a black oat, rye, and wheat cover crop. Their results indicated that use of a roller-crimper in combination with glyphosate at 0.84 kg ae/ha at anthesis was as effective as using glyphosate at 1.68 kg ae/ha alone when treatments were applied at the same crop growth stage, for all cover crops evaluated. Few growers are currently utilizing roller-crimpers to manage cover crops; however, grower interest in this management technique exists due to the potential to reduce soil erosion and increase water infiltration and soil moisture storage (Truman et al., 2002).

While research has evaluated weed-suppressive qualities of winter cover crops, few experiments have evaluated peanut response. Therefore, our objective was to evaluate weed control provided by black oat, rye, and wheat as winter cover crops within three herbicide input systems, compared to winter fallow, for conservation-tilled peanut using the Brazilian system of managing cover crops. Peanut yield was also evaluated for each cover crop and herbicide input system.

Field experiments were conducted from autumn 1994 through autumn 1997 at the Alabama Agricultural Experiment Station's Wiregrass Research and Extension Center (31°24′N, 85°15′W), located near Headland, AL. The soil was a Dothan fine sandy loam (fine-loamy, siliceous, thermic Plinthic Paleudults). The experimental area had been in conservation tillage (strip-tillage) for the previous 8 y and had a high population of Palmer amaranth (Amaranthus palmeri S. Watts.).

The experimental was set up as a strip-plot design with a factorial arrangement of treatments with four replications. Horizontal strips consisted of black oat, rye, wheat, or fallow. The seeding rate for all cereal cover crops was 120 kg/ha and were established utilizing a Great Plains^H 5 no-till drill6 in early November of 1994, 1995, and 1996. Ammonium nitrate was applied at 56 kg/ha to the cover crops in fall of 1994 and 1995 after establishment. Ammonium nitrate was not applied in 1997 due to an oversight. Cover crops were terminated 3 wks prior to planting peanut in early May each year with an application of glyphosate at 1.12 kg ae/ha utilizing a compressed CO₂ backpack sprayer delivering 140 L/ha at 147 kPa. Biomass from black oat, rye, wheat, and fallow plots was measured immediately before glyphosate application in all years. The above-ground portion of each cover crop and weeds in the winter fallow plots was clipped from three randomly-selected 0.25-m² sections in each plot, dried at 60 C for 72 h, and weighed.

Within 3 d following glyphosate application, the cover crops were rolled with a mechanical roller-crimper as described by Ashford and Reeves (2003) to flatten all residues on the soil surface. The peanut variety GK 7 was planted each year at a rate of 28 seed per meter of row with a four-row John Deere MaxiMerge^H planter7 equipped with Martin® row cleaners8 and ACRA-PLANT^H retrofit seeding double-disk openers9. Plots were four 92 cm rows wide and 9.1 m long.

Vertical plots were herbicide input systems consisting of no herbicide, a low input system consisting of preemergence (PRE) followed by (fb) postemergence (POST) herbicides, or a high input system consisting of PRE fb sequential POST herbicides. The PRE fb POST herbicide low input system consisted of pendimethalin at 1.12 kg ai/ha fb an additional early POST application of paraquat at 0.14 kg ai/ha plus bentazon at 0.56 kg ai/ha. The PRE fb sequential POST herbicide high input system contained the aforementioned herbicides fb 2,4-DB at 0.22 kg ai/ha plus chlorimuron at 0.14 kg ai/ha applied late POST.

In 1995, because the site had a well-developed hardpan, the experimental area was in-row sub-soiled prior to planting with a KMC^H narrow-shanked parabolic subsoiler10, equipped with pneumatic tires to close the subsoil channel (Raper, 2005). In 1996 and 1997, the area was in-row subsoiled prior to planting with a bent-leg subsoiler (Paratill11) 2 wks prior to planting. In all years, residue disturbance was minimal and residue formed a thick (approximately 10 cm) mat over the soil surface with exception of the winter-fallow plot treatment. Weed control was determined by visual ratings a 0%  =  no control, 100%  =  complete control) early in the season [30 d after planting (DAP)] and late in the season (51–98 DAP). Only late season ratings are reported. All weed species present at both ratings were evaluated for control, as a reduction in total above ground biomass resulting from both reduced emergence and growth, and the combined average for each rating and treatment was calculated. Late season weed control ratings were averaged over all dominant weed species.

Alabama Cooperative Extension System recommendations were used for insect control and nutrient management. Peanut yield was determined by machine-digging the middle two rows of each plot and harvesting with a combine.

All data were subjected to analysis of variance (ANOVA) using the general linear models procedure in SAS (SAS, 1998) to evaluate the effect of a three (herbicide input level) by four (cover crops) factorial treatment arrangement. Herbicide input levels and winter cover crops were considered fixed effects while year effects were considered random variables. Non-transformed data for visual evaluations were presented because arcsine square root transformation did not affect data interpretation. Means for appropriate main effects and interactions were separated using Fisher's protected LSD test at P ≤0.10 a priori. Where interactions occurred, data were presented separately and where interactions did not occur, data were combined.