Deep Ripping Fall-Planted Wheat After Fallow to Improve Infiltration and Reduce Erosion

Authors: William Schillinger, Agronomist with Washington State University Cooperative Extension at Ritzville, WA; Dale Wilkins, Agricultural Engineer with the USDA-ARS Columbia Plateau Conservation Research Center at Pendleton, OR; Roger Veseth, Extension Conservation Tillage Specialist with Washington State University and the University of Idaho at Moscow, ID.

Abstract

Water runoff and soil loss from winter fields are often severe during the winter when rain or snow melt occur on frozen soils in the inland Pacific Northwest (PNW). In a 2-year field study near Benge, WA, we tilled seeded wheat plots on slopes > 40% in late fall to a depth of 10 or 24 inches with shanks spaced 12 or 20 feet apart. In a dry winter, no soil loss was measured in ripped plots compared to 1.3 tons/acre soil loss for the control. Soil drying occurred near the tillage channels in ripped plots, reducing over-winter soil water storage. In a winter with higher than average precipitation and frequent frozen soil conditions, soil loss was 2.8 and 9.0 tons/acre for ripped and control treatments, respectively. Ripping significantly improved water infiltration into the soil to a depth of 6 feet as far as 3 feet down slope from the tillage channel. In both years, grain yield was reduced in the row most disturbed by the tillage shank, but was increased in adjacent rows. On a whole-plot basis, there were no differences in grain yield between ripped and control treatments either year. Results suggest that deep ripping seeded wheat fields in late fall is an effective soil and water conservation practice which does not reduce grain yield.

Introduction

Water erosion is often severe during winter in the dryland winter wheat production areas of the inland PNW. Precipitation in this region varies from 8 inches to 24 inches and characteristically 60% occurs between November and March. A winter wheat – fallow rotation is practiced on about 5 million acres receiving < 15 inches annual precipitation (Douglas et al., 1992).

Water erosion in the wheat-fallow rotation is most severe during the winter of the crop year because of the winter precipitation pattern, long steep slopes, very little ground cover from crop residue or wheat seedlings and low water infiltration rates through frozen soil. Soil freezing may occur to depths of 4 inches several times during the winter with occasional freezing to 16 inches (Papendick and McCool, 1994). Partial or complete soil thawing frequently occurs between freezing events. Soil loss from water erosion can be especially high when snowmelt or rain occur on thawed soil overlying a subsurface frozen layer. In north-central Oregon, Zuzel et al. (1982) reported that 86% of soil erosion on winter wheat was caused by rapid snowmelt or rainfall on thawing soil. Water infiltration rates of the silt loam soils in this region are about a half inches per hour, but approach zero when the soils are frozen (Zuzel and Pikul, 1987). Infiltration rate into frozen soil decreases with increasing soil water content at the time of freezing (Willis et al., 1961).

Erosion control practices include tillage, crop residue management, strip cropping, green cover (crop cover), contouring and terracing. Even the best management plans are often not sufficient to control soil erosion during events of rain and/or melting snow on frozen soil (McCool, 1990). Inclement weather conditions frequently limit the amount of crop residue available for erosion control. Dry seedzone soil conditions in the fall often reduce winter wheat stand establishment and consequently limit the amount of crop residue produced. Winter kill, drought, fire and disease also may result in insufficient residue for erosion control.

Additional management practices are needed for the Inland PNW to combat erosion events associated with frozen soil. Where soil freezing is common, growers routinely chisel or subsoil wheat stubble after harvest at the beginning of the fallow cycle (Ramig and Ekin, 1991), but do not employ this practice during the crop year. Pikul et al. (1992), using a rainfall simulator, found that contour ripping increased water infiltration into frozen soil during the crop winter. Saxton et al. (1981) reduced water runoff and erosion during the crop winter by filling soil slots with crop residue to prevent freezing and surface sealing. Tillage slots and cracks that extend from the surface to below the frozen soil layer form preferential flow paths. Surface water moves down these channels through the frozen layer and infiltrates into the permeable subsoil (Pikul et al., 1992). For these systems to be successful, a connecting water channel from the soil surface to below the frozen layer must be maintained during the time frozen soil/runoff events occur. Forming tillage slots at the time of seeding in late summer-early fall does not improve infiltration because the loose dry surface soil sloughs back into the tillage channel.

One solution to the problem of maintaining the tillage slot integrity during the critical erosion period is to till the soil after it freezes. Wilkins et al. (1991) developed a special tillage tool to create tillage channels through the frozen layer with minimum soil disturbance. This tool was used in northeast Oregon on frozen fields that were seeded to winter wheat. Frozen soil tillage destroyed wheat plants in and adjacent to the tillage slot but grain yields were not reduced due to stand reduction or plant disease (Wilkins and Zuzel, 1994). Although this technique of tilling when the soil is frozen is successful, there is a narrow window of time when tillage can be performed. Wilkins and Zuzel (1991) tilled when there was 2-to 4-inches of frozen soil. If the frozen layer is thin it will not support the tractor and ruts are formed and the soil compacted. The other extreme is a thick layer of frozen soil that requires excessive power to fracture and large chunks of soil are uprooted causing difficulty at harvest.

Another approach for providing tillage channels through frozen soil in wheat fields is to seed in the fall, wait for rain, and then make channels with deep tillage shanks prior to soil freezing. Fall rains moisten and firm the loose dry surface mulch created during summer fallow. We felt that tillage channels formed with these soil conditions would tend to stay open during the winter when frozen soil runoff occurs. It is also more practical than tilling frozen soil, allowing growers a longer time period in late fall to create the tillage channels. Harold Clinesmith, the grower cooperator for our study, reports he has reduced soil erosion for many years on his farm by tilling seeded wheat fields with a ripper fabricated in his shop.

The objectives of this research were to determine the effects of deep ripping fall-sown wheat on steep slopes prior to soil freezing on soil loss, water infiltration into the soil, and grain yield components and crop characteristics.

Materials and Methods

A 2-year on-farm experiment was conducted during the 1993-1994 and 1994-1995 crop cycles on the Harold Clinesmith farm near Benge, Washington. The farm receives an average of 13 inches annual precipitation and cropping pattern is a wheat-fallow rotation. Soil was a Walla Walla silt loam overlying basalt bedrock. The experimental design was a randomized complete block with six replications of two treatments: ripped and control. For the 1993-1994 crop cycle, the experimental site was a 41% north-facing slope. In 1994-1995, the site was a 43% north-facing slope. At both sites, soil was 3-to 4-feet deep along the backslope and extended to depths > 6 feet towards the base of the hill.

Seeding and Ripping Operations

In 1993, Hyak winter wheat (soft white club) was seeded on September 22 in 16-inch rows with deep furrow split-packer drills at a rate of 55 lb/acre along the contour of the hillside. Uniform wheat seedling emergence and plant establishment was attained. On December 13, ripped plots were established by tilling unfrozen soil about 11 inches deep with a single 1-inch wide shank with attached rotary subsoil spider (Wilkins et al., 1991). Contour tillage slots were established at 20 foot intervals along the slope. The ripper, pulled by a small crawler tractor, was lifted out of the soil when crossing control plots.

Seedzone soil water was insufficient for early fall seeding in 1994. Seeding was conducted on November 7 after fall precipitation had wet the soil surface. Eltan winter wheat (soft white common) was seeded in 6-inch rows with double-disc drills at a rate of 80 lb/acre. Ripped plots were contour tilled 24 inches deep with 2-inch thick shanks at 12 foot intervals along the slope on November 11. The tillage implement was a subsoiler which was modified by removing all but the end shanks. The implement was pulled with a Caterpillar D6 crawler tractor. Wheat seedlings had not yet emerged at the time of the ripping operation.

Water Infiltration, Soil Loss, and Yield Component Measurements

Soil water content was measured within 3 days of the ripping operation and periodically thereafter throughout both winters. Volumetric water content of the 1- to 6-foot soil depth was measured in 0.5-foot increments with a neutron probe. Water content of the 0-to 1-ft soil depth was measured gravimetrically, as described by Gardner (1986), in two 0.5-foot core samples. In 1993, neutron probe access tubes were installed 1, 3 and 5 feet down-slope of a tillage channel near the base of the hill. In 1994, four access tubes were placed in each plot: in the tillage channel, and 1, 3, and 5 feet down-slope of the tillage channel. Access tubes were placed in the same general lateral locations in control plots, i.e. where the tillage channel would have been if the ripper had not been lifted out of the soil when crossing control plots.

Soil loss from rill erosion during the winter was measured using the voided rill method (Everts and Riehle, 1980). Total cross-sectional area of rills near the top, middle and base of each plot were averaged to determine soil loss on a whole-plot basis. Precipitation, minimum-maximum air temperature, and soil temperature at 2, 4, 8, and 12 inch depths were recorded hourly throughout the study period at a weather station placed near the experimental sites. Head density was measured from hand-cut samples obtained from 3-foot row sections near each access tube in all plots at harvest in August. Clean grain yield, kernels per head, 1000 kernel weight, and dry matter were determined from these samples.

An analysis of variance was conducted for water content in 0.5 foot depth increments and for the total 6-foot soil profile on each sampling date, soil loss from rill erosion on each sampling date, and yield components and crop characteristics. Treatments were considered significantly different if the P-value was <0.05. Treatment means were separated by Fisher’s protected least significant difference.

Results and Discussion

Water Infiltration

The 1993-1994 winter was dry (Fig. 1) and water recharge into the soil was poor. There were only 3 days between December and February where precipitation exceeded 0.2 inches. In March, the soil near the tillage channel was significantly drier than in control plots but there were no differences in soil water content 90 cm down-slope of the tillage channel (Fig. 2).

Precipitation was greater than average and frequently occurred when the soil was frozen during the 1994-1995 winter (Fig. 3). By late January, treatment differences in water infiltration within 1 foot of the tillage channel were measured to soil depths of 5 feet (Fig. 4).

By late February, highly significant differences in soil water content were measured to 6 feet depth within 3 feet from the tillage channel (Fig. 5), but there were no differences in water content 5 feet from the tillage channel.

Soil water differences between treatments were slightly less pronounced in late March (Fig. 5 and Fig. 6) because March precipitation fell on unfrozen soils.Ripping does not generally improve water infiltration when the soil is unfrozen (Pikul et al., 1992). At harvest in August, available soil water was depleted in control plots but still present within 1 foot of the tillage channels, especially at deeper depths (Fig. 7). We were surprised to find these differences because winter wheat is an efficient water extractor, and speculate that diminished plant stand near the tillage channel reduced water demand.

Soil Loss

Ripping significantly reduced soil loss by rill erosion during both dry and wet winters. In 1994, most soil loss occurred during a 3 day period in early January when > 1 inch of rain fell on thawed soil overlying a frozen layer extending to 4 inches depth (Fig. 1 and Table 1). The remainder of the 1994 precipitation occurred on unfrozen soil (Fig. 1), causing little soil loss. In March, there were no measurable rills in the ripped treatment compared to a modest soil loss of 1.3 tons/acre in control plots (Table 1).

Recurrent precipitation on frozen soils during the 1994-1995 winter produced cumulative soil loss from rill erosion (Fig. 3 and Table 1). Ripping significantly reduced soil loss throughout the winter compared to control treatments (Table 1). Tillage channels generally stopped rills, whereas many rills extended the entire length of the hillside in control plots. No precipitation occurred on frozen soils after February. Rate of soil loss in ripped plots, low throughout the winter, had increased by early spring (Table 1), perhaps because the surface soil was saturated and tillage channels had filled with sediment by this time.

Table 1. Cumulative soil loss (tons/acre) from ripped and control plots during the crop cycle at the Benge, WA site in 1994 and 1995.

Yield Components and Crop Characteristics

Ripping decreased grain yield in the row closest to the tillage channel both years by reducing head density (Table 2 and Table 3). However, yield in adjoining rows as far as 3 feet from the tillage channel was increased compared to control plots (Table 2 and Table 3).

Ripping had no effect on grain yield 5 feet from the tillage channel either year. On a whole-plot basis, ripping did not significantly affect any yield component or crop characteristic either year (Table 2 and Table 3).

Conclusion

Creating deep tillage channels on contour in seeded wheat on steep slopes in the fall: (1) reduced soil loss by retarding rill erosion during both dry and wet winters; (2) increased water infiltration during the wet winter, and; (3) did not reduce grain yield either year. In areas of the PNW where soil freezing is common, growers routinely rip wheat stubble after harvest to reduce the risk of soil erosion and increase water storage during the fallow winter. Our research suggests that growers will benefit by employing the same practice, but with wider shank spacings, during the crop year.