Minimum and Delayed Conservation Tillage for Wheat-Fallow Farming

Chapter 2, no. 27 – Conservation Tillage Systems and Equipment, May 2001

Authors: William Schillinger, Harry Schafer, and Bruce Sauer; Department of Crop and Soil Sciences; Washington State University, Dryland Research Station, Lind, WA

Abstract

Maintaining crop residue, clods, and roughness on the soil surface during summer fallow is critical for wind erosion control in the low-precipitation (< 12 in. annual) dryland wheat production region of the inland Pacific Northwest, USA. Conventional farming practices are intensive, involving eight or more tillage operations during the fallow cycle. The objective was to evaluate fallow conservation tillage management systems for soil water storage, residue retention, surface and subsurface soil cloddiness, surface roughness, wheat stand establishment, and grain yield during 6 years at Lind, Washington. The soil is a Shano silt loam. Treatments were: i) conventional (tillage), ii) minimum (herbicides and tillage), and iii) delayed minimum (herbicides and delayed tillage). Averaged over years, precipitation storage efficiency in the soil was 51, 54, and 57% over-winter, and 24, 26, and 26% at the end of the fallow cycle, for conventional tillage (CT), minimum tillage (MT), and delayed minimum tillage (DMT), respectively. Surface residue and surface clod mass were consistently reduced by 45% or more in CT compared with MT and DMT. There were no differences among treatments in seed-zone water content at time of sowing in September nor in grain yield in any year or when averaged across years. Results show that the long-term practice of minimum and delayed minimum tillage during fallow significantly increased surface residue and clod retention for erosion control with no adverse agronomic affects compared with conventional tillage.

Introduction

Conventional tillage practices in the semiarid (6-to 12-inch annual precipitation) wheat-summer fallow region of the inland Pacific Northwest often bury excessive quantities of residue and reduce soil cloddiness and surface roughness. Blowing dust from fallow, or newly-sown winter wheat after fallow, causes recurrent soil loss and air quality problems (Papendick, 1998). Conservation tillage has been advocated for several decades to control soil erosion (Horning and Oveson, 1962), but many wheat growers are reluctant to change tried-and-proven conventional tillage practices. Growers in the low-precipitation region generally do not practice no-till summer fallow because of increased evaporative loss of seed-zone soil water during the dry summer months compared with tillage (Hammel et al., 1981; Schillinger and Bolton, 1993). In Washington State, less than 25% of crop production is with conservation tillage compared with a national average of 36% (Conservation Technology Information Center, 1997). With Shano and Ritzville soils, that cover a wide geographic wheat-fallow area, clods are frequently pulverized during tillage (Schillinger and Papendick, 1997). By the end of the 13-month-long fallow period, the surface soil is often powdery and lacks residue cover.

Development and adoption of agronomically feasible and more environment-friendly fallow management methods are needed. The objective of the study was to determine the long-term effects of conventional (CT), minimum (MT), and delayed minimum (DMT) tillage systems during fallow on soil water storage, residue retention, surface and subsurface soil cloddiness, surface roughness, wheat stand establishment, and grain yield.

Materials and Methods

A 6-year tillage management study was conducted from August 1993 to July 1999 at the Washington State University Dryland Research Station at Lind, WA. Annual precipitation averages 9.61 inches. The Shano silt loam soil is more than 6 ft deep with less than 2% slope. The experimental design was a randomized complete block with three tillage treatments replicated four times. Individual plots were 60 by 150 ft, which allowed use of commercial-size farm equipment. Wheat and fallow phases of the study were present each year.

Tillage treatments and field operations

The three tillage management treatments were: (i) Conventional Tillage (CT) – standard frequency and timing of tillage operations using implements commonly utilized by growers; ii) Minimum Tillage (MT) – standard frequency and timing of tillage operations, but herbicides were substituted for tillage when feasible and a non-inversion “undercutter” V-sweep implement was used for primary spring tillage, and; iii) Delayed Minimum Tillage (DMT) – similar to minimum tillage except primary spring tillage with the undercutter V-sweep was delayed until at least mid-May. A complete list of field operations and timing for each treatment throughout the study are shown in Table 1.

With conventional tillage, post-harvest tillage was conducted in August of 1993, 1994, and 1995 with overlapping 14-in.-wide V-sweeps on 12-in. spacing at 5-in. depth to kill Russian thistle (Salsola iberica Sennen and Pau) by severing the tap root. Russian thistle was not present in August 1996, 1997 and 1998, thus post-harvest sweep operations were not required. Plots were chiseled in November, after the surface soil was wetted by fall rains, to a depth of 10 in. with straight-point shanks spaced 24 in. apart to create channels for controlling frozen soil runoff during the winter. In late February, 12 oz./a glyphosate (Roundup) herbicide was applied to control grass weeds. Primary spring tillage was conducted in mid-to-late March with two passes of a duck-foot cultivator at 5 in. depth with staggered 9-in.-wide blades spaced 7 in. apart with an attached 4-bar spring-tooth harrow. After high-residue-production years, primary spring tillage involved a single pass with a tandem disk with 22-in.-diameter blades to a depth of 5 in. (Table 1). Plots were fertilized with 40 lb/a anhydrous NH3-N in April with a shank applicator and rodweeded three times at 4-in. depth during late spring and summer to control Russian thistle and other broadleaf weeds. Soft white winter wheat (cv. Eltan) was sown in 16-in.-wide rows in early September all years with a John Deere HZ deep furrow drill. Deep-furrow drills are the standard for sowing winter wheat into carryover soil water in summer fallow in the inland Pacific Northwest.

Minimum tillage treatments were sprayed with a nonselective herbicide for post harvest control of Russian thistle instead of tillage in August of 1993, 1994 and 1995, but not in 1996, 1997 and 1998 when Russian thistle were not present (Table 1). In November, the plots were chiseled to depths of 10-to-16 in. with straight-point shanks spaced 48 inches apart (twice the shank spacing of conventional tillage). Chiseling was not conducted in 1996. A rotary “shark’s tooth” subsoiler that caused little residue disturbance and created one 16-in.-deep pit with 1 gallon capacity every 8 ft2 was used in 1997 and 1998 in lieu of chiseling. Glyphosate was applied in late winter followed by primary tillage at 5 in. depth with the undercutter equipped with overlapping 32-in.-wide V-blades spaced 28 in. apart. A rolling harrow was attached behind the undercutter to break up large clods and fill air voids. The plots were rodweeded three times at 4-in. depth during late spring and summer and fertilized with aqua NH3-N injected between every other row of the John Deere HZ deep furrow drill when sowing winter wheat in early September. Though sowing depth varied each year depending on soil water content, it was always the same for each treatment.

The delayed minimum tillage treatment was identical to the minimum tillage treatment except that: (i) plots were not chiseled or rotary subsoiled in 1996, 1997, and 1998, (ii) primary spring tillage was delayed until mid-May or early June, and (iii) only two rodweedings were conducted during late spring and summer (Table 1).

To alleviate rough areas and other problems associated with conducting all tillage operations in the same track, the rodweeder implement was pulled perpendicular or at an angle to plot direction over the entire experimental area whenever weeding was required in all three treatments. All treatments were sown at the same time. Winter wheat stand establishment failed in September 1994 due to insufficient seed zone water and all plots were resown to hard red spring wheat (cv. Butte 86) at 60 lb/a in 6-in. rows with a double-disc drill in March 1995. In-crop broadleaf weeds were sprayed with 1.5 pints of Bronate plus 0.1 oz. Harmony Extra per acre applied in March (winter wheat) or April (spring wheat) during tillering stage of wheat growth.

Table 1. Field operations for the three tillage management systems during the six fallow cycles (1993-1999).

Measurements

Water measurements in the 6 ft soil profile were made immediately after grain harvest in late July-early August (beginning of fallow), in March prior to primary tillage, and again in late August-early September just before sowing winter wheat. Soil volumetric water content in the 1-to 6-ft depth was measured in 6-in. increments by neutron attenuation. Volumetric water in the 0-to 1-ft depth was determined from two 6-in. core samples using gravimetric procedures (Gardner, 1986). In addition, volumetric water content in the seed zone was determined in 0.8-in. increments to a depth of 9 inches just before sowing winter wheat in early September of 1994, 1995, and 1996 using an incremental soil sampler (Pikul et al., 1979). All soil water measurements were obtained from three locations in each plot.

Surface soil cloddiness was determined at the end of the fallow cycle in 1994, 1995, 1996, and
1998 by measuring the diameter of individual soil clods within a one-yard diameter sampling hoop randomly positioned at three locations in each plot. Wheel tracks were avoided. All clods with diameters equal or greater than 2 inches were sorted into 0.5-in. size increments and the mass of each size group measured in the field with a battery-powered digital scale. Clod mass was not measured in 1997 because clod structure was dispersed during an intense rain shower. Subsurface soil cloddiness was measured by gently dry sieving 0.35 ft3 of soil from the 0-to 4-in. tillage mulch layer through stacked 2 in.2, 1 in.2, and 0.5 in.2 mesh screens. Clods not passing through each of the three mesh screens were then weighed. Subsurface clod measurements were obtained from the same one-yard diameter area where surface clods had just been removed, therefore surface clods greater than 2 in. diameter were excluded from subsurface samples. Oriented roughness in all plots was measured soon after sowing in September using the chain method (Saleh, 1993).

Surface residue remaining from the previous crop cycle was measured several times throughout the fallow period by clipping and gathering all aboveground dry matter within a one-yard-diameter hoop. Three samples were always obtained from each plot. Wheat straw and dead Russian thistle plants were separated, placed in paper bags, and allowed to air dry in a low-humidity greenhouse before weighing.

Winter wheat stand establishment was determined by counting individual plants in one-yard-long row segments 21 days after sowing. Grain yield was determined in mid-to-late July by harvesting a 20-ft swath through each 150-ft-long plot with a commercial-size combine and auguring grain into a weigh wagon.

Precipitation was recorded at a standard U.S. Weather Bureau shelter located < 0.5 mi from the experiment site (Table 2). A frost depth tube (McCool and Molnau, 1984) was installed in undisturbed winter wheat stubble near the weather shelter and freeze-thaw status of the soil was recorded daily throughout the winter.

Analysis of variance was conducted for total soil water content in the 6 ft profile, seed-zone water content, quantity of surface residue, surface cloddiness, subsurface cloddiness, surface roughness, wheat stand establishment, and grain yield. The procedure used to compare treatment means was Fisher’s protected least significant difference. All statistical tests were done at the 5% level of significance.

Table 2. Precipitation during six 13-mo-long fallow cycles compared with the 80-year average at Lind, Washington.

Results and Discussion

Surface Residue

Residue at the beginning of fallow (just after harvest) ranged from 2000 to 5000 lb/A during the 6-year period, and was lowest after a spring wheat crop (1993-1994 and 1995-1996 fallow cycles, Table 3). Russian thistle produced more dry matter by the time of grain harvest than dry matter for the spring wheat crop. The important contribution of dead Russian thistle plants as residue during fallow after low crop production years in this study has been reported (Schillinger et al., 1999). By using herbicides instead of tillage for post-harvest Russian thistle control, and non-inversion undercutter V-sweeps for primary spring tillage, significantly greater quantities of surface residue were maintained with MT and DMT when compared with CT on all sampling dates (after the onset of tillage) during all years (Table 3). Retention of sufficient residue for government farm program compliance (> 350 lb/a) during fallow was not a problem, even with CT, when the previous winter wheat crop produced 3500 lb/A or more straw.

Surface Clods, Subsurface Clods, and Roughness

Averaged across years, the mass of surface soil clods 2 in. or more in diameter at time of sowing in September was 9, 16, and 20 tons/a for CT, MT, and DMT, respectively (Fig.1). The DMT treatment generally resulted in the greatest surface clod mass because the undercutter V-sweep did little mixing or disturbance to the surface soil layer, which was dry when primary spring tillage occurred in mid-May to early-June. The greater mass of surface clods in 1996 compared with other years is likely due to extensive and prolonged freezing of the soil during the 1995-1996 winter (data not shown) which generally promotes a more stable clod structure. The skew treader tillage operations in March and May of 1998, to cut and incorporate large quantities of surface residue into shorter lengths, resulted in a major reduction in surface clods compared with other years when the skew treader was not used (Fig. 1). Quantities of blowing dust collected from portable wind tunnel (Pietersma et al., 1996) tests on soils at the Lind Dryland Research Station (Horning et al., 1998), and on the tillage management experiment at Lind(K.E. Saxton, unpublished data), were significantly reduced with increasing levels of residue, clods, and roughness.

Table 3. Quantity of surface residue as affected by conventional, minimum, and delayed minimum tillage during the 13-mo-long fallow cycle†. Values in parenthesis for 1993-1994 and 1995-1996 fallow cycles are the percentage of total dry biomass comprised of dead Russian thistle plants.

In the subsurface (0-4-in.) tillage mulch layer, masses of 0.5-to 1.0-in. and 1.0-to 2.0-in. diameter clods in 1995 and 1996 were greater in MT and DMT treatments compared with CT, but there were no differences for subsurface clods 2 in. and larger (Fig. 2). Growers feel that finely divided soil aggregates within the subsurface tillage mulch layer are desirable for retarding evaporative water loss from summer fallow, thus there is a perception that CT is required for optimum soil water conservation. Similar to surface clods, the skew treader implement used in 1998 reduced subsurface clod mass compared with previous years and there were no differences in subsurface clods among treatments (Fig. 2).

There were no differences in oriented soil roughness between MT and DMT measured just after sowing winter wheat. Both MT and DMT had rougher surfaces than CT, except in 1998, when there were no differences among treatments (data not shown), presumably due to use of the skew treader on all plots.

Soil Water Content

Figure 3 shows that there was little or no difference in soil water content among treatments immediately after grain harvest (beginning of fallow) during any year. Less than 6 inches of water remained
in the 6 ft soil profile at the beginning of fallow, except in 1993 (Fig. 3a) when 2.1 in. of rain fell in July after wheat had reached physiological maturity and was no longer extracting soil water.

Precipitation storage efficiency (SE) is defined as the percentage of precipitation occurring during the fallow cycle that is stored in the soil. Over-winter SE (from late-July to March) was extremely low (range 3 to 14%) in 1993-1994 (Fig. 3a) because: i) water from the 2.1 in. July rainfall event likely evaporated after beginning-of-fallow soil water measurements were obtained, and ii) only 4.1 in. of over-winter precipitation occurred, the least for any year of the study and well below the long-term average (Table 2). Over-winter SE was significantly improved with MT and DMT compared with CT in 1994-1995 (Fig. 3b) when several January and February precipitation events occurred when surface soil was frozen to depths as great as 12 inches (data not shown). The 16-in.-deep channels created by wide-spaced chiseling in MT and DMT likely allowed precipitation to infiltrate into unfrozen soil, whereas infiltration was probably impeded in the CT treatment because the soil was frozen below the 10 in. chiseling depth.

During the 1996-1997 fallow cycle, when no fall chiseling was conducted in the MT and DMT treatments, several winter precipitation events occurred on frozen soil, or on thawed soil overlaying a frozen layer, extending as deep as 9 inches (data not shown). The CT treatment stored significantly more water in the soil than MT or DMT (Fig. 3d). This was presumably due to frozen soil restricting water infiltration in non-chiseled soils while channels in the chiseled soils extended below the depth of frost. The 1995-1996, 1997-1998, and 1998-1999 winters were relatively open and mild, and there were no over-winter SE differences among treatments (Figs. 3c, 3e, and 3f). Averaged across 6 years, over-winter SE was 51, 54, and 57% for CT, MT, and DMT, respectively. These over-winter SE values are considerably higher (61, 62, and 65% for CT, MT, and DMT, respectively) if 1993-1994 is excluded from the data set.

Water remaining in the soil profile at the end of fallow cycle was always in the same relative ranking among treatments as measured over-winter, i.e., highest for MT and DMT in 1995 (Fig. 3b) and for CT in 1997 (Fig. 3d), but not different in the other years (Figs. 3a, 3c, and 3e). The 5-year average end-of-fallow SE was 24, 26, and 26% for CT, MT, and DMT, respectively.

There were no differences in seed zone water content among treatments in 1994, 1995, and 1996 (Fig 5). Sowing of winter wheat was attempted in 1994, but the deep-furrow drill could penetrate no deeper than 8 inches, which was not adequate to reach the minimum soil water content (11% by volume, Fig. 4a) required for emergence from deep sowing conditions on silt loam soils (Lindstrom et al., 1976). Seed zone water content was adequate in 1995 (Fig. 4b) and 1996 (Fig. 4c) as well as in subsequent years (comparative measurements not taken) in all treatments. Many growers feel that it is necessary to create a fine “dust mulch” on the soil surface to retard evaporative soil water loss from fallow during the summer. These data, however, show that the greater mass of surface clods (Fig. 1) and sub-surface clods (Fig. 2) in the MT and DMT treatments did not adversely affect seed zone water content compared with CT.

Stand Establishment and Grain Yield

Winter wheat seedling stand establishment was not affected by tillage treatment, except in 1996, when CT stands were greater than in MT and DMT (data not shown). Differences in 1996 are likely due to the larger quantities of surface clods in MT and DMT compared with CT (Fig. 1), as many seedlings were unable to elongate around clods that rolled into the furrow during sowing. These stand differences in 1996 did not affect grain yield.
There were no differences in grain yield among treatments during any year or when averaged over 5 years (Fig. 5). Due to generally favorable precipitation and growing conditions, grain yield averaged across treatments and years in this study was 52 bu/a compared with the long-term (30 year) average winter wheat grain yield at Lind of 35 bu/a. Janosky (1999), who conducted an economic analysis of this study over the 6-year period, reported that MT and DMT slightly out performed the CT system in terms of net returns over total costs.

Summary and Conclusions

1. Surface residue retention during fallow was consistently and significantly increased using minimum tillage (MT) and delayed minimum tillage (DMT) compared with conventional tillage (CT). When wheat straw production was low, the minimum quantity of surface residue (350 lb/a) required for highly erodible soils for government farm program compliance could not be achieved or was marginally met using CT, whereas ample surface residue was always present in all years with MT and DMT.
2. On average, twice the surface soil clod mass and a rougher surface was achieved with MT and DMT compared with CT.
3. Chiseling in late fall with straight-point shanks benefited over-winter precipitation storage efficiency (SE) in two out of six years when water runoff or snowmelt occurred on frozen soils, but had no effect during open winters when runoff was not a factor. Creating deep tillage channels at wide spacing in late fall with MT and DMT was equal to or greater for over-winter SE than with CT, where chisel shanks were more closely spaced and operated at a shallower depth. The best option to maximize over-winter water storage with minimal residue disturbance may be to create narrow pits about 16 inches deep with a long-tooth rotary subsoiler, with approximately one pit every 8 ft2.
4. Seed-zone water at the end of fallow (measured 3 years) was not affected by tillage treatment. This suggests that finely divided soil particles within the tillage mulch may not be as important for retarding evaporative water loss during the summer as previously thought. Rather, creating of an abrupt break between the tilled and non-tilled layer with primary spring tillage, which severs capillary channels from the subsoil to the surface, appears to be the dominant factor regulating over-summer evaporative water loss.
5. Delaying primary spring tillage until mid-May and beyond had no adverse agronomic affects compared with MT and CT. The late winter application of a nonselective herbicide provided excellent control of downy brome (Bromus tectorum) and broadleaf weeds until at least 1 May in all years in the DMT treatment. Downy brome, the most problematic grass weed in the region, was well controlled in all tillage treatments during all years of the experiment.
6. Surface residue in excess of 1800 lb/a at time of sowing in MT and DMT treatments plugged the deep-furrow drill in 1997. This problem was easily remedied by ‘blind sowing’ (with only the drill packer wheels) prior to actual sowing. Several implements, such as the coil packer, rotary hoe, and skew treader will bury, align, or otherwise cut straw to allow effective drill operation in heavy residue; but these implements also pulverize soil clods and, therefore, are not recommended in low-residue situations.

In conclusion, conventional tillage during fallow held no agronomic or economic (Janosky, 1999) advantages over MT or DMT in this 6-year experiment. The CT system had distinct environmental disadvantages, especially when straw production from the preceding wheat crop was less than 3,000 lb/a. This research showed that, with judicious use of herbicides, tillage operations during fallow can be effectively reduced from eight (CT) to as few as three (DMT). If MT and DMT fallow management were widely practiced on Shano soils in the Columbia Plateau of eastern Washington, it is reasonable to expect a sharp reduction in wind erosion and suspended dust emissions, with associated benefit to air quality. Minimum tillage and delayed minimum tillage practices, as outlined in this paper, can be implemented by wheat growers with little or no hardship to their livelihood.

References

Conservation Technology Information Center. 1997. 1997 National crop residue management survey. pp. 81. Cons. Tech. Info. Center, West Lafayette, IN.

Gardner, W.H. 1986. Water content. p. 493-544. In A. Klute (ed.) Methods of soil analysis. Part 1. 2nd ed. Agron. Monogr. 9. ASA and SSSA, Madison, WI.

Hammel, J.E., R.I. Papendick, and G.S. Campbell. 1981. Fallow tillage effects on evaporation and seedzone water content in a dry summer climate. Soil Sci. Soc. Am. J. 45:1016-1022.

Horning, L., L. Stetler, and K.E. Saxton. 1998. Surface residue and soil roughness for wind erosion protection. Trans. ASAE 4(4):1061-1065.

Horning, T.R., and M.M Oveson. 1962. Stubble mulching in the Northwest. Agric. Info. Bulletin No. 253. USDA-ARS and Oregon Agric. Exp. Station, Corvallis, OR.

Janosky, J.S. 1999. An economic analysis of conservation tillage cropping systems in eastern Washington. M.S. thesis, Washington State Univ.

Lindstrom, M.J., R.I. Papendick, and F.E. Koehler. 1976. A model to predict winter wheat emergence as affected by soil temperature, water potential, and depth of planting. Agron. J. 68:137-141.

McCool, D.K., and Molnau, M. 1984. Measurement of frost depth. p. 33-41. In B.A. Shafer (Ed.) Proc. Western Snow Conf., 17-19 April, Sun Valley, Idaho.

Papendick, R.I. (Ed.). 1998. Farming with the wind: Best management practices for controlling wind erosion and air quality on Columbia Plateau croplands. Washington State Univ. College of Agric. and Home Econ. Misc. Publ. MISC0208.

Pietersma, D., L.D. Stetler, and K.E. Saxton. 1996. Design and aerodynamics of a portable wind tunnel for soil erosion and fugitive dust research. Trans. ASAE 39(6):2075-2083.

Pikul, J.L., Jr., R.R. Allmaras, and G.E. Fischbacher. 1979. Incremental soil sampler for use in summer-fallowed soils. Soil Sci. Soc. Am. J. 43:425-427.

Saleh, A. 1993. Soil roughness measurement: Chain method. J. Soil Water Cons. 48(6):527-529.

Schillinger, W.F., and F.E. Bolton. 1993. Fallow water storage in tilled vs. untilled soils in the Pacific Northwest. J. Prod. Agric. 6:267-269.

Schillinger, W.F., and R.I. Papendick. 1997. Tillage mulch depth effects during fallow on wheat production and wind erosion control factors. Soil Sci. Soc. Am. J. 61:871-876.

Schillinger, W.F., R.I. Papendick, R.J. Veseth, and F.L. Young. 1999. Russian thistle skeletons provide residue in wheat-fallow cropping systems. J. Soil Water Cons. 54(2):506-509.