| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a USDA-ARS, Northern Grain Insects Research Laboratory, 2923 Medary Ave., Brookings, SD 57006
b USDA-ARS, Northwest Irrigation and Soils Research Laboratory, 3793 N. 3600 E., Kimberly, ID 83341, USA
* Corresponding author (jpikul{at}ngirl.ars.usda.gov)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
444 mm of water in 15 d. Results reveal a difficult soil management problem. Subsoiling initially improves infiltration, but no additional water storage was discernable after 15 d. Further, excess water percolation has potential to leach nitrate-N from the profile.
Abbreviations: 
v, soil water measurements • 
b, soil bulk density • NoSS, not subsoiled • SS, subsoiled • SSplus, subsoiled plus secondary tillage • PR, soil penetration resistance • SSI, soil subsidence index • SSR, soil strength response
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
Shallow soil pans can impede water and gas movement. Occurrence of pans may be a consequence of soil type or management or both. Many soils of the Southeastern Coastal Plain of the USA are weakly structured and prone to soil compaction (Busscher et al., 1986; Sojka et al., 1990). These loamy sands are easily compacted when exposed to traffic or other consolidation forces. Clay loams of the Southern Great Plains (Unger, 1993a) have dense Bt1 and Bt2 horizons that limit water infiltration and plant root penetration. Silt loams of the Pacific Northwest frequently have shallow layers of high strength that have been implicated as the cause of reduced water infiltration (Pikul et al., 1990) and poor plant performance (Sojka et al., 1993). Repeated shallow tillage of sandy loam soils in eastern Montana has elevated bulk density (b) and PR at a depth of 10 cm (Pikul and Aase, 1995). These pans are commonly referred to as tillage pans and are a consequence of pressure applied by normal tillage operations. The cases cited are but a small sample of typical pans that are mentioned in the literature. In each case, the cause of the pan and the climate were different, but tillage was sought as the means to loosen the pan and improve water relations and crop performance.
Objectives were to (i) determine effect of subsoiling on water infiltration and storage and (ii) evaluate longevity of soil structure created by tillage following three wetting and draining cycles.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
Annual precipitation at the research site was 357 mm, with 283 mm (80%) occurring during April through September. Plots were laid out on portions of the farm that had been managed in a fallow–wheat cropping sequence since about 1975. Customarily, primary tillage was in the spring with a tandem disk or with medium-crown sweeps, 0.45 m wide, operated 0.1 m deep. Tillage for summer fallow was with sweeps at 0.1 m deep and rod weeder.
Subsoiling
Experiment 1 was a randomized complete block, with three replications. Tillage treatments were: (i) SS with paratill, (ii) subsoil with paratill and secondary tillage with disk (SSplus), and (iii) NoSS. Infiltration plots were 15 m long and 12 m wide. The paratill subsoiler had four shanks (two left facing and two right facing) with point spacing at 0.66 m. Our implement was similar to that illustrated by Unger (1993b)(Fig. 1 and 3). Speed was 1.3 m s-1 and depth of subsoiling was 0.3 m. Disk tillage was 0.1 m deep. Tillage was conducted in early September and field measurements were terminated at the end of October.
Experiment 2 was a randomized complete block with four replications. These plots were established outside of the area used for studies in Year 1 for Exp. 1. Customary fallow tillage was deferred to avoid disturbance of standing wheat residue from the previous crop. Herbicides were used to kill plants on the infiltration plots. Tillage treatments were: (i) ripped with parabolic subsoiling shanks to a depth of 0.3 m (SS), (ii) ripped and tilled with subsurface sweeps (SSplus), and (iii) NoSS. Sweep tillage was 0.1 m deep. Tillage was conducted in mid-May and all field measurements were completed by the end of July.
Water Infiltration
A Palouse rainfall simulator (Bubenzer et al., 1985) was used to apply water at a rate of 40 mm h-1 to 1.16- by 1.16-m infiltration frames. Electrical conductivity of Missouri River water (Culbertson, MT, municipal water supply) used for the infiltration tests was 0.7 dS m-1, concentration of cations was 0.157 g L-1, and SAR was 13.6. This simulator was designed to mimic low intensity rainfall characteristics of the inland Pacific Northwest. Typical summer rainstorms in the northern Great Plains are high intensity and short duration. The Palouse simulator produces drop sizes that are 1.3 to 1.8 mm diameter. By comparison, natural rainfalls with intensities of 50 mm h-1 have drop sizes that are 1 to 5 mm in diameter (Wischmeier and Smith, 1958). Therefore, the test soil was not exposed to rainfall energy that exceeded that of naturally occurring storms.
Infiltration frames were constructed of heavy gauge steel. Frames were placed so that the path of only one subsoiling shank was within each infiltration frame. To install a frame to a depth of at least 0.3 m, we carefully dug a trench around the outside of the frame. As layers of soil were removed, the infiltration frame was forced downward to enclose an undisturbed soil monolith. Particular attention was given to soil outside the frame in the vicinity of the soil disturbance created by subsoiling shanks. This soil was removed, backfilled, and packed to eliminate lateral flow of water from inside the frame to outside the frame. Inside edges of the infiltration frames were sealed with bentonite clay to prevent any water leakage along the metal-soil interface. Frames were installed on each replication of each tillage treatment.
Surface conditions within infiltration frames on each treatment were protected from rainfall energy of naturally occurring storms during the course of both Exp. 1 and 2. Calibration pans were placed over the infiltration frames when rain was likely. Following infiltration tests, the frames were kept covered to minimize evaporative water loss.
Water application rate from the rainfall simulator was measured at the start and finish of each infiltration test by collecting the water from a 1.35-m2 calibration pan placed over the infiltration frame. Application rate was calculated as the average of these measurements. Water infiltration was calculated as the difference between application rate and runoff rate. Ponded water (water that would have run off) was removed from within the infiltration frame by vacuum. Water was applied for 3 h on Day 1, 2, and 3. The soil drained for 20 h following each water application. Tests on NoSS for Exp. 1 in Year 1 were left unfinished because of the onset of winter weather. Thus, only duplicate measurements for Day 1 and 2 of infiltration tests are shown for Exp. 1.
Profile Water Content
Soil water content was measured using neutron attenuation in Exp. 2. Equipment was calibrated as described by Pikul and Aase (1998). Within each infiltration frame, a permanent access tube was installed, enabling volumetric soil water measurements to a depth of 2.0 m at 0.1-m increments. A ring of bentonite clay around and overlain with a small mound of soil prevented water from ponding and running down the soil-tube interface during infiltration tests. Soil water content was expressed as an average of four replications for each tillage treatment. Measurements were made before and after each of the three infiltration runs and at 1, 5, and 15 d following the last water application.
Penetration Resistance, Soil Water, and Bulk Density
Soil PR was measured with a Soiltest CL-700 pocket penetrometer for Exp. 1. This penetrometer has a blunt tip of 6.35-mm diameter. Before use, the penetrometer was calibrated against a load cell (Bradford, 1980) and measurements are presented as soil PR. Penetration resistance profiles to a depth of 0.44 m were obtained by excavating a trench perpendicular to the direction of tillage. This trench was dug through the area of the infiltration test 7 d after the last water application. On the trench wall, facing the infiltration test area, a 25-mm grid was established having width (horizontal component) of 0.42 m and length (vertical component) of 0.44 m. Within the grid, we measured PR at 306 points.
Samples for b and gravimetric water (Gardner, 1986) were obtained by extracting a 25-mm core horizontally from a freshly prepared face of the trench wall. The sampler had a cutting tip of 19.6 mm and three cores were bulked per depth. Sample depths were centered at 0.013 m and at 30-mm increments from a depth of 0.05 m to a depth of 0.38 m. Cores were taken from the path of the subsoiler. Gravimetric water content was determined on each increment. Volumetric water content was calculated as the product of b and gravimetric water content.
Penetration resistance for Exp. 2 was measured with a 30° cone penetrometer that had a base area of 645 mm2. Measurements were taken within each frame at spatial-intervals of 0.10 m along each 0.9-m transect (nine positions). A depth profile of PR was obtained at increments of 0.075 m to a depth of 0.38 m (six depths). In each frame there were two transects oriented perpendicular to the direction of tillage, and penetrometer resistance measurements for like depth and transect positions were averaged.
Soil b and gravimetric water were each measured on the same day as PR. Bulk density samples were taken with a tube sampler described by Allmaras et al. (1988). The sampler had a cutting tip of 19.6 mm. Four cores were taken vertically within each infiltration frame in positions not disturbed by ripping. Each 0.30-m core was cut into ten 0.03-m depth increments. Gravimetric water content was determined on each increment and volumetric water content was calculated.
Soil subsidence is defined as the downward movement of the soil surface caused by the collapse of underlying tillage-induced macroporosity. Tilled soil subsides due to factors acting on the soil surface or within the tilled layer (Onstad et al., 1984). A single-value index of soil subsidence was calculated using PR data arrays from each treatment (Pikul and Aase, 1999). Surface maps using transect position, depth, and PR were prepared for each plot sampled in Exp. 1 and 2. Briefly, a three-dimensional plot of PR data yielded a response surface showing spatial variation of soil strength. Smoothed surface maps were prepared using three-dimensional surface plotting (Surfer 8, Golden Software Inc., Golden, CO). Default geostatistical gridding method of kriging was used to smooth PR data arrays. Standard deviation of residuals [PR(raw data)-PR(smooth data)] provided a way to evaluate the fit of each map to the original data.
Soil strength response (SSR), in units of force, was calculated for each smoothed surface in a manner analogous to calculating volume. Strength response was defined as the product of (transect length, m) x (soil depth, m) x (cone index, Pa) and had the units of force (N). Surfer software provides numerical integration procedures for calculating volume. Relative error was estimated by comparing numerical results from three numerical integration methods.
Nondimensional soil subsidence indices (SSIs) were calculated as:
 | [1] |
In a previous study, SSI was used to estimate soil subsidence 2.5 yr after subsoiling (Pikul and Aase, 1999). Positive values from that study ranged from 0.049 to 0.314. A normal probability plot of SSI using the Ryan-Joiner goodness-of-fit test (Minitab Statistical Software, Minitab Inc. State College, Pa) suggests a normal distribution (p = 0.1) of SSI values.
Values of final infiltration rate, cumulative infiltration, change in soil water content, and subsidence index were tested for significance using ANOVA and LSD.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
b and volumetric water for SSplus, SS, and NoSS treatments in the top 0.4 m are shown in Fig. 1
(Exp. 1) and Fig. 2
(Exp. 2) after the third water application. Comparison of Fig. 1 (NoSS) and Fig. 2 (NoSS) shows that there was a zone of maximum b at 0.10 m on the test plots for both Exp. 1 and 2. This maximum value was 1.64 and 1.74 Mg m-3, respectively, for Exp. 1 and 2.
|
|
10 g kg-1. We think that the zone of maximum b at 0.1 m depth was a consequence of the historic use of shallow sweep tillage on this farm and not a consequence of soil textural differences.
|
b in tillage studies because unconsolidated soil having large clods is difficult to sample. Soil b is necessary for volumetric water determination, and volumetric water is necessary for interpretation of PR measurements. We used two methods to obtain soil samples from the tillage zone. In Exp. 1, samples were extracted horizontally after excavating a pit. In Exp. 2, profile samples were extracted by vertical sampling. The methodology used in Exp. 1 provided a way to accurately measure b within the path of subsoiling (Fig. 1). Soil b at 0.14 m 7 d after three wetting and drainage cycles was 1.44 Mg m-3 for SSplus, 1.15 Mg m-3 for SS, and 1.64 Mg m-3 for NoSS treatments. Results suggest that soil, within the subsoiled zone under SSplus, consolidated following three artificial rainstorms. Our water infiltration measurements support these observations.
Vertical sampling (Exp. 2) in the rip zone was not possible because of the difficulty in obtaining a representative soil core. Samples for this experiment were taken 0.20 m off-center of the subsoiled rip. There were no differences in soil b among treatments (Fig. 2). In comparison of SSplus and SS to NoSS, these measurements suggest that the soil was not loosened 0.20 m either side of the parabolic subsoiling shank (Fig. 2).
Differences in soil water content can cause differences in soil strength. Our soil water measurements (v) and b measurements were taken concurrently with PR measurements. Profiles of v (Fig. 1 and 2) show little difference among treatments. Differences in PR are therefore expected to be a consequence of soil subsidence or soil loosening.
Penetration Resistance
Penetration resistance measurements were used as an index of soil subsidence following three consecutive artificial rainstorms. Measurements were mapped as a three-dimensional surface plot to visualize tillage induced soil structure and as a way to quantify changes in structure as a consequence of repeated wetting and drainage. Examples of these maps for Exp. 1 are shown in Fig. 4
for soil under NoSS and Fig. 5
for soil under SS.
|
|
b shown for NoSS in Fig. 1 and Fig. 2 corresponds to the depth of maximum PR. Together, measurements of PR and b suggest that the tillage pan is a layer of low total porosity rather than a layer of high strength due to cementation of the soil fabric.
Tillage-induced structure provides preferential water flow paths which are important to maintain rapid water infiltration. As will be shown, either paratill subsoiling (Exp. 1) or ripping with a parabolic subsoiling shank (Exp. 2) created flow paths by fracturing the tillage pan. An example of a surface map of transect position (x), soil depth (y), and PR (z) for SS in Exp. 1 is shown in Fig. 5. Measurements followed three artificial rainstorms. This figure shows the cross-sectional area disturbed by the paratill. In this figure, the chisel portion of the subsoil shank is centered at about transect position 0.22 m. Low PR values outline a zone of soil that was fractured by the subsoil tool. This zone is roughly in the shape of a vee, extending upward and outward to the surface from a depth of 0.30 m (Fig. 5).
A single-value index of soil subsidence (SSI, Eq. [1]) was calculated using surface maps of the type shown in Fig. 4 and 5. For the illustrated case on Exp. 1, SSR for NoSS (Fig. 4) was 151 253 and 90 460 N for SS (Fig. 5), resulting in an SSI of 0.40 for the SS treatment. In the case of SSplus (not shown) SSI was -0.026. Values of SSI approach zero as differences between NoSS and tillage treatment become less. The combination of secondary tillage and three wetting and drainage cycles on the SSplus treatment were enough to destroy structure created by subsoiling.
On Exp. 2, before water application, average SSI was 0.37 under SS and 0.30 under SSplus, showing that sweep tillage following subsoiling reconsolidated soil disturbed by tillage, and likely destroyed some surface-connected macropores. Following three artificial rainstorms, average SSI decreased to 0.206 on SS and 0.188 on SSplus (significant at P = 0.088). These results indirectly suggest that there were differences in tillage-induced soil structure between the two tillage treatments following the simulated rainstorms. Surface tillage (sweeps) on SSplus produced a relatively smooth surface that slaked quickly, as evidenced by reduced SSI and, as will be shown, water infiltration.
Water Infiltration and Storage
Measured water infiltration rates support local observations. Early in the growing season, runoff is rarely seen; by midsummer, however, runoff can be severe on smooth tilled fields after high-intensity thunderstorms. Infiltration measurements (Table 1) show that final infiltration on all treatments decreased with each subsequent artificial rain. On the SSplus, Exp. 1, final infiltration rate decreased from 34.6 mm h-1 on Day 1 to 6.2 mm h-1 on Day 3. Similarly, in Exp. 2, final infiltration decreased from 37.8 mm h-1 on Day 1 to 3.8 mm h-1 on Day 3. Final infiltration rate on SS for Day 3 was nearly five times greater than SSplus in Exp. 1 and three times greater than SSplus in Exp. 2.
|
b and PR, provide evidence that tillage-induced preferential flow paths, continuous with the soil surface, likely contributed to the greater water infiltration under SS treatment. For tests conducted in both Exp. 1 and 2, there was a marked decrease in water infiltration on Day 2. Tests on Day 1 are difficult to interpret because differences in profile water content among treatments affect water infiltration. Following water application on Day 1 (Exp. 2), water content in the top 1.83 m was 516 mm on SSplus, 521 mm on SS, and 501 mm on NoSS (Fig. 8)
. The decline in infiltration on Day 2 (Exp. 2) likely reflects a change in pore space configuration at the surface because soil water content was nearly the same among treatments.
|
|
|
Measured change in soil water content (Exp. 2) following each infiltration test was generally less than measured cumulative water infiltration (Table 1). For example, on Day 2, cumulative water infiltration (water addition) under SS was 69.8 mm, and measured change in soil water of the top 1.83 m was only 38.6 mm. Results were similar for other tillage treatments and suggest that internal soil water drainage was very rapid. Regardless of large differences among treatments in cumulative water infiltration (Table 1), soil profile water content (top 1.83 m) returned to 480 mm within a day following water application on all treatments (Fig. 8). Average drainage rate was 1.4 mm h-1 following infiltration tests on Days 2 and 3. During the following 4 d, water drained from the profile at a rate of 0.23 mm h-1. Subsequent drainage (Days 7 to 15) was at a rate of 0.09 mm h-1 (Data not shown in Fig. 8) with a final measured water content of 444 mm on Day 15 (data not shown in Fig. 8).
Water infiltration (Table 1) and drainage characteristics (Fig. 8) provide evidence that runoff and deep percolation occur rapidly on the Williams loam. These data have important bearing on the conduct of water balance studies (crop water use). Neutron access tubes are often used to measure soil water status at the beginning and ending of a cropping period. Soil water status and precipitation provide the information required to calculate crop water use or soil water storage. Often, out of necessity and for lack of better information, researchers (including the authors of this report) claim that water runoff and deep percolation were negligible (Aase and Pikul, 2000; Pikul et al., 2000). Knowing the upper drained limit of a soil as suggested by Aase and Pikul (2000) and as shown here can help identify those rain events where runoff and drainage might not be negligible.
On a large-scale crop rotation experiment involving 76 plots on this same research farm (Williams loam), soil-water use was measured under six crop rotations across 5 yr (Aase and Pikul, 2000). We found that spring soil-water content (top 1.8 m) at the end of 5 yr in each rotation sequence was nearly the same as that measured in Year 1. Experimental evidence led us to the conclusion that this soil can hold only 450 mm of water in the top 1.83 m (0.25 m m-3). Results from the current small-plot infiltration experiment support this finding.
Soil developed from glacial till has large variations in soil texture (Aase and Pikul, 2000) that causes variability in soil water storage and hydraulic conductivity. Water draining past the root zone has implications for development of saline seeps and leaching of excess nutrients (Brown et al., 1983). On annual wheat vs. wheat–fallow rotation on this farm we found that plots under wheat–fallow had 257 kg ha-1 more NO3–N in soil between the depths of 1.2 and 3.4 m than plots under annual wheat with spring tillage (Pikul and Aase, 1994, unpublished data). Primarily, spring wheat extracts the bulk of its nutrients and water from the top 1.2 m, so nutrients at depths greater than 1.2 m may be unavailable for spring wheat. Our results show that soil water in excess of 444 mm in the top 1.83-m profile drained quickly and this drainage could account for NO3–N leaching from the upper profile and accumulation of nutrients in lower portions of the profile.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Received for publication April 1, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. L. Baumhardt, O. R. Jones, and R. C. Schwartz Long-Term Effects of Profile-Modifying Deep Plowing on Soil Properties and Crop Yield Soil Sci. Soc. Am. J., May 1, 2008; 72(3): 677 - 682. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Machado, S. Petrie, K. Rhinhart, and R. E. Ramig Tillage Effects on Water Use and Grain Yield of Winter Wheat and Green Pea in Rotation Agron. J., January 11, 2008; 100(1): 154 - 162. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. M. Sainju, T. Caesar-TonThat, A. W. Lenssen, R. G. Evans, and R. Kolberg Long-Term Tillage and Cropping Sequence Effects on Dryland Residue and Soil Carbon Fractions Soil Sci. Soc. Am. J., September 28, 2007; 71(6): 1730 - 1739. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Pikul Jr., J. K. Aase, and V. L. Cochran Water Use and Biomass Production of Oat-Pea Hay and Lentil in a Semiarid Climate Agron. J., January 1, 2004; 96(1): 298 - 304. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
0.05.
0.5 h (ANOVA, P 
