Soil Science Society of America Journal 64:710-714 (2000)
© 2000 Soil Science Society of America
DIVISION S-6-SOIL & WATER MANAGEMENT & CONSERVATION
Temporal Changes in Small Depth-Incremental Soil Bulk Density
S.D. Logsdon and
C.A. Cambardella
National Soil Tilth Laboratory, USDA–Agricultural Research Service, 2150 Pammel Dr., Ames, IA 50011 USA
logsdon{at}nstl.gov
 |
ABSTRACT
|
|---|
Farmers are concerned that soil compaction will be a problem in the first few years after conversion to strict no-till. The objective of this study was to determine if the changes in depth-incremental soil bulk density during transition to no-till were greater than density changes due to natural variation. We sampled six locations from a no-till field and six locations from a disk field of a soybean [Glycine max (L.) Merr.]–corn (Zea mays L.) rotation and took 12 samples at each of the 12 locations. The soil types sampled were Clarion (fine-loamy, mixed, superactive, mesic, Typic Hapludoll), Nicollet (fine-loamy, mixed, superactive, mesic, Aquic Hapludoll), and Canisteo (fine-loamy, mixed, superactive, calcareous, mesic, Typic Endoaquoll) soils. We divided each 300-mm soil sample into 15 depth-increments, and pooled the 12 samples for each location for each 20-mm depth-increment. Bulk densities ranged from 0.9 to 1.3 Mg m-3 in the top 0 to 20 mm, and 1.2 to 1.6 Mg m-3 for the depth-increments below 20 mm. Comparing the first and last sampling dates for the no-till field, no significant changes in bulk density were detected for any depth-increment; however, for the disk field, increases in significant bulk density were detected in 10 out of 15 depth-increments. Natural causes probably contributed to the bulk density changes that occurred over time in both fields. None of these bulk densities were high enough to restrict root growth. We concluded that concern about soil compaction under no-till management is inappropriate for these structured soils.
Abbreviations: COLE, coefficient of linear expansion
|
INTRODUCTION
|
|---|
MANY FARMERS ARE RELUCTANT TO SWITCH from disk or chisel systems to no-till practices because of concerns about compaction, especially in the early years after conversion. Bruce et al. (1990) observed that the top 200 mm of a sandy soil after 8 yr was more dense for the no-till treatment than for the disk treatment. After 10 yr on a silt loam, the no-till system soil was more dense than the disk–harrow system soil in the fall and spring (Horne et al., 1992). Averaging across rotations, soil in the no-till treatment was significantly more dense than soil in the disk treatment of wheat (Triticum aestivum L.) for 42% of the time at the 0- to 50-mm depth, 83% of the time at the 50- to 125-mm depth, and 21% of the time at the 125- to 200-mm depth (Franzluebbers et al., 1995).
Well-structured soils and soils with macropores and fractures provide a pore network for root growth (Stypa et al., 1987; Logsdon et al., 1992) and water infiltration, often resulting in no yield reduction, even when the soil is compacted (Voorhees et al., 1989; Lowery and Schuler, 1991; Logsdon et al., 1992; Kaspar et al., 1995). Other studies have shown that moderate compaction may benefit crop yield, especially during dry years (Graham et al., 1986; Johnson et al., 1990), because of better seed–soil contact and better soil continuity contributing to capillary rise of water to the root zone (Lipiec and Simota, 1994).
Documentation of management practices on soil density is obscured by natural variations in soil density (Blake et al., 1976; Voorhees and Lindstrom, 1984; Logsdon et al., 1992; Franzluebbers et al., 1995). Soil water content at the time of tillage or traffic, depth of winter freezing, water content before winter freezing, shrinking and swelling, and action of soil fauna and roots all cause variations in density, apart from the direct-management effect.
Small depth-incremental sampling has been used to examine the depth distribution of soil bulk density, residue distribution, and organic C content (Pikul and Allmaras, 1986; Allmaras et al., 1988; Staricka et al., 1991; Allmaras et al., 1996). These studies examined effects of long-term tillage, but the small incremental technique might be even more valuable for examining short-term effects of tillage system. The small depth-increments might detect trends that would be diluted in larger-increment samples. Sampling below 300 mm was considered unnecessary since the bulk densities are not expected to be different for different tillage systems below 300 mm (Logsdon et al., 1990).
There is a need for research in farmers' fields to complement field-plot research. Results from traffic-controlled studies at the plot-scale are hard to extrapolate to the field- or farm-scale. Equipment dealers simply do not make tractors, combines, and other field equipment with the same wheel spacing or swath width. We can retrofit our small plot equipment, but farmers rarely adjust their larger field equipment to have the same wheel spacing and swath width for all operations. The nonuniformity of equipment within a farmer's field causes a greater percentage of the field to be subject to wheel-traffic compaction, compared with controlled-traffic field plots.
The primary objective of this study was to quantify the temporal changes in soil bulk density of 20-mm depth-increments sampled to a depth of 300 mm during the first 3 yr after the change to no-till management in a farmer's field. Temporal changes in soil bulk density are also examined for a disk field to help distinguish density changes due to natural processes from management-influenced density changes. A second objective was to compare the no-till system with the disk system.
|
Materials and methods
|
|---|
Field Layout
This study was conducted on two fields in central Iowa, both on similar soils and farmed by the same operator. The fields were located 1.6 km apart, and each field or section of field considered was about 20 ha. Both fields had been under a disk management system until 1992 and were in a corn–soybean rotation with corn grown in even-numbered years. The two fields had been under the same management practices by the same operator with the same soil types before the study began. This does not preclude minor variations in soil properties between the two fields.
The disk management used deep disking to 180 mm in the fall after corn but not after soybean. One of the fields was converted to strict no-till after the corn harvest in 1992. No-till practices were maintained throughout the study until the end of the growing season in 1996. The other field was continued under the disk management system. Detailed cultural history for the fields is included in Tables 1 and 2
. The combine and tractor did not have the same wheel spacings. Since combine traffic is more closely spaced, combine traffic covered additional areas of the field not covered by previous tractor traffic. The rows were oriented the same every year, but they did not occur at the same location each year. This resulted in a variation in traffic location from year to year. Tillage, planting, and anhydrous applications caused additional soil disturbances. The disk field in the late spring had three to six trips through the field during this study compared with one to three trips in the no-till field (Tables 1 and 2). Most of the spring operations in the disk field included soil disturbance as well as tractor traffic, but most of the trips in the no-till field only involved tractor traffic. Disks can cause compaction below the depth of disking if the soil is too wet.
Small Incremental Sampling
We sampled two locations in each of three soil types for each field, for a total of 12 general locations. The soil types were Clarion, Nicollet, and Canisteo soils. The Canisteo sites were intermingled with Webster soils (fine-loamy, mixed, superactive, mesic, Typic Endoaquoll), which differed only in the absence of carbonates in the surface soil. Each sampling location was pooled from 12 individual samples taken within the general location. We took each sample to a total depth of 300 mm using the technique of Pikul and Allmaras (1986). The sampling tool had a relief cutting tip of 19 mm in diam., which screwed onto a 425-mm long cylinder (including the cutting tip). The cylinder had been welded on to a fitting that allowed pin-attachment of a handle. Because of the relief cutting tip, the inner diameter of the cylinder was slightly larger than the cutting tip, which greatly reduced friction between soil and tube. In addition, the sampling device was inserted into the soil using gentle pressure after it had been sprayed with cooking spray lubricant. Inspection of the samples showed that there was no compression of height compared with the sampling depth. We used a solid metal rod to gently push the sample out of the sampling cylinder on to a tray, with the top of the sample being the first to be pushed out of the sampling cylinder. The tray was marked in 20-mm increments to aid in cutting the 300-mm soil sample into 20-mm depth-increments. We cut each of the samples into 15 subsections, each 20 mm. For each 20-mm depth-increment at each location, we pooled the subsections from the 12 samples and stored them together in a soil moisture can.
We sampled the same six general locations (12 samples pooled each site) in each field on 18–20 Aug. 1993, 14 Apr. 1994, 24–25 Oct. 1994, 13 Apr. 1995, 13 June 1995, 11 Apr. 1996, and 16 Aug. 1996. We took additional samples on 15 Nov. 1994 after the disk treatment field had been deep disked.
Sample Processing
For bulk density determination, we air-dried the pooled subsections in the laboratory for 2 or 3 wk. Then we took a small sample from each pooled subsection to oven-dry and determine the air-dry water content. (We did not oven-dry the full pooled subsection because that would invalidate analysis of residue and organic C.) From the air-dry water content we determined the oven-dry mass; and we calculated bulk density from the oven-dry water content and known volume of the pooled subsection at the time of sampling (cylinder of 19-mm diam. and 20-mm length x 12 samples). Because soil water content influences bulk density, we determined water content at the time of sampling by weighing the pooled subsections before air-drying for the sampling dates of 13 Apr. 1995, 13 June 1995, 11 Apr. 1996, and 16 Aug. 1996.
For additional site characterization, we determined residue depth distribution for each pooled subsection of the 11 June 1996 sampling date, dispersed 40 g of each pooled subsection with sodium hexametaphosphate, and then shook them overnight on a shaker. Then we washed residue out of the dispersed, pooled subsections as they were submerged on a 1-mm sieve and expressed residue amount on an aerial basis. We also determined organic C and total N for each of the eight top-depth pooled subsections for sampling done on 16 Aug. 1996. After careful subsampling, we removed carbonates with 1 M H2SO4 (if necessary), and then determined organic C and total N using dry-combustion methods in a Carlo-Erba NA1500 NCS elemental analyzer (Haake Buchler Instruments, Paterson, NJ1)
. We expressed both organic C and total N on an aerial basis.
Additional whole-field properties (especially C and N pools) from intensive spatial sampling for these fields before the no-till management was started are discussed in Cambardella et al. (1994). The disk (pothole) field had been sampled spatially in April 1992, and the no-till field had been sampled spatially in October 1992 (Cambardella et al., 1994). Our sampling areas were on the other side of the no-till field from the section sampled by Cambardella et al. (1994). In addition, the two sides of the field were in alternate sequences of the corn–soybean rotation. The sampled areas of the fields had different starting points for organic C and total N (5.4 vs. 4.7 kg C m-2 and 0.45 vs. 0.42 kg N m-2) for disk and no-till fields, respectively (Cambardella et al., 1994). Mean soil bulk densities for the 0- to 150-mm depth were 1.32 Mg m-3 for the disk field and 1.14 Mg m-3 for the sampled area of the no-till field; and soybean-residue cover in the other side of the no-till field for the fall of 1992 was 514 g m2. We did not measure surface residue in this study because of nonuniform distribution. The intense rain and runoff in the spring and summer of 1993 washed the previous year's corn residue in the no-till field (the side sampled in this study) into piles producing very nonuniform residue distribution throughout the remainder of this study.
Statistics
The first type of comparison was between sampling dates within each field. Since we returned to the same six sampling areas in each field on each sampling date, we paired by sampling location to test for differences between sampling dates. We used the 95% confidence interval of the paired differences (Karlen and Colvin, 1992) to test for statistical significance. Confidence intervals all in the positive range would indicate significant increases, and confidence intervals all in the negative range would indicate significant decreases. For each depth-increment we tested for differences between adjacent sampling dates and between the first and last sampling dates; for the disk field, we also tested between the October 1994 and April 1995 sampling dates.
The second type of comparison was between tillage systems. Because the tillage systems were located in separate fields, the pairing mechanism was the three soils. We averaged the data from the two replicates for each soil within a field, resulting in three soil means for each field (tillage system), depth, and date. We limited the whole-field comparisons to the first and last sampling dates. The three soils were the same in the two fields, and we analyzed the pairs using the same confidence-interval technique (Karlen and Colvin, 1992). Logsdon and Kaspar (1995) analyzed several comparisons between two fields using a paired t-test analysis.
|
Results and discussion
|
|---|
Depth distribution of residue, organic C, and organic N were similar for the two fields (Table 3)
. Previous small depth-incremental studies have shown that no-till, chisel, sweep, and disk management systems all resulted in most of their residue being in the top 100 mm (Allmaras et al., 1988; Staricka et al., 1991; Allmaras et al., 1996). Spatial discussion of the two fields (Cambardella et al., 1994) showed similar trends in these fields for organic C and N.
View this table:
[in this window]
[in a new window]
|
Table 3 Small depth-incremental residue distribution, organic C, and total N for no-till and disk fields on the spring 1996 measurement date
|
|
Rainfall patterns may have affected some of the variability in bulk density (Fig. 1)
. In 1993 for the 2 wk before pre-plant and planting field activities, there had been 79 and 51 mm of rain, respectively, in the no-till and disk fields. Similarly for the 2 wk before pre-plant and planting field activities in 1994, the rain amounts were 46 mm for both fields; in 1995, rain amounts were 75 and 72 mm; and in 1996, they were 7 and 42 mm. Except for the 0- to 20-mm depth-increment of the disk field, mean soil water contents in 1995 and 1996 (Table 4)
may have been high enough for traffic and/or tillage tools to compact the soil. Measurements on a nearby field indicated coefficient of linear expansion (COLE) values of 0.028, 0.050, and 0.097 m m-1 for Clarion, Nicollet, and Canisteo soils, respectively, indicating some shrink–swell potential. Because we measured bulk density (based on field-sampled volumes) rather than clod density (based on changing volume during shrinkage), soil volume changes during shrink–swell in the field would only affect our measurements if the soil surface had subsided (not measured).
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 1 Rainfall patterns for April 1993 through August 1996 for the mean of the two fields. Arrows indicate times of small incremental sampling
|
|
There were not many significant bulk density changes for the no-till field (Table 5)
. The disk field had significant bulk density increases between the summer of 1993 and the spring of 1994. In the disk field (Table 6)
the reduction in bulk densities due to the fall disking in 1994 were not significant within the 95% confidence interval except for the 140- to 160-mm depth-increment. Comparing the first and last sampling dates for the no-till field (Fig. 2)
, there were no significant changes in soil bulk density at any depth (significance not shown for this time-interval comparison). Comparing the first and last sampling dates for the disk field (Fig. 2), there were significant increases in density for the depth-increments of 0 to 120, 180 to 200, 220 to 240, and 260 to 300 mm.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 2 Comparison of depth-incremental bulk densities for the no-till and disk fields for the first and last sampling dates
|
|
Non-management factors probably contributed to soil bulk density changes in this study. At other sites in the no-till field, frequent surface measurements of soil water content (not shown) showed a mean volumetric water content range (maximum–minimum) throughout the 4 yr of this study to be 0.20 m3 m-3 for somewhat poorly drained and well-drained soils, and 0.25 m3 m-3 for poorly drained and very poorly drained soils. Combined with COLE data, shrink–swell would have been expected to contribute to some of the observed density changes. Freeze–thaw would also have been expected to contribute to soil density changes. Depth of freezing measured in this field and another field within 10 km (Sauer et al., 1998; and other data not shown) showed that every winter, freezing depths extend at least 0.25 m for at least 30 d, often freezing to 0.5 or 1.0 m for drier soils with less residue and snow cover. Biopores were present as well. Activity of surface-feeding earthworms (Lumbricus terrestris L.) was observed in sections of the no-till field (even before conversion to no-till) but not in the disk field; however, subsurface earthworms (Apporecteda spp.) were observed in both fields (E. Berry, personal communication, 1994).
Unger (1991) measured overwinter changes in bulk density for no-till soil. He observed for the 40- to 70-mm depth that bulk density increased 1 yr, decreased 1 yr, and did not change significantly the third year. For the 140- to 170-mm depth, the bulk density increased 2 yr and did not change significantly the third year.
In our study, comparing tillage system differences in 1993 revealed that no bulk density differences between the two fields were significant within the 95% confidence interval (Table 7)
. Although this was after the management switch, only minor differences in management had occurred up to the time of sampling in August of 1993 (Tables 1 and 2). This again demonstrated the similarities of the fields at the start of the experiment. For the last sampling date, none of the bulk densities were significantly different between the fields within the 95% confidence interval except for a significantly lower bulk density for the no-till field in the depth-increment of 240 to 260 mm (Table 7). The reason the differences were not significant at other depths was because of inconsistencies among the soils. Bulk densities of Nicollet and Clarion soils were numerically higher for the disk field, but the lack of differences in the Canisteo soil resulted in no significant differences at the other depths (not shown).
None of these bulk densities were high enough to cause a compaction concern. Even if the densities had been larger, the cracks and biopores present would reduce the negative effect of compaction (Logsdon et al., 1992). Crop yields were numerically higher for the no-till field, but a direct comparison was not possible because of variety differences. Soybean yields, averaged for 1993 and 1995, were 2.8 and 2.4 Mg ha-1 for the no-till and disk fields, respectively. Corn yields, averaged for 1994 and 1996, were 11.5 and 11.0 Mg ha-1 for the no-till and disk fields, respectively.
In summary, there were minor variations in soil densities for both the no-till and disk management systems. Much of the soil-density variation was due to natural causes rather than by management. None of the bulk densities were unusually high, but presence of macropores would have prevented any root restriction or infiltration problems. We conclude that concern about compaction under no-till on these structured midwestern soils is unwarranted.
|
ACKNOWLEDGMENTS
|
|---|
We acknowledge the efforts of Gavin Simmons in data collection.
|
NOTES
|
|---|
1 Instrument information is provided for the benefit of the reader and does not imply endorsement by the USDA. 
Received for publication August 5, 1998.
|
REFERENCES
|
|---|
- Allmaras R.R., Copeland S.M., Copeland P.J., Oussible M. Spatial relations between oat residue and ceramic spheres when incorporated sequentially by tillage. Soil Sci. Soc. Am. J. 1996;60:1209-1216.[Abstract/Free Full Text]
- Allmaras R.R., Pikul J.L., Jr., Kraft J.M., Wilkins D.E. A method for measuring incorporated crop residue and associated soil properties. Soil Sci. Soc. Am. J. 1988;52:1128-1133.[Abstract/Free Full Text]
- Blake G.R., Nelson W.W., Allmaras R.R. Persistence of subsoil compaction in a Mollisol. Soil Sci. Soc. Am. J. 1976;40:943-948.[Abstract/Free Full Text]
- Bruce R.R., Langdale G.W., Dillard A.L. Tillage and crop rotation effect on characteristics of a sandy surface soil. Soil Sci. Soc. Am. J. 1990;54:1744-1747.[Abstract/Free Full Text]
- Cambardella C.A., Moorman T.B., Novak J.M., Parkin T.B., Karlen D.L., Turco R.F., Konopka A.E. Field-scale variability of soil properties in central Iowa soils. Soil Sci. Soc. Am. J. 1994;58:1501-1511.[Abstract/Free Full Text]
- Franzluebbers A.J., Hons F.M., Zuberer D.A. Tillage-induced seasonal changes on soil physical properties affecting soil CO2 evolution under intense cropping. Soil Tillage Res. 1995;34:41-60.
- Graham J.P., Blackwell P.S., Armstrong J.V., Christian D.G., Howse K.R., Dawson C.J., Butler A.R. Compaction of a silt loam by wheeled agricultural vehicles: II. Effects on growth and yield of direct-drilled winter wheat. Soil Tillage Res. 1986;7:189-203.
- Horne D.J., Ross C.W., Hughes K.A. Ten years of a maize/oats rotation under three tillage systems on a silt loam in New Zealand: I. A comparison of some soil properties. Soil Tillage Res. 1992;22:131-143.
- Johnson J.F., Voorhees W.B., Nelson W.W., Randall G.W. Soybean growth and yield as affected by surface and subsoil compaction. Agron. J. 1990;82:973-979.[Abstract/Free Full Text]
- Karlen D.L., Colvin T.S. Alternative farming system effects on profile nitrogen concentration on two Iowa farms. Soil Sci. Soc. Am. J. 1992;56:1249-1256.[Abstract/Free Full Text]
- Kaspar T.C., Logsdon S.D., Prieksat M.A. Traffic pattern and tillage system effects on corn root and shoot growth. Agron. J. 1995;87:1046-1051.[Abstract/Free Full Text]
- Lipiec J., Simota C. Role of soil and climate factors influencing crop responses to soil compaction in Central and Eastern Europe. In: Soane B.D., van Ouwerkerk C., eds. Soil compaction in crop production. Amsterdam: Elsevier Sci, 1994:365-390.
- Logsdon S.D., Allmaras R.R., Nelson W.W., Voorhees W.B. Persistence of subsoil compaction from heavy axle loads. Soil Tillage Res. 1992;23:95-110.
- Logsdon S.D., Allmaras R.R., Wu L., Swan J.B., Randall G.W. Macroporosity and its relation to saturated hydraulic conductivity under different tillage practices. Soil Sci. Soc. Am. J. 1990;54:1096-1101.[Abstract/Free Full Text]
- Logsdon S.D., Kaspar T.C. Tillage influences as measured by ponded and tension infiltration. J. Soil Water Conserv. 1995;50:571-575.
- Lowery B., Schuler R.T. Temporal effects of subsoil compaction on soil strength and plant growth. Soil Sci. Soc. Am. J. 1991;55:216-222.[Abstract/Free Full Text]
- Pikul J.L., Allmaras R.R. Physical and chemical properties of a Haploxeroll after fifty years of residue management. Soil Sci. Soc. Am. J. 1986;50:214-219.[Abstract/Free Full Text]
- Sauer T.J., Hatfield J.L., Prueger J.H., Logsdon S.D. Energy balance of a corn residue-covered field during snowmelt. J. Am. Water Resour. Assoc. 1998;34:1401-1414.
- Staricka J.A., Allmaras R.R., Nelson W.W. Spatial variation of crop residue incorporated by tillage. Soil Sci. Soc. Am. J. 1991;55:1668-1674.[Abstract/Free Full Text]
- Stypa M., Nunez-Barrios A., Barry D.A., Miller M.H., Mitchell W.A. Effects of subsoil bulk density, nutrient availability and soil moisture on corn root growth in the field. Can. J. Soil Sci. 1987;67:292-308.
- Unger P.W. Overwinter changes in physical properties of no-tillage soil. Soil Sci. Soc. J. 1991;55:778-782.
- Voorhees W.B., Johnson J.F., Randall G.W., Nelson W.W. Corn growth and yield as affected by surface and subsoil compaction. Agron. J. 1989;81:294-303.[Abstract/Free Full Text]
- Voorhees W.B., Lindstrom M.J. Long-term effects of tillage method on soil tilth independent of wheel traffic compaction. Soil Sci. Soc. Am. J. 1984;48:152-156.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
R. T. Venterea and J. M. Baker
Effects of Soil Physical Nonuniformity on Chamber-Based Gas Flux Estimates
Soil Sci. Soc. Am. J.,
August 20, 2008;
72(5):
1410 - 1417.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C.J. Dell, P.R. Salon, C.D. Franks, E.C. Benham, and Y. Plowden
No-till and cover crop impacts on soil carbon and associated properties on Pennsylvania dairy farms
Journal of Soil and Water Conservation,
May 1, 2008;
63(3):
136 - 142.
[Abstract]
[PDF]
|
|
|
|
|
|
|
D. Ginting and B. Eghball
Nitrous Oxide Emission from No-Till Irrigated Corn: Temporal Fluctuation and Wheel Traffic Effects
Soil Sci. Soc. Am. J.,
May 6, 2005;
69(3):
915 - 924.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. Lampurlanes and C. Cantero-Martinez
Soil Bulk Density and Penetration Resistance under Different Tillage and Crop Management Systems and Their Relationship with Barley Root Growth
Agron. J.,
May 1, 2003;
95(3):
526 - 536.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
