| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a The Amalgamated Sugar Company LLC, P.O. Box 700, Paul, ID 83347-0700
b Agriculture and Agri-Food Canada, Lethbridge Research Centre, P.O. Box 3000, Lethbridge, AB, Canada T1J 4B1
c Battelle Washington Operations, 901 D Street, 370 L'Enfant Promenade SW, Suite 900, Washington, DC 20024-2115
d Dep. of Renewable Resources, Univ. of Alberta, Edmonton, AB, Canada T6G 2E3
Corresponding author (larney{at}em.agr.ca)
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: Cpar, soil roughness measured parallel to tillage direction • Cper, soil roughness measured perpendicular to tillage direction • DASD, dry aggregate size distribution • EF, erodible fraction • GMD, geometric mean diameter
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Merrill et al. (1999) defined soil wind erodibility as the tendency of surface soil to resist or be vulnerable to transport by wind. Both soil and plant factors contribute in the estimation of wind erodibility risk for any particular environment. The soil factors include inherent wind erodibility, soil surface roughness, and soil wetness, while the plant factors include surface cover of the soil by living plants and their residues as well as the vertical effect of standing plants or residues.
Dry aggregate size distribution and soil surface roughness are important indicators of wind erodibility (Zobeck, 1991) and are influenced by management (tillage, cropping) and climatic factors (freeze–thaw, wetting–drying, freeze-drying). Overwinter breakdown of soil aggregates increases wind erosion risk, especially after conventionally tilled fallow (heavy-duty cultivator, disk harrow), a practice that leads to low crop residue levels or bare soils on the semiarid Canadian prairies (Larney et al., 1995).
In southern Alberta, overwinter changes in DASD on five fallow management systems of a clay loam soil were related to climatic factors such as cumulative snowfall, snow cover days, and degree of freeze–thaw activity (Larney et al., 1994). Increased snowfall and snow cover days reduced aggregate breakdown, while increased freeze–thaw cycles stimulated aggregate breakdown. In North Dakota, Merrill et al. (1995) found that climatic factors influenced DASD more than tillage treatments in a spring wheat–fallow cropping system. They found significant relationships between DASD and climatic factors such as number of snow cover days, number of freeze–thaw cycle days with no snow cover, and fall precipitation. From fall to spring, snow cover and fall precipitation increased aggregate GMD, while the number of freeze–thaw cycles decreased GMD. Furthermore, Merrill et al. (1999) reported that crop rotation and sampling date were significant sources of variation in DASD, while tillage effects were nonsignificant.
In Kansas and Texas, Chepil (1954) reported increases in erodibility from fall to spring for five soils ranging in texture from fine sandy loam to clay, with greatest increases on the finer textured soils. Erodibility increases were attributed to the freezing of moist soil during winter, which caused expansion of ice crystals within aggregates and subsequent shattering.
Layton et al. (1993) discussed overwinter climatic processes and DASD changes for three tillage systems with different amounts of residue cover in Kansas. Precipitation was the driving force behind DASD changes. However, precipitation effects were also influenced by residue amounts. A zero-till treatment showed the most consistent fall to spring decrease in GMD, even though the high residue cover was thought to have decreased freeze–thaw and freeze-dry effects. Conversely, overwinter studies on a clay loam soil in Saskatchewan showed that bare soil surfaces exposed to freeze–thaw and freeze-drying cycles were more vulnerable to breakdown than those protected by snow cover (Anderson and Bisal, 1969).
Surface roughness is closely associated with DASD and was identified as a governing principle of erosion risk by Chepil (1950), who concluded that the degree of surface roughness is related to the size, shape, distribution, and proportion of surface projections present as nonerodible aggregates. Römkens and Wang (1986) scaled field roughness into four classes including that due to individual particles or aggregates of 0- to 2-mm diam., surface variations or random roughness on the order of 10-cm diam., tillage or oriented roughness in the range of 10- to 30-cm diam., and roughness on a field topography scale. Large clods and tillage ridges perpendicular to the prevailing wind are effective for erosion control (Fryrear, 1984).
Zobeck and Onstad (1987) reviewed the many devices for measuring soil roughness along with the various indices used to summarize the data. Potter and Zobeck (1990) and Potter et al. (1990) developed a microrelief or surface roughness index that estimated microrelief effects on soil susceptibility to wind erosion. Most methods for measuring soil roughness have used pin meters, but recently a roller chain method was developed by Saleh (1993). The advantages and disadvantages of using the roller chain method have been discussed (Merrill, 1998; Saleh, 1997; Skidmore, 1997).
Previous research on overwinter effects on DASD has consisted of only fall and spring sampling (Larney et al., 1994; Layton et al., 1993; Merrill et al., 1995). This type of two-point sampling precludes the development of equations to predict relationships between aggregate size changes and climatic variables. Since overwinter changes in DASD and surface roughness are dynamic phenomena, this study was initiated to monitor overwinter changes in these parameters and to determine relationships between these changes and climatic variables for three clay loam soils in southern Alberta.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The Fairfield site (1992-1993) is located 6 km east of Lethbridge. The 0- to 2.5-cm soil layer contains 39% sand, 28% silt, 33% clay, and 17 g kg-1 organic C. Silage barley (Hordeum vulgare L.) was grown on the site from 1989 to 1991 using conventional tillage (heavy-duty cultivator, disk harrow). In 1992, the site was fallowed with herbicides and one pass of a wide-blade cultivator. Stubble residue was pulverized by a severe hail storm in August 1992. On 18 Sept. 1992, a 20 by 20 m plot was chisel-ploughed to a 12-cm depth. This resulted in a ridged, roughened surface with negligible residue cover.
The Wilson site (1993-1994) is located 15 km south of Lethbridge. The soil contains 35% sand, 31% silt, 34% clay, and 17.2 g kg-1 organic C in the 0- to 2.5-cm layer. The site had been continuously cropped under no-till since 1984. Spring wheat (Triticum aestivum L.) was grown in 1992. From May to August 1993, a 200-m-diam. circle (3.14 ha) was conventionally fallowed with three passes (11 May, 28 June, 2 August) of a heavy-duty cultivator equipped with a hydraulically driven rod weeder. A final pass with a heavy-duty cultivator, rod weeder, and tine harrows to a depth of 8 cm was conducted on 26 Oct. 1993, resulting in minimal surface residue.
The Lethbridge site (1994-1995) is located 1 km north of Lethbridge, and the soil contains 41% sand, 27% silt, 32% clay, and 15 g kg-1 organic C in the surface 7.5-cm layer. The site had been in a spring wheat–fallow rotation since 1955 (Johnston et al., 1995). It was cropped in 1993 and fallowed in 1994 with three passes (24 May, 8 July, 30 August) of a one-way disk.
Dry Aggregate Size Distribution
At all three sites, 5 kg of soil was sampled from the 0- to 2.5-cm depth with a flat shovel for DASD analysis. Sampling interval varied, being dictated by the absence of snow cover. At the Fairfield site, the 20 by 20 m chisel-plowed plot was divided into 20 subplots 1 m wide by 20 m long for DASD sampling. A different subplot was randomly chosen on each of 15 sampling dates between 18 Sept. 1992 and 12 May 1993. Ten subsamples were taken at each sampling date.
At the Wilson site, samples for DASD were taken on nine dates between 26 Oct. 1993 and 29 Apr. 1994. Locations were at 10-m intervals along a 200-m west–east transect on the circular plot for a total of 20 subsamples per sampling date.
At the Lethbridge site, two 6 by 20 m plots were each divided into 20 subplots 1 m wide by 6 m long. A different subplot on each of the two plots was randomly chosen on each of 14 sampling dates between 30 Aug. 1994 and 24 May 1995. Five DASD samples were taken from each of the subplot for a total of 10 per sampling date.
All DASD samples were air-dried and passed through an improved rotary sieve (Chepil, 1952) with 38-, 12.6-, 7.1-, 1.9-, 1.2-, and 0.47-mm diameter sieves. Dry aggregate size distribution was expressed as GMD, the aggregate diameter at which 50% of the sample weight passed (Gardner, 1956), and EF, the percentage of aggregates <0.84 mm in diameter. Since an 0.84-mm sieve was not used, EF was calculated by substituting 0.84 into the regression equation for cumulative percentage passing vs. log of sieve size. Geometric mean diameter has a nonlinear, inverse relationship to EF. Generally, aggregates <0.84 mm in diameter are erodible by wind, and soils with an EF >60% are considered at relatively high erosion risk (Chepil, 1941, 1942).
Soil Surface Roughness
Soil surface roughness parallel to the direction of tillage or random roughness (Cpar), which is roughness caused by aggregates only, was measured at the Wilson and Lethbridge sites, by the chain method of Saleh (1993) using a 1-m roller chain with 0.95-cm links. This method expresses roughness as a percentage value (Cr) as calculated in Eq. [1]:
 | (1) |
In fall, 10 positions were randomly chosen for soil surface roughness measurements at the Wilson site, and 12 at the Lethbridge site. These positions were permanently marked and visited on each measurement date. Cpar was measured 11 times at the Wilson site between 26 Oct. 1993 and 10 May 1994 and 14 times at the Lethbridge site between 30 Aug. 1994 and 24 May 1995. Soil roughness was measured when there was no snow cover.
Climatic Variables
Mean hourly air temperature (1-m height) was measured on each site, commencing on 23 Oct. 1992 at Fairfield, 1 Nov. 1993 at Wilson, and 30 Aug. 1994 at Lethbridge. From air temperature data, the number of freeze–thaw cycles was determined for all days and for days without snow cover. Cutoffs of -2°C for freezing and +2°C for thawing were used to estimate the number of soil freeze–thaw cycles. Rainfall, snowfall, snow depth, and total precipitation were measured at the Agriculture and Agri-Food Canada Research Centre weather station, located 
5 km from the Fairfield site, 15 km from the Wilson site, and 1 km from the Lethbridge site. A trace of snow constituted a snow cover day.
Statistical Analysis
Mean GMD, EF, and Cpar values and standard errors were calculated for each sampling date. For all three sites, regression analysis (SAS Institute, 1989) was performed between the mean values of the dependent variables GMD and EF and cumulative days from initiation of the study. For the Wilson and Lethbridge sites a similar regression was performed with Cpar as the dependent variable.
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
8 cm of snow depth) followed by evaporative drying allowed the saturated soil surface to crust by the 1 Feb. 1993 sampling date. Dry aggregate size distribution sampling of a crusted soil surface created clods of considerable size, strong enough to resist breakdown during rotary sieving, which increased GMD. Soils in northern climatic zones are prone to crusting during winter because of a decline in wet aggregate stability from fall to spring caused by freeze–thaw cycles (Bullock et al., 1988). Merrill et al. (1995) also reported increases in GMD after snowmelt in spring. However, climatic conditions during February and March 1993 allowed GMD to return to precrusting values (Fig. 1a). The decline in GMD after surface crusting followed a similar pattern to the post-tillage decline in the fall (Fig. 1a). If the 1 February value is treated as an outlier and deleted, the R2 value of the exponential relationship increased to 0.73 (P = 0.05). At the Wilson site, GMD ranged from a maximum of 9.05 mm after fall tillage (26 Oct. 1993) to a minimum of 1.17 mm (14 Mar. 1994). Although the maximum and minimum GMD values were greater than those of the other study sites, the decline in GMD over time was greater than at the other sites, as denoted by the different form of exponential equation (Fig. 1b).
At the Lethbridge site, GMD ranged from 4.71 mm after fall tillage (30 Aug. 1994) to 0.80 mm (17 May 1995). Although the maximum fall GMD value at the Lethbridge site was much lower than that at the Wilson site (4.71 vs. 9.05 mm), minimum GMD values in spring were quite similar (0.80 vs. 1.17 mm). This demonstrates that similar levels of erosion risk may exist in spring on soils that had widely different DASD values in the previous fall.
Erodible Fraction
All sites showed a general increase in EF from post-tillage values in fall to final values in spring (Fig. 2a–2c)
. Linear regressions of EF data with cumulative days from initial fall sampling resulted in R2 values (P = 0.05) of 0.57 for the Fairfield site, 0.91 for the Wilson site, and 0.62 for the Lethbridge site. The EF ranged from a minimum value of 12.6% (26 Oct. 1993, Wilson site) to a maximum of 74% (22 Apr. 1992, Fairfield site). Anderson and Wenhardt (1966) determined that soils with EF values >60% were generally erodible by wind. This threshold level was exceeded at the Fairfield site only from 4 Dec. 1992 to 1 Feb. 1993 and from 3 March to 12 May 1993. Like the GMD data, EF data from the Fairfield site reflected the effects of surface crusting, with a large decrease in the EF on 1 Feb. 1993 (Day 137). However, EF values rapidly returned to precrusting values (Fig. 2a). If the EF value for 1 February is deleted, the R2 value for the linear regression increased from 0.57 to 0.68 (P = 0.001).
|
Soil Surface Roughness
Soil surface roughness (Cpar) values for the Wilson site (Fig. 3a) ranged from 15.1 (26 Oct. 1993) to 3.7% (30 Mar. 1994). The Cpar values for the Lethbridge site (Fig. 3b) ranged from 14.4 (30 Aug. 1994) to 3.3% (19 Apr. 1995). Both sites exhibited an overwinter smoothing trend following the rough conditions created by fall tillage. Merrill et al. (1999) reported a range of Cpar from 12.5 to 3.1% for a low residue till (<10% residue cover) spring wheat–fallow study in North Dakota, which was very similar to the ranges found at our sites.
|
Climatic Variables and Wind Erodibility Changes
The total number of freeze–thaw cycles, using the air temperature cutoffs of +2°C for thawing and -2°C for freezing, was 54 for Fairfield in 1992-1993, 65 for Wilson in 1993-1994 and 50 for Lethbridge in 1994-1995 (Table 1). The number of cycles on days with no snow cover was 45 for Fairfield, 58 for Wilson, and 39 for Lethbridge. Therefore, soils were exposed to 78 to 89% of the total freeze–thaw cycles occurring during the study periods.
|
The second or "winter snow" period represented the time from the last rainfall event in fall to the first rainfall event in spring when all precipitation fell as snow (Table 1). For all three sites the "winter snow" period produced the largest increases in EF (Table 1). The Wilson site showed the largest increase in erodibility (+25.1%) and had more than twice the number of freeze–thaw cycles without snow cover (40 vs. 18) as the other two sites. Previous studies have shown a positive correlation between freeze–thaw cycles and EF increases (Larney et al., 1994; Merrill et al., 1995). Additionally, the higher snowfall amount at the Wilson site (82.4 mm) compared with the other sites (42–43 mm) may have allowed greater wetting of the soil surface during the thaw periods, leading to greater soil disruption upon subsequent freezing.
The Fairfield and Lethbridge sites had similar snowfall amounts and numbers of freeze–thaw cycles (Table 1), but the Lethbridge site had a larger increase in EF compared with Fairfield (+17.0 vs. +11.2). Other processes such as freeze-drying of the soil surface or soil abrasion by blowing snow (Bullock et al., 1992, 1999; de Jong and Kachanoski, 1988) may have contributed to this difference.
The third and final period was the "spring snow/rain" period, which represented the time from the first rainfall event in spring and included precipitation as snow and rain until the end of each study year (Table 1). During the "spring snow/rain" period, only the Fairfield and Wilson sites showed increases in EF, and these increases were the smallest of all three overwinter periods.
The Lethbridge site showed a decrease (-9.2%) in EF for this period, which may be related to the high amount of total precipitation (155.7 mm) and a daily maximum rainfall of 24.4 mm. Most of the precipitation (78%) during this period at the Lethbridge site was in the form of spring rain near the end of the study (May) when the frequency of freeze–thaw cycles was declining. We believe this precipitation caused saturation of the soil surface, slaking of aggregates and soil crusting as reported by Kemper et al. (1987) and Uehara and Jones (1974). In addition, Anderson and Wenhardt (1966) found that a clay loam soil had a lower EF in spring than the previous fall in 5 of 7 yr and attributed this to a positive correlation with precipitation. The reduced frequency of freeze–thaw cycles at this time of year also decreased soil disruption. Similar decreases in EF did not occur at Fairfield and Wilson as total precipitation amounts were lower than at Lethbridge for the "spring snow/rain" period (Table 1).
At the Lethbridge site, even though the "winter snow" and the "spring snow/rain" periods had similar numbers of freeze–thaw cycles (26 vs. 24, Table 1), total precipitation in the "winter snow" period was only 23% of that in the "spring snow/rain" period. Trends in erodibility were quite different with the "winter snow" period showing a large increase (+17.0% change in EF) and the "spring snow/rain" period showing a decrease (-9.2% change in EF). This illustrates that the type and amount of precipitation, rather than the number of freeze–thaw cycles, is an important factor in the magnitude and direction of erodibility changes.
Climatic factors affecting overwinter changes in EF also played a role in overwinter changes in soil surface roughness (Cpar). The overwinter decline in Cpar (Fig. 3) was related to the breakdown of large aggregates by the same climatic forces that induced declines in EF. However, unlike EF, the impact of rainfall may account for a larger component of the decline in Cpar, due to its smoothing effect (Römkens and Wang, 1987; Zobeck and Popham, 1990). Even though the Wilson and Lethbridge sites differed in the amounts and form of precipitation, the number of days with snow cover, and the number of freeze–thaw cycles (Table 1), final Cpar values were similar at both sites (3.7% at Wilson, 3.3% at Lethbridge, Fig. 3). This illustrates that similar levels of soil roughness can exist in spring, despite being influenced by a different set of climatic factors over the previous winter.
Management Effects on Dry Aggregate Size Distribution
Although the sites in this study had quite similar surface textures and organic C levels, their management histories varied. The Fairfield site had been continuously cropped with conventional tillage and all surface residue was removed for 3 yr prior to the study. This site had the highest initial EF value (38.9%, Fig. 2a) and was the only site where EF increased to >60% (wind erosion threshold; Anderson and Wenhardt, 1966) during the winter period. Physical disruption from a history of tillage and the removal of crop residue seems to have increased erosion risk.
The Wilson site had been in zero tillage for 7 yr prior to the fallow season in 1993. Although this site had the lowest initial EF value (12.6%, Fig. 2b) of all three sites, it showed the fastest overwinter increase in EF (Fig. 2b). This agrees with the findings of Larney et al. (1994) and Layton et al. (1993), who indicated that zero till soils may have higher erosion risk than tilled soils if residue cover is jeopardized. In addition, the Wilson site was the only one where earthworm casts were observed on the soil surface. Clapperton et al. (1997) reported significantly higher populations of earthworms (Aporrectodea calignosa Savigny) under zero tillage than under conventional tillage in southern Alberta. Earthworm casts are known to have lower bulk densities and have a propensity to absorb more water than surface soil, making them more susceptible to freeze–thaw disruption (Lal and Akinremi, 1983).
The Lethbridge site, which had been under long-term wheat–fallow management with conventional tillage, had an initial EF value (30.4%, Fig. 2c) that fell between the values for the Fairfield (38.9%, Fig. 2a) and Wilson (12.6%, Fig. 2b) sites but had the lowest organic C levels of all sites (15 g kg-1). The EF values at the Lethbridge site did not increase above the threshold value of 60%, unlike the Fairfield site, and increased very slowly over the winter period (Fig. 2c), unlike the Wilson site.
Our results suggest that management systems and cropping histories may play a significant role in overwinter changes in wind erodibility risk. They influence the starting value of EF in fall as well as the slope of the overwinter relationship between EF and time.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
The study also demonstrated the effect of cropping history on wind erodibility and pointed to the fragility of zero tillage soils once the protective layer of surface residue is removed. A soil that was cultivated after 7 yr of zero tillage, although nonerodible in fall, was not resilient to overwinter climatic forces experienced in southern Alberta, and had approached erosion risk in spring.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Received for publication April 3, 2000.
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
