Quantifying wind erosion on summer fallow in southern Alberta.
The correct and reliable estimation of wind erosion losses is important in the evaluation of erosion control systems (conservation tillage, chemical fallow, residue management, stripcropping, shelterbelts). In the past, wind erosion losses have been estimated by the Wind Erosion Equation (WEQ) (Woodruff and Siddoway). The WEQ was derived in laboratory wind tunnels from measurement of basic physical processes. Field verification was not possible as field erosion measuring equipment was not available at the time. To investigate field wind erosion processes, information on meteorological variables, soil flux variables, and temporal soil surface properties for individual erosion events is required. Recently developed methodology (Fryrear et al.) and field equipment, such as airborne dust samplers (Fryrear, 1986), surface creep and saltation samplers (Stout and Fryrear), and the SENSIT wind erosion sensor (Gillette and Stockton), have provided mechanisms to describe field erosion losses on a single storm basis. New equations have been developed to describe the vertical (Fryrear and Saleh; Vories and Fryrear) and horizontal (Stout) components of mass flux. New methodologies for measuring temporal soil properties affecting erodibility have been reported (Zobeck). These measurements and procedures are needed to test the process-based Wind Erosion Prediction System (WEPS), which will replace the empirically based WEQ (Argabright; Hagen; Larney, Bullock, and McGinn).
This paper reports on the erosion processes and soil losses occurring on a fallow field in southern Alberta. The field was monitored for three periods (corresponding to early, mid, and late fallow conditions) during the normal 20-month fallow period. This site served as a validation site for WEPS by quantifying soil loss resulting from wind erosion as it occurred in the field.
We established a wind erosion study site in November 1990 on a Dark Brown Chernozemic soil (Typic Haploboroll) about 15 km (9 mi) southeast of Lethbridge, Alberta (49 [degrees] 37[minutes] N, 112 [degrees] 38[minutes] W). The surface texture was clay loam (30% sand, 38% silt, 32% clay) with an organic carbon content of 1.83%. The land had been cropped continuously since 1984 using zero tillage. In 1990, the entire field was cropped to canola (Brassica napus L.).
The study site consisted of a single circular plot (Fryrear et al.; Larney, Bullock, and McGinn; Stout) 200 m (656 ft) in diameter with an area of 3.14 ha (7.75 ac). In November 1990, the study circle was cultivated with one pass of an offset disc and three passes of a double disc to create an erodible surface surrounded by a non-erodible surface protected by canola stubble seeded to winter wheat (Triticum aestivum L.). The circular design was used to allow erosion data collection regardless of the wind direction and provide a range of field lengths with a minimum number of dust samplers (Fryrear et al.).
We installed 14 clusters of Big Spring Number Eight (BSNE) dust samplers (Fryrear) on January 30, 1991. A cluster comprised an upright with four samplers positioned approximately 0.1, 0.2, 0.5, and 1 m (0.33, 0.66, 1.63, and 3.28 ft) above the soil surface. Six equidistant radii, 60 [degrees] apart, were located on the circle using a transit instrument. On each radius, one cluster was installed at 95 m (312 ft) from the center of the circle or 5 m (16 ft) from the protected surface and one 60 m (197 ft) from the center of the circle or 40 m (131 ft) from the protected surface. One cluster was located in the center of the circle. A background cluster was positioned in the protected surface to the west of the circle (prevailing winds are westerly) to check whether soil from distant erodible areas upwind moved onto the study site during windstorms.
In the center of the circle we installed a meteorological instrument tower (Larney, Bullock, and McGinn). Wind direction and wind speeds at 0.2, 0.5, 1, and 2 m [TABULAR DATA OMITTED] (0.66, 1.63, 3.28, and 6.56 ft) above ground were logged and integrated at 2 min intervals. A SENSIT (Sensit Company, 1226 Milner Lane, Longmont, CO 80503) wind erosion sensor (Gillette and Stockton) was also installed at 5 cm (2 in) above the soil surface close to the duster in the center of the circle. The SENSIT contains a piezoelectric quartz crystal that indicates the exact moment an erosion event begins (threshold conditions) and ends by registering the kinetic energy of particle impact (integrated over 2 minute intervals).
We monitored erosion events in three separate periods between April 1991 and May 1992 representative of different stages in the fallow phase. An erosion event was defined by weights of dust [greater than] 10 g in the bottom samplers of some clusters with corresponding high wind speeds and SENSIT output. The SENSIT measured storm duration. After each erosion event, we measured the height from the soil surface to the center of each of the 14 bottom dust samplers and transferred the contents of each sampler into plastic bags. Sub-samples of dust were taken for moisture determination. Oven-dry dust weights, height from the soil surface to the bottom sampler, and mean wind direction of the storm were used to calculate total soil loss per event. A computer program developed by the U.S. Department of Agriculture at Big Spring, Texas, using equations for vertical distribution of material moving in saltation and surface creep (Fryrear; Fryrear et al.; Vories and Fryrear) and horizontal distribution of material across an eroding surface (Fryrear et al.; Stout) was used for this analysis.
On April 4 and December 10, 1991, and April 5, 1992, we took 20 samples [each 5-to 7-kg of aggregates (11-15 lbs)] from 0- to 2.5-cm (1 in) depth in a transect across the circle for aggregate size distribution analysis. Geometric mean diameter (GMD) of the aggregates was determined (Kemper and Rosenau) using an improved rotary sieve (Chepil 1952). GMD is the screen diameter which 50% of the sample by weight passes.
Results and discussion
Soil losses. Table 2 shows soil losses for the 16 erosion events that occurred during the study monitoring periods. The losses varied from 0.3 Mg/ha (0.1 ton/ac) on April 25, 1991, to 30.4 Mg/ha (13.6 ton/ac) on April 3, 1992. The background cluster located on the protected surface trapped negligible amounts of dust, which verified that all measured losses were from the erodible study circle. The longest storm monitored by the SENSIT was 11.7 hr on December 9, 1991, and the shortest was 0.2 hr on April 25, 1991.
Soil loss tolerances (T-values) have been defined as "the maximum rate of soil erosion that may occur and still permit a high level of crop productivity to be obtained economically and indefinitely" (Wischmeier and Smith). The USDA-SCS has set a maximum T-value of 11.2 Mg/ha (5 ton/ac) annually (Schertz). The T-value was originally established in relation to soil loss caused by sheet and rill erosion by water and is an average annual value. We have used the T-value for comparative purposes only since wind erosion on a single storm basis is a different physical process.
Under our study conditions, six of the 16 storms exceeded the annual T-value. These are substantial losses and illustrate the fragility of the soil even though the field had been in continuous zero tillage for 6 years. Admittedly, the study site represented a "worst-case scenario" as we made no attempt to conserve surface residue. The site was cultivated five times over the fallow period (four times in the first winter and once in the second winter) and three applications of herbicide (May, July, and August 1991). This fallow management system is not remarkably different from that widely used in the semi-arid Canadian prairies where herbicides have been substituted for some tillage operations. However, the implements we used (offset and double discs) have largely been replaced by blade cultivators which leave more surface residue. Also, the timing of tillage, in November and February, was different than the normal May-September tillage operations for fallow.
The total soil loss for all 16 events was 144.4 Mg/ha (64.5 ton/ac). Assuming the bulk density of the soil surface is 1 g/[cm.sup.-3], then the average depth of topsoil removed from the circle during the study period was 14.4 mm (0.6 in). Appreciably more topsoil was lost from the eastern half of the circle, which was furthest downwind from the protected surface during erosion events, since the prevailing wind is westerly, than from the western half. Most of the topsoil was deposited by the wind in a fan-shaped area on the stubble to the east of the circle.
Renewal rates for A horizon material have been estimated at 25.4 mm (1 in) per 30 years (Schertz). Therefore, if no further erosion occurred on this site, it would take about 17 years of natural soil formation to regain the topsoil lost during this study. Other reports have put the natural rate of soil formation at 25.4 mm (1 in) per 100 years (Hall, Daniels, and Foss). If we use this more conservative rate then it would take 60 years to recover the soil removed.
Wind speeds. In general, there was a decline in erosion event mean wind speed at 2 m (6.6 ft) with time. The storm with the highest mean wind speed was the first one, April 4, 1991 (15.1 m/s, 31.6 mi/hr), and that with the lowest mean wind speed was April 9, 1992 (8.4 m/s, 17.6 mi/hr). The decline in storm mean wind speed with time was not necessarily a result of less severe storms as the study progressed, but of a decline in threshold conditions. Threshold conditions may be defined as the wind speed necessary to initiate particle movement and hence, SENSIT measurements. The minimum wind speed associated with each storm is a close approximation of threshold conditions. The decline in threshold conditions was caused by changing surface conditions over the study period. Surface conditions at the time of the first storm were rougher, with more coarse particles and an average aggregate GMD value of 2.53 mm (0.1 in) for the site on April 4, 1991. The GMD value was 0.31 mm (0.012 in) on December 10, 1991 and 1.10 mm (0.043 in) on April 5, 1992.
The change in surface erodibility and threshold conditions due to weathering can be further illustrated by comparing two storms of about equivalent duration but one year apart: #2 (April 8, 1991) and #10 (April 4, 1992). The mean wind speed of storm #2 was 14.3 m/s (30 mi/hr) and this resulted in a soil loss of 2.4 Mg/ha (1.1 ton/ac). The mean wind speed of storm #10 was lower [11.0 m/s, 23 mi/hr] but the total soil loss was 6.1 Mg/ha (2.7 ton/ac), 2.5 times higher than storm #2.
Figure 1 shows that the first three events had minimum wind speeds at 2 m (6.6 ft) greater than 12.5-13.0 m/s (26.327.3 mi/hr). After a season of chemical fallow and one pass with pin harrows the minimum wind speeds for the five December 1991 events dropped to 8.5-9.9 m/s (17.8-20.8 mi/hr). The first of these events (#4, December 6) had the lowest minimum wind speed (8.5 m/s, 17.8 mi/hr); the others had speeds 1.0-1.4 m/s (2.1-2.9 mi/hr) higher. This difference in minimum wind speed is most likely a result of the considerable amount of loose erodible material (LEM) present on the soil surface for storm #4, as no erosion had occurred since the previous April. Although we did not measure LEM, we believe the LEM pool was depleted by storm #4, thereby increasing threshold conditions for subsequent storms. This finding is supported by the soil loss data which showed that, among the December 1991 events, storm #4 caused the greatest loss. [TABULAR DATA OMITTED] This storm was slightly shorter (8.1 vs. 8.3 hr) and had a similar mean wind speed to storm #7. However, soil loss for storm #7 was only 62% of that for storm #4 even though the events were separated by a period of only 5 days (during which no precipitation occurred).
We also monitored erosion at the end of the fallow season. The first six events of this period (storms #9-14, April 1992) all had minimum wind speeds in the range 5.9-7.6 m/s (12.4-15.9 mi/hr), substantially lower than those of a year earlier near the beginning of the fallow period (April 1991), or toward the middle of the fallow period (December 1991).
The mean minimum wind speeds for the three monitoring periods were 12.7 m/s (26.6 mi/hr) for storms #1-3; 9.5 m/s (19.9 mi/hr) for storms #4-8; and 6.7 m/s (14 mi/hr) for storms #9-14. We attribute most of the decline in mean wind speed (3.2 m/s (6.7 mi/hr) between April and December 1991 to a summer of aggregate weathering by rainfall and one cultivation with pin harrows (November 18, 1991), which resulted in a more erodible surface. The decline (2.8 m/s, 5.9 mi/hr) between December 1991 and April 1992 we attribute to a winter of freeze-thaw-wet-dry activity, although that winter was one of the driest on record for the Lethbridge area. The winter and summer declines in mean wind speed were of similar magnitude, suggesting that one pass of pin harrows and a summer of weathering by raindrop action had the same effect on erodibility as a winter of freeze-thaw-wet-dry activity.
There was a noticeable increase in minimum wind speed for the last two events (#15, #16) of the study. Event #15 (April 27, 1992) occurred after precipitation on April 17, 18, 21, and 22 totaling 15 mm (0.59 in). The average gravimetric moisture content of the 0-2.5-cm (0-1 in) depth was 4.4% on April 16; 16.8% on April 20; and 12.9% on April 23, 1992. Surface soil moisture is one of the best deterrents to wind erosion (Bisal and Hsieh; Chepil 1956) as particle movement is prevented until evaporation and drainage has removed sufficient moisture to lower soil cohesion and increase susceptibility to entrainment by wind. The precipitation events also caused crusting of the soil surface. Crust thickness, crust stability and crust fraction (fraction of soil surface, on an area basis, covered by crust) are important properties affecting erosion (Zobeck). Under drier, non-crusted conditions, the minimum wind speed required to register SENSIT output on April 27, 1992 would have been close to 6.7 m/s (14 mi/hr), the average for the previous six events. However, erosion was precluded by surface soil moisture and a 2-5-mm (0.08-0.2-in) thick crust and did not begin until wind speeds reached 10.2 m/s (21.4 mi/hr).
The increase in minimum wind speed to 12.1 m/s (25.4 mi/hr) for the final storm of the study period (#16; May 11, 1992) can be explained by the increase in surface roughness caused by the spring wheat (Triticum aestivum L.) seeding operation on April 27, 1992. Rows were oriented north-south, perpendicular to the direction of the prevailing wind, to help protect emerging seedlings.
Another Alberta study found that pre-seeding tillage decreased erodibility (Black and Chanasyk). However, our study site was direct seeded with an air seeder with hoe attachments that left ridges about 6 cm (2.4 in) high and 20 cm (8 in) apart. This increased surface roughness, delaying the onset of erosion until sustained wind speeds greater than 12.1 m/s (25.4 mi/hr) were recorded 14 days after seeding. Sustained wind speeds ([greater than] 3 hr) greater than threshold conditions in early April (6.7 m/s, 14 mi/hr), were recorded on 7 dates between April 27 and May 11, 1992. Under non-seeded, non-ridged conditions, potential erosion events might have occurred on these 7 dates as there was no precipitation and the soil surface was very dry.
We have quantified, for the first time in Canada, actual soil losses caused by wind erosion on a single storm basis. The total soil loss [144.4 Mg/ha (64.5 ton/ac)] during the fallow period points to the fragility of the soil surface, even after 6 years of continuous zero tillage. The magnitude of erosion losses was closely related to temporal soil properties. Precipitation events increased storm minimum wind speed (threshold conditions) by initially increasing surface moisture and subsequently causing crust formation. Management factors, such as seeding perpendicular to the prevailing wind, also increased threshold conditions.
Based on the fastest rate of natural soil renewal reported for cultivated land, and assuming no further erosion occurred, it would take about 17 years to restore topsoil lost during one injudiciously managed fallow period.
Argabright, M.S. 1991. Evolution in use and development of the wind erosion equation. Journal of Soil Water Conservation 46: 104-105.
Black, J.M.W., and D.S. Chanasyk. 1989. The wind erodibility of some Alberta soils after seeding: Aggregation in relation to field parameters. Canadian Journal of Soil Science 69: 835-847.
Bisal, F., and J. Hsieh. 1966. Influence of moisture on erodibility of soil by wind. Soil Science 102: 81-86.
Campbell, C.A., R.P Zentner, H.H. Janzen, and K.E. Bowren. 1990. Crop Rotation Studies on the Canadian Prairies. Publ. 1841/E, Research Branch, Agriculture Canada, Ottawa, Ontario 133 pp.
Chepil, W.S. 1952. Improved rotary sieve for measuring state and stability of dry soil. Soil Science Society of America Proceedings 16: 113-117.
Chepil, W.S. 1956. Influence of moisture on erodibility of soil by wind. Soil Science Society of America Proceedings 20: 288-292.
Dormaar, J.F. 1987. Quality and value of wind-movable aggregates in Chernozemic Ap horizons. Canadian Journal of Soil Science 67: 601-607.
Fryrear, D.W. 1986. A field dust sampler. Journal of Soil and Water Conservation 41: 117-120.
Fryrear, D.W., and A. Saleh. 1993, Field wind erosion: Vertical distribution. Soil Science 155: 294-300.
Fryrear, D.W., J.E. Stout, L.J. Hagen, and E.D. Vories. 1991. Wind erosion: Field measurement and analysis. Transactions of the American Society of Agricultural Engineers. 34: 155-160.
Gillette, D.A., and P.H. Stockton. 1986. Mass momentum and kinetic energy fluxes of saltating particles. In: W.G. Nickling (ed.) Aeolian Geomorphology. Allen and Unwin, Boston, Mass. pp. 35-56.
Hagen, L.J. 1991. A wind erosion prediction system to meet user needs. Journal of Soil and Water Conservation 46: 106-111.
Hall, G.F., R.B. Daniels, and J.E. Foss. 1982. Rate of soil formation and renewal in the USA. In: Determinants of Soil Loss Tolerance. Special Publication No. 45, American Society of Agronomy, Madison, Wise. pp. 23-39.
Kemper, W.D., and R.C. Rosenau. 1986. Aggregate stability and size distribution. In: A. Klute (ed). Methods of Soil Analysis. Agronomy No. 9 (Part 1), 2nd ed., American Society of Agronomy, Madison, Wisc. pp, 425-442.
Larney, F.J., M.S. Bullock, and S.M. McGinn. 1992. The wind erosion prediction system (WEPS) and wind erosion processes in southern Alberta. Proceedings of the 29th Annual Alberta Soil Science Workshop, February 19-20, 1992, Lethbridge, Alberta, pp. 79-89.
Nkemdirim, L.C. 1986. Chinooks in southern Alberta-some distinguishing nocturnal features. Journal of Climatology 6: 593-604.
Schertz D.L. 1983. The basis for soil loss tolerances. Journal of Soil and Water Conservation 38: 10-14.
Stout, J.E. 1990. Wind erosion within a simple field. Transactions of the American Society of Agricultural Engineers 33: 1597-1600.
Stout, J.E., and D.W. Fryrear. 1989. Performance of a windblown-particle sampler. Transactions of the American Society of Agricultural Engineers 32: 2041-2045.
Vories, E.D., and D.W. Fryrear. 1991. Vertical distribution of wind-eroded soil over a smooth bare field. Transactions of the American Society of Agricultural Engineers 34: 1763-1768.
Wischmeier, W.H., and D.D. Smith. 1978. Predicting rainfall erosion losses - a guide to conservation planning. Agriculture Handbook 537. USDA, Washington, D.C.
Woodruff, N.P., and F.H. Siddoway. 1965. A wind erosion equation. Soil Science Society of America Proceedings 29: 602-608.
Zobeck, T.M. 1991. Soil properties affecting wind erosion. Journal of Soil and Water Conservation 46: 112-118.
F. J. Larney is a soil scientist, M. S. Bullock a research technician, and S. M. McGinn an agricultural meteorologist with Agriculture Canada, Research Station, Lethbridge, Alberta T1J 4B1. D. W. Fryrear is an agricultural engineer with the Agricultural Research Service, U.S. Department of Agriculture, Cropping Systems Research Laboratory, Big Spring, Texas 79721-0909. This research was undertaken as part of the Soil Quality Evaluation Program through funding provided from National Soil Conservation Program to the Canada-Alberta Soil Conservation Initiative. The cooperation of I.A. Lanier who willingly provided land for this study is appreciated. Company names are included for the benefit of the reader and do not imply endorsement or preferential treatment of the product by Agriculture Canada or USDA.
|Printer friendly Cite/link Email Feedback|
|Author:||Larney, F.J.; Bullock, M.S.; McGinn, S.M.; Fryrear, D.W.|
|Publication:||Journal of Soil and Water Conservation|
|Date:||Jan 1, 1995|
|Previous Article:||Soil erosion - crop productivity relationships for six Alberta soils.|
|Next Article:||Well sampling for agrichemicals in high capacity systems.|