NSCP Research

Contribution Agreements
Soil Structure and Erosion Modelling

Contribution agreements were supported to continue two ongoing research projects which were deemed extremely important to understanding soil structure and erosion modelling.


  1. The Relationship between Landscape Position, Tillage Practices, and Soil Loss: Model Development - University of Guelph - Dr. R. G. Kachanoski - $76,528.00
        View/Download report  [136 KB pdf]
  2. Methodologies for Assessing Soil Structure and for Predicting Crop Response to Changes in Soil Quality - University of Guelph - Dr. B. D. Kay - $83,472.00
       View / Download Report [1001 KB pdf]


NSCP Research List | NSCP Program

The Relationship between Landscape Position, Tillage Practices,
and Soil Loss: Model Development

D.A. Lobb, Dr. R.G. Kachanoski,
Centre for Land and Water Stewardship
Department of Land Resources Science
University of Guelph, Guelph, ON N1G 2W1

September, 1993


In 1987 the University of Guelph initiated a soil erosion study, Management of Farm Field Variability II. Soil Erosion Processes on Shoulder Slope Landscape Positions (SWEEP/TED), at two field sites in southwestern Ontario, one in Brant County and the second in Middlesex County. The study measured tillage translocation and tillage erosion on convex upper slope landscape positions. The estimated rate of soil loss resulting from net downslope translocation was in excess of 6.5 kg m-2 yr-1 at the Brant Co. field site and in excess of 4.5 kg m-2 yr-1 at the Middlesex Co. field site. Subsequent examination of that data recognized that tillage erosion was responsible for at least 70 % of the total soil lost on the upper slope landscape positions based on estimates of total soil loss using resident 137Cs.

A second study, Soil Loss by Tillage Erosion: The Effects of Tillage Implement, Slope Gradient, and Tillage Direction on Soil Translocation by Tillage (SWEEP/TED), by the University of Guelph from 1990 to 1991 at two field sites in Huron County was conducted to determine the effect of tillage implement type on the magnitude of tillage translocation and tillage erosion under a range of slope gradients in topographically complex landscapes. All four tillage implements, the chisel plough, mouldboard plough, tandem disc and field cultivator, were found to be erosive, causing soil loss on upper slope landscape positions and soil accumulation in lower slope landscape positions.

Tillage erosivity, the potential for tillage events to erode soil within a landscape, was recognized to be a function of several physical and human parameters, including: tillage tool shape and arrangement within a tillage implement, tillage implement length and width, tractor-implement match, tillage depth and tillage ground speed, and tillage operator response to varying landscape conditions. The tillage parameters are controlled by selection and varied by the operator. Landscape erodibility, the potential for the soil within the landscape to be eroded by tillage events, was recognized to be a function of the topographic and soil parameters, including: slope gradient and curvature, and field soil bulk density, soil moisture content, and the ability of the soil to resist displacement and translocation (internal friction due to cohesion and adhesion). The landscape parameters evolve through erosion.

The objective of this, the third study conducted by the University of Guelph, was to define the relationship between tillage erosion and landscape position in the form of a model based on the data collected in the Huron County study.

In the proposed model, tillage erosion was calculated as the net translocation at specified points in the landscape, the difference between the soil translocated into a point and the soil translocated out from that point during a single tillage operation. Tillage translocation was related to slope gradient and slope curvature by a simple linear function. The translocation in to and out from a point was calculated from forward and backward differences in topographic conditions. Therefore, the model predicted soil redistribution from forward tillage translocation along two-dimensional landscape profiles.

The proposed tillage erosion model was calibrated using experimental data from the Huron Co. study Soil Loss by Tillage Erosion: The Effects of Tillage Implement, Slope Gradient, and Tillage Direction on Soil Translocation by Tillage (SWEEP/TED).

The proposed tillage erosion model was validated using data collected during two preceding studies, Management of Farm Field Variability I. Quantification of Soil Loss in Complex Topography (SWEEP/TED) conducted in Brant County and Soil Loss by Tillage Erosion: The Effects of Tillage Implement, Slope Gradient, and Tillage Direction on Soil Translocation by Tillage (SWEEP/TED) conducted in Huron County. Resident 137Cs radioactivity was used to estimate soil redistribution within the landscapes of the field sites. These estimates of soil loss and accumulation were compared to those predicted by the tillage erosion model based on the topography of the field sites.

The proposed tillage erosion model provided a reasonably accurate prediction of soil redistribution at the Brant County field site when compared to that estimated using resident 137Cs radioactivity. The tillage erosion model provided a relatively poor prediction of soil redistribution at the Huron County field site when compared to that estimated using resident 137Cs radioactivity. There is some indication that the poor prediction for the Huron site was due in part to the model's simplicity (not able to predict the effect of curvature asymmetry on tillage erosion - a problem which would be greater at this site than the Brant site because of smaller scale of the ridge). Soil losses, based on the 137Cs data, were situated on the convex upper slope landscape positions, but they were greater in severity on the shoulder slope position of the steeper of the ridge's two slope faces. Although the model correctly predicted the general pattern of soil losses and accumulations, the model underpredicted the magnitude, or severity, of soil losses at both field sites. Too few data of soil accumulation estimates were available to make a similar inference about soil accumulation. Several possible reasons for this underprediction of soil loss were identified:

  1. the tillage implements and the tillage sequence used to predict the soil redistribution may have been less intensive than those responsible;

  2. inaccuracies associated with the use of resident 137Cs may have caused overestimation of soil redistribution (the problem associated with point measurements resulting in apparent losses on backslope positions, as well, the current level resident 137Cs for a non-eroded site may be much less than the assumed 2500 Bq m-2 in Huron County);

  3. wind and water erosion may have caused soil redistribution in addition to that caused by tillage erosion (the redistribution pattern is inconsistent with that of soil erosion by overland water flow).

For a first attempt at modelling tillage erosion in complex landscapes the performance of the proposed model was considered very good. Clearly, there are limitations to the complexity and consequently predictive capabilities of the model due to the lack of experimental data for calibration procedures, particularly tillage depths. At the time the study was initiated the number of parameters involved and the complexity of the relationships was not fully appreciated. This was exploratory research, and therefore presuming that a model could be developed on such a data set was very ambitious.

The fact that the proposed tillage erosion model predicts greater rates of soil loss on convex upper slope landscape positions where severe soil loss occurs, and soil accumulation in concave lower slope landscape positions where soil accumulation is observed, indicates that this model is more appropriate than water erosion models for predicting soil erosion in topographically complex landscapes. Consequently, it can be presumed that the proposed tillage erosion model is more appropriate than water erosion models for basing soil conservation decisions relating to soil degradation and soil productivity. Comprehensive soil erosion models including submodels for erosion by wind, water and tillage may provide the best prediction of soil redistribution in topographically complex landscapes.



NSCP Research List | NSCP Program

Methodologies for Assessing Soil Structure and for Predicting
Crop Response to Changes in Soil Quality

University of Guelph
B. D. Kay, A. da Silva, K. Denholm, N. Eshraghi, E. Perfect and V. Rasiah


The objectives of this study were:

  1. to identify a method(s) for measuring soil structural changes which may be related to soil management systems and which can be shown to be useful for characterizing changes in soil quality across a range of soil conditions and;

  2. to evaluate existing crop productivity in terms of their suitability for predicting crop response to changes in soil quality.

The budget associated with the contract was directed to field and laboratory studies related to objective (a) and the collection of field data to be used in the evaluation of crop productivity models [obj. (b)]. The research related to objective (b) has been part of the work plan of an Agriculture Canada Research Branch staff person and the salary expenditures associated with this part of the project have not been charged to the contract.

The field studies for the project were located on the farm of Mr. Don Lobb, Huron County. This site was one of the T-2000 sites investigated during the Ontario Land Stewardship program and is one of the longest running field scale side-by-side comparisons of zero and conventional tillage in Ontario. The comparison is maintained as a strip about 0.5 km in length which traverses soils with clay contents ranging from 7 to 35%. The site was maintained in corn production in 1991 and 1992. (The study has been extended to 1993 and supported by funds from alternative sources). Thirty-six locations (soils) were identified on each transect (tillage treatment) for detailed studies on soil structure.

Soil structure can be defined in terms of structural form and structural stability. Structural form relates to the arrangement or "architecture" of solid and void spaces whereas structural stability refers to the resistance of structural form to deformation (including fragmentation) when stress is applied. Structural form can progressively change subsequent to a change in soil or crop management practices through changes in the level of stress applied to a soil or by changing the population of soil organisms (e.g. earthworms). Structural form will also change if the stress remains constant but stability changes. Management practices can cause changes in stability by causing changes in the level of stabilizing materials (primarily organic in origin) in soils. Methodologies to assess both structural stability and structural form were assessed in this study. Pedotransfer functions were developed, where possible, in order to describe the contribution of inherent soil properties to the magnitude of the different parameters that were measured.

Parameters which were used to describe structural stability related to the resistance of soil to deformation by two types of stress: moving water and mechanical stress causing fragmentation. Stability parameters related to moving water were assessed at two different scales: that of aggregates > 0.25 mm, and that of clay-sized particles (< 0.002 mm). The resistance to mechanical stress was assessed using tensile strength and the distribution of aggregate sizes created by tillage.

Preliminary studies using rainfall simulation techniques indicated that the amount of runoff and the amount of sediment in the runoff arising from a rainfall event were related to dispersible clay and time to ponding; and that these parameters became more important as the extent of surface cover by crop residues decreased. Time to ponding is related to infiltration characteristics and was found to be strongly dependent on wet aggregate stability. Stability parameters at the scale of aggregates and at the scale of dispersible clay both appeared, therefore, to be important in describing runoff and sediment load in the runoff. Studies were therefore initiated to assess both characteristics in more detail.

A turbidimetric technique was developed to expedite characterization of dispersible clay across the range of soils on the study site. The technique involved developing a standard curve (turbidity as a function of concentration of dispersible clay) which can be described as a function of inherent soil characteristics (clay and organic matter content), and then characterizing the dispersibility of clay. Variation in the characteristics of the standard curve with soil properties appeared to be due to the concentration range in which the standard curve was determined and the mean weight diameter of the dispersed clay fraction. A single curvilinear standard curve was found to be applicable to all of the soils on the study site since the curvilinear representation incorporated the influence of both concentration and mean weight diameter. The dispersible clay content was found to increase with increasing clay content, increasing water content and decreasing organic matter content; the variation in dispersible clay content with tillage appeared to be due primarily to the influence of tillage in reducing the organic matter content.

Wet aggregate stability was found to increase with clay, water and organic matter content. The reduction in stability with tillage appeared to be related to the reduction in organic matter content with tillage.

The response of soil to mechanical stress was assessed by considering tensile strength measurements and the dry aggregate size distribution created in seedbed by tillage. Tensile strength increased with increasing clay content, wet aggregate stability and decreasing organic matter content. Aggregate size distributions were assessed using different approaches. A description of the distribution by fractal theory was found to be most accurate. The analyses indicated that the number of aggregates in the largest size fraction increased with increasing clay content, wet aggregate stability and decreasing organic matter content. A comparison of tensile strength and aggregate size distribution characteristics showed a highly significant correlation indicating increasing fragmentation with decreasing tensile strength. The analyses suggest that one parameter could be predicted from the other and that, for a given application of stress through tillage, either parameter could be predicted from inherent soil characteristics.

Parameters that were used to describe structural form included both static and dynamic parameters. Bulk densities and relative bulk densities were measured. The concept of least limiting water range (LLWR) was used to describe the combined effects of structural form on aeration, resistance to penetration and available water and represented measurements under "static" conditions. Structural form was characterized under dynamic conditions using infiltration measurements. Once again the sensitivity of these parameters to inherent soil properties and to management was determined.

Bulk density was found to vary with clay and organic matter contents and was higher on the no till than the conventional till treatment. The relative bulk densities were determined by dividing the observed bulk density of each soil by the bulk density determined after compacting each soil with a compressive stress of 200 kPa.The bulk density after compaction was also found to vary with clay and organic matter content. The relative bulk density was however constant across all soils for a given tillage treatment and was 11% higher on the no till treatment. This type of analysis has not been done before and obviously has important implications for all laboratory studies in which bulk density and inherent soil properties are variables.

Values of LLWR were determined by establishing the functional dependence of the water release curve (potential versus water content) and the soil resistance curve (resistance to penetration versus water content) on bulk density, clay and organic matter content. Limiting values were then assigned, using generally accepted criteria in the literature, for aeration (10% air filled porosity), field capacity (0.01 MPa), permanent wilting point (1.5 MPa) and resistance to penetration (2.0 MPa) to these functions in order to define the LLWR for each soil. Analyses showed a wide variation in LLWR with clay and organic matter content for a given tillage treatment. Correlation of LLWR with plant growth parameters indicated a strong correlation between LLWR and plant population. Analyses are still underway relating soil water content and LLWR to leaf extension during the growing seasons.

Infiltration was measured in the non-trafficked inter-rows on all 36 locations under both tillage treatments. The field saturated hydraulic conductivity, Kfs, was found to be higher under the no-till treatment than under conventional till and may reflect greater continuity in macropores in the no-till treatment. A statistically significant, but poor, correlation existed between Kfs and inherent soil properties.

Data were collected that could be utilized in evaluating plant growth models. Climatic records were obtained from a weather station maintained on the site. Additional information on plant response parameters (yields, root distributions) were also recorded.

Adaptation of current crop productivity models is being undertaken by Mr. Ken Denholm, Agriculture Canada Research Branch, Guelph as part of this project. This activity has not progressed as rapidly as originally anticipated. However, once the models are developed a complete data set is available to assess the models in terms of their ability to predict yield response on soils of different structure and under the dramatically different climatic conditions that existed in 1991 and 1992.


NSCP Research List |NSCP Program


Created: 09-21-1996
Last revised: Sunday, May 08, 2011 02:50:13 PM