| | Soil-landform (SL) | Crop rotation (CR) | Tillage intensity (TI) | Climate (soil zone, ecoregion) | Area - weighted means | Ecodistrict (regional) representation | Strengths and limitations of the process and tools
The process was very labour intensive and generated a large amount of data (Appendix D - diskette). Examples have been extracted to illustrate the specific impact of soil-landform, crop rotation, tillage intensity options and area weightings, climate/ soil zone impacts as well as a final district/ regional summation.
Soil-Landform (SL)
The effect of soil-landform varies somewhat from region to region (Table 2).
Table 2. The impact of Soil-Landforms (SL) on predicted soil quality after 30 years under present management. 1
Ecodistrict | Soil Code | Slope | Soil Rating |
T0 | T30 | Max. | Min. |
593 | ESH
DON
RYF
TAG | A
A
A
B | 47
51
54
59 | 48
53
54
59 | 50
54
54
62 | 47
52
54
58 |
596 | DVS
JUH
JUH
NMA | B
B
C
A | 44
65
65
52 | 45
66
65
52 | 48
69
69
53 | 44
64
63
52 |
598 | AGH
HZM
LAD | A
A
A | 62
51
53 | 62
52
53 | 63
53
54 | 61
51
53 |
680 | LCY
SDN
FRY
COA | A
A
A
A | 54
62
66
55 | 55
63
66
56 | 55
63
66
57 | 54
62
66
55 |
681 | UCS
ATO
LNN
PIB
MLA | A
A
A
A
A | 65
70
59
59
63 | 66
71
59
59
63 | 66
71
60
59
64 | 65
70
59
59
63 |
727 | MMO
POK
AGS
FLU
WKN | A
A
A
B
A | 78
81
66
65
50 | 78
81
66
65
50 | 78
81
66
66
53 | 78
81
66
65
49 |
728 | AGS
POK
BVH
CMO | A
A
B
A | 61
76
67
51 | 61
76
67
51 | 62
76
67
52 | 61
76
67
51 |
|  |
Ecodistrict | Soil Code | Slope | Soil Rating |
T0 | T30 | Max. | Min. |
730 | EO2
EO3
IRM
FLU | B
C
B
B | 68
68
50
57 | 68
69
51
57 | 69
69
51
57 | 68
68
50
56 |
746 | AT2
AT3
PED | B
A
A | 84
84
82 | 84
84
82 | 84
84
82 | 84
84
82 |
743 | HND
PRO | B
B | 49
56 | 51
57 | 51
57 | 50
56 |
781 | DMH
HND
DMH
CNN | A
B
B
B | 70
45
70
48 | 71
46
71
49 | 71
47
71
49 | 71
45
71
48 |
793 | LET
CIO
CRD
CMY | A
A
A
A | 53
48
50
29 | 52
49
49
23 | 55
50
50
25 | 50
47
47
20 |
806 | HDY
HUK
ROL | B
A
B | 30
22
29 | 30
22
30 | 31
23
31 | 29
22
29 |
821 | MAB
FMT
CFD | B
C
A | 36
33
38 | 33
34
38 | 33
34
38 | 33
33
37 |
828 | MAB
MAB
CFD
ROL | B
A
A
A | 35
35
38
29 | 36
36
37
30 | 39
39
39
31 | 34
34
36
39 |
|
1 Ratings are averages of all crop rotation and tillage options.
The largest differences are associated with features that impact on erosion. Sandy texture, such as the CMY soil in the Dark Brown Ecodistrict 793, is particularly noticeable. Another major factor was the original organic matter content. Those soils which were initially low in organic matter, such as the Luvisols in Ecodistricts 593 or 594 or the Brown Chernozems in 821 or 828, showed the greatest variations in response to different management options. It appears that the soils with higher initial quality show the least change which could be interpreted as being the most resilient or most resistant to degradation.
Crop Rotation (CR)
Crop rotations that represent the two extremes of cover, the degree of erosion control and the amount of vegetative material that will be returned to the soil had a marked impact on predicted soil quality change (Table 3, Figure 2).
Table 3. The impact of crop rotation (CR) on predicted soil quality after 30 years under present conditions. 1
Ecodistrict Number |
| Crop Rotation2 | 593 | 596 | 598 | 680 | 681 | 727 | 728 | 730 | 746 | 743 | 781 | 793 | 806 | 821 | 828 |
| Initial(To) | 50.1 | 55.3 | 55.8 | 58.3 | 64.5 | 71.6 | 64.7 | 64.2 | 83.4 | 52.2 | 56.8 | 49.2 | 27.1 | 35.3 | 35.2 |
| A | | | | | | | | | | | 57.4 | 46.9 | 26.6 | 33.9 | |
| B | 50.7 | 55.0 | 56.2 | | | 71.6 | 64.7 | 64.4 | 83.4 | | 58.1 | 47.9 | | | 34.8 |
| C | 50.7 | 54.7 | 56.1 | | | 71.6 | 64.7 | 64.6 | | | 58.1 | | | | |
| D | | | | | 64.5 | 71.6 | 64.7 | 64.5 | 83.4 | | 57.5 | 48.5 | | | |
| E | | | 56.2 | 58.8 | 64.5 | 71.6 | 64.7 | 64.5 | 83.4 | | | 49.3 | | | |
| F | | | 56.2 | 59.1 | 65.1 | 71.6 | 65.3 | | 83.4 | | | | | | |
| G | 52.1 | 59.0 | 57.0 | 58.8 | 65.4 | 71.8 | 65.3 | 64.8 | 83.4 | | | | | | |
| H | 50.7 | 54.7 | 55.6 | 58.8 | 65.1 | 71.6 | 64.7 | 64.7 | 83.4 | 53.2 | 57.8 | 47.0 | 26.6 | | 34.3 |
| J | | | | 59.3 | 65.2 | 71.9 | 65.3 | 64.7 | 83.4 | | | 48.0 | 28.2 | | 36.4 |
| L | | | | | | | | | | | | 48.9 | | | 37.3 |
1 Ratings are averages of all crop rotation and tillage options.
2 Rotation symbols refer to farms dominated by: A = wheat-fallow; B = wheat; C = wheat-oilseeds; D = wheat-barley; E = barley; F = barley-forages; G = forages; H = mixed (no dominant); J = significant specialty crops; L = significant irrigation. See appendix 4 for more detail.
Figure 2. Predicted 30 year percent change in land suitability ratings for the various crop rotations.
Those rotations with a high proportion of fallow, A and H in the Prairie Zone, were predicted to have a 4% decrease in LSRS rating. At the other extreme, rotations with forages (rotation G) or irrigated forages (rotation L) were predicted to realize a 6% increase in ratings. With minor exceptions, the other rotations, that are mainly combinations of annual cereal crops with less than 15% fallow, showed little impact. The sandy CMY unit in Ecodistrict 793 was particularly sensitive to CR variation.
Some of the rotations that have rather loose definitions, such as L (any rotation with over 15% irrigation), or J (any rotation with over 15% specialty crops) and those which vary from one region to another such as J or H (mixed crops), show considerable variation in predicted response. This is seen in absolute values but is enhanced in the percentage calculation because of their occurrence in the Prairie Zone with low initial ratings. The large variation for the forage rotation (G), and its rather low (2%) average increase, reflects both environmental and rating considerations. This rotation occurs in the more humid regions and, while always resulting in an increase in organic matter content (Appendix D), the impact is greater for those soils which had low initial organic levels such as the Luvisols in Ecodistrict 596.
Tillage Intensity (TI)
It was anticipated that this factor would have a significant impact on soil quality trends but that was not the case in this analysis (Table 4). Of the 15 Ecodistricts, 5 showed a slight increase in soil quality with reduced tillage, 8 showed no difference and 2 actually had a decrease.
Table 4. The impact of tillage intensity (TI) on predicted soil quality after 30 years under present management conditions
Ecodistrict Number |
Tillage intensity1 | 593 | 596 | 598 | 680 | 681 | 727 | 728 | 730 | 746 | 743 | 781 | 793 | 806 | 821 | 828 |
1 | 51.6 | 55.8 | 56.2 | 58.9 | 64.9 | 71.6 | 64.9 | 64.6 | 83.4 | 52.7 | 57.7 | 47.9 | 27.1 | 34.1 | 35.6 |
2 | 50.5 | 55.8 | 56.2 | 58.9 | 64.9 | 71.7 | 64.9 | 64.6 | 83.4 | 53.7 | 57.8 | 48.3 | 27.1 | 33.6 | 35.8 |
1 1 = conventional tillage; 2 = reduced tillage;
2 averaged for all soil-landforms and crop rotations.
There were some trends but the absolute value differences were generally small. Reduced tillage tends to increase organic matter content (reduces mineralization). However, the LSRS calculation does not recognize increased organic amounts over a threshold of 4% so, as expected, the increases were mainly in the prairie zone which started with a rather low initial content. The two decreases, in Ecodistricts 593 in the north and 821 in the south, also occurred in areas of low initial carbon content. However, the increased erosion associated with conventional tillage was predicted to increase the clay content of the surface horizon (Appendix D) and the concomitant increase in water holding capacity more than compensated for the decrease in organic matter.
In this study we only defined "conventional" and "conservation" tillage regimes with essentially one less tillage operation in the latter (see Appendix 3 for descriptions). We did not include a "zero till" option because it represents, at the present time, a relatively small proportion of the area. It is possible that this extreme would have shown some significant differences.
Climate (Soil Zone, Ecoregion)
A comparison of equivalent soil-landforms and crop rotations in various regions indicates that climate affects soil quality assessments (Tables 2, 3, 4). However, this is not a simple cause and effect relationship. It is largely delivered through the impact of climate on crop rotation and management. There might be some direct impacts such as the effect of temperature on organic matter mineralization or wind on erosion but these are relatively minor compared to the vegetation cover aspects.
Area - Weighted Means
This is not, strictly speaking, a factor in soil quality assessment but rather a fallout of the procedure. However, the procedure can have a marked impact on values generated to represent large areas such as Ecodistricts. Where the component soil - landforms and crop rotations are similar or where the areas of each are similar, the impact may be negligible. But, in districts such as 828 (Table 6) where one crop rotation dominates, the difference is significant. Here the values for the various combinations range from -5% to +8% and a simple average would be +1%. However, the rotation with the -5 value occupies 70% of the area and the weighted mean of -4% is clearly a more appropriate representation.
Another example which illustrates the impact of area weighting is Ecodistrict 793 with markedly different soil ratings, cropping systems and the proportions of each (Table 5).
Table 5. Land suitability ratings for soils in Ecodistrict 793 under different crop rotations.
Crop Rotation |
Soil | Slope | Prop | initial | A
(7%) | B (22%) | D
(6%) | E
(3%) | H (25%) | L
(5%) | J
(33%) | weighted mean | % Change |
| LET | A | 55% | 53 | 50 | 51 | 51 | 55 | 50 | 52 | 52 | 51.2 | -3.4% |
| CIO | A | 20% | 48 | 48 | 49 | 49 | 49 | 48 | 50 | 49 | 48.8 | +1.7% |
| CRD | A | 15% | 50 | 47 | 50 | 50 | 50 | 47 | 50 | 49 | 48.8 | -2.4% |
| CMY | A | 10% | 29 | 20 | 22 | 24 | 25 | 22 | 25 | 20 | 21.6 | -25.5% |
Weighted mean | 49.15 | 46.15 | 47.55 | 47.75 | 50.05 | 46.35 | 48.60 | 47.75 | 47.62 | -3.1% |
% change | | -6.5% | -3.3% | -2.8% | +1.8% | -5.7% | -1.1% | -2.8% | -3.1% | |
The CMY soil is predicted to have a 9 point (31%) reduction in quality under crop rotations which have a high proportion of fallow (rotation A) or significant specialty crops (rotation J). These two rotations make up about 40% of the total in the area giving a mean reduction of 25.5%. However, CMY only accounts for 10% of the Ecodistrict. At the other extreme is the CIO soil which is predicted, under the L rotation (which has significant irrigation with forages), to have a 2 point or 4.2% increase in quality. It represents about 20% of the area. The dominant soil is LET (55%) with a predicted 30 year response of -3.4%. The resultant reduction in quality considering the proportions of all the soils is about 1.5 units or 3.3%.
Ecodistrict (Regional) Representation
The area-weighted mean of the SL, CR and TI options provides an index of overall soil quality in a region. The values for the 15 pilot areas based on a 30 year prediction of present conditions, suggest some clear trends (Tables 6 and 7, Figure 3).
Table 6. Land suitability ratings for various crop rotation and tillage options in the 15 pilot areas using 30 year EPIC predictions.
CR 1 | TI 2 | Ecodistrict Number |
593 | 596 | 598 | 680 | 681 | 727 | 728 | 730 | 746 | 743 | 781 | 793 | 806 | 821 | 828 |
| T0 (initial) | 50.1 | 55.3 | 55.8 | 58.3 | 64.5 | 71.6 | 64.7 | 64.2 | 83.4 | 52.2 | 56.8 | 49.2 | 27.1 | 35.3 | 35.2 |
| A | 1 | | | | | | | | | | | 57.4 | 46.2 | 26.6 | 34.1 | |
| 2 | | | | | | | | | | | 57.5 | 47.7 | 26.6 | 33.6 | |
| B | 1 | 51.3 | 55.0 | 56.4 | | | 71.6 | 64.7 | 64.5 | 83.4 | | 58.1 | 47.6 | | | 34.9 |
| 2 | 50.2 | 55.0 | 56.0 | | | 71.6 | 64.7 | 64.2 | 83.4 | | 58.1 | 48.3 | | | 34.7 |
| C | 1 | 51.3 | 54.7 | 56.0 | | | 71.6 | 64.7 | 64.6 | | | 58.1 | | | | |
| 2 | 50.2 | 54.7 | 56.2 | | | 71.5 | 64.7 | 64.7 | | | 58.1 | | | | |
| D | 1 | | | | | 64.5 | 71.6 | 64.7 | 64.5 | 83.4 | | 57.5 | 48.6 | | | |
| 2 | | | | | 64.5 | 71.6 | 64.7 | 64.5 | 83.4 | | 57.5 | 48.4 | | | |
| E | 1 | | | 56.2 | 58.8 | 64.5 | 71.6 | 64.7 | 64.5 | 83.4 | | | 50.1 | | | |
| 2 | | | 56.2 | 58.8 | 64.5 | 71.6 | 64.7 | 64.5 | 83.4 | | | 48.5 | | | |
| F | 1 | | | 56.2 | 59.1 | 65.1 | 71.6 | 65.3 | | 83.4 | | | | | | |
| 2 | | | 56.2 | 59.1 | 65.1 | 71.6 | 65.3 | | 83.4 | | | | | | |
| G | 1 | 52.6 | 59.0 | 57.0 | 58.8 | 65.4 | 71.8 | 65.3 | 64.8 | 83.4 | | | | | | |
| 2 | 51.5 | 59.0 | 57.0 | 58.8 | 65.4 | 71.8 | 65.3 | 64.8 | 83.4 | | | | | | |
| H | 1 | 51.3 | 54.7 | 55.4 | 58.8 | 65.1 | 71.6 | 64.7 | 64.7 | 83.4 | 52.7 | 57.5 | 46.4 | 26.6 | | 34.7 |
| 2 | 50.2 | 54.7 | 55.8 | 58.8 | 65.1 | 71.6 | 64.7 | 64.7 | 83.4 | 53.7 | 58.1 | 47.7 | 26.6 | | 34.0 |
| J | 1 | | | | 59.3 | 65.2 | 71.8 | 65.3 | 64.7 | 83.4 | | | 47.8 | 28.2 | | 36.3 |
| 2 | | | | 59.3 | 65.2 | 72.1 | 65.3 | 64.8 | 83.4 | | | 48.2 | 28.2 | | 36.6 |
| L | 1 | | | | | | | | | | | | 48.6 | | | 36.7 |
| 2 | | | | | | | | | | | | 49.2 | | | 38.0 |
Summary of T30 Ratings |
maximum | 52.6 | 59.0 | 57.0 | 59.3 | 65.4 | 72.1 | 65.3 | 64.8 | 83.4 | 53.7 | 58.1 | 50.1 | 28.2 | 34.1 | 38.0 |
minimum | 50.2 | 54.7 | 55.4 | 58.8 | 64.5 | 71.5 | 64.7 | 64.2 | 83.4 | 52.7 | 57.4 | 46.2 | 26.6 | 33.6 | 34.0 |
wtd mean | 51.7 | 55.2 | 56.3 | 58.8 | 65.0 | 71.7 | 64.8 | 64.6 | 83.4 | 53.0 | 57.7 | 47.7 | 26.6 | 34.0 | 34.7 |
1 CR = crop rotation: A = wheat-fallow; B = wheat; C = wheat-oilseeds; D = wheat-barley; E = barley; F = barley-forages; G = forages; H = mixed (no dominant); L = significant irrigation; J = significant specialty crops. See appendix 4 for more detail.
2 TI = tillage intensity: 1 = conventional; 2 = reduced (see Appendix 3 for more detail)
The subhumid (cooler) forested regions, the Boreal Transition Ecoregion with Grey and Dark Grey soils, are represented by Ecodistricts 593, 596 and 598 in the Peace region and by 680 and 681 in the Barrhead and Bonnyville areas. Predicted changes range from +3.3% to -0.8%. Most of the values are in the less than 1% range which is essentially a neutral or sustaining situation. A notable exception is the 80% hay rotation in district 596 which predicts a change of +7% for mainly Grey soils. This was based on a substantial increase in organic matter content and decreased erosion on some strongly undulaing landscapes. However this rotation only accounted for about 10% of the area with the other 90% being slightly negative so the overall impact was not large.
Table 7. Percentage change in land suitability ratings after 30 years.
ecodistrict number | maximum | minimum | Weighted mean |
593 | 5.1% | 0.2% | 3.3% |
596 | 6.7% | -1.2% | -0.2% |
598 | 2.2% | -0.7% | 0.9% |
680 | 1.6% | 0.8% | 0.8% |
681 | 1.3% | 0.0% | 0.8% |
727 | 0.6% | -0.1% | 0.2% |
728 | 0.9% | 0.0% | 0.2% |
730 | 0.7% | 0.0% | 0.6% |
746 | 0.0% | 0.0% | 0.0% |
743 | 3.0% | 1.0% | 1.5% |
781 | 2.2% | 1.0% | 1.6% |
793 | 1.8% | -6.1% | -3.1% |
806 | 4.3% | -1.9% | -1.7% |
821 | -3.3% | -4.7% | -3.7% |
828 | 8.0% | -3.3% | -1.3% |
Figure 3. Predicted percent change in land suitability ratings for the pilot areas
. 
Figure 4. Areal distribution of land suitability trends in Alberta
The subhumid Parkland region with Black soils (Ecodistricts 727, 728 and 730) predicted positive changes of 0 to about 1%. There is very little variation among the crop rotation and tillage options in this region. There are several factors contributing to this result. First, nearly all the rotations are dominated by cereals with very little fallow. Second, those systems with substantial amounts of hay, such as G and F, did show an increase in organic carbon but these soils are already quite high in organic matter and any amount over a 4% threshold is not considered by the LSRS method.
The semi-arid to subhumid Grasslands with Dark Brown soils are represented by Ecodistricts 743, 781 and 793. There is a general positive increase of 1 to 2 units (Table 6) in the Provost (743) and Drumheller (781) areas. However in the Lethbridge area (793), the high fallow rotations (A and H) showed significant declines in quality rating. This was particularly true for the sandy CMY soil which was marginal to start with.
The drier, semi-arid Grasslands with Brown soils were dominated by rotations with 40% to 50% fallow (A and H). All showed declines in soil quality (LSRS rating) of 1 to 2 units under dryland conditions. Where the rotations were identified with hay (L), these were assumed to be irrigated and a marked increase in organic matter resulted in higher ratings. However, these rotations only accounted for a small proportion of the area and the weighted means were still negative.
In general, the EPIC generated predictions suggest that present practices are maintaining soil quality in the more humid areas (Grey, Dark Grey, Black and some Dark Brown soils). In the drier areas (Brown soils) and where there are marginal soil conditions (sandy or hilly), soil quality is predicted to decline. The principal contributing factor to the assessment is the level of organic matter. Those factors such as fallow and erosion, which decrease the amount of organic matter in the soils, decrease the ratings. Factors such as continuous cropping and inclusion of hay in the rotation (combined with low initial values of organic matter) result in a predicted increase in organic matter and soil quality rating.
Strengths and Limitations of the Process and Tools.
The EPIC model has performed well for specific Alberta situations (T. Goddard and C. Izaurralde, pers. comm.) and was accepted for this study as given. It was developed using site applications and many of the detailed inputs were simplified or averaged for the regional approach (cf. Haugen-Kozyra et al. 1995, 1996, 1997). For example, seeding rates and row spacings were standardized, climate attributes and seeding dates were assigned by region. While the absolute values may not be correct for specific situations, the regional assessment should be close and it will not affect the comparative analysis made here. Again the nature of the input data will tend to dampen the amount of variation at a regional level.
The LSRS method was designed to operate using reconnaissance soil survey information and matches the level of data input for this study quite well. It was constructed to identify and rate limitations to plant growth and as such is sensitive to changes at the boundaries of growing conditions but not at the centre of the range. For example, a one half unit change in pH from 6.8 to 6.3 does not affect the output at all but a change from 5.8 to 5.3 results in a drop in capability or soil quality rating. Similarly, any change in organic carbon greater than 4% will result in no change in the rating while a change from 1.5% to 1%, where it affects both nutrient status and structure, causes a 5 point reduction in rating. Therefore, while the system is not sensitive to small changes in the inputs under all situations, it can identify major impacts on plant growth potential. Again, the averaging of data inputs as well as the results reduces variability.
The area weighting procedure, by its very nature, dampens variability. It provides an excellent representative summary for regional assessment but loses the extreme situations. Using Ecodistrict 793 (Table 5) as an example, the effect of averaging (consolidating or pro-rating) can be clearly illustrated. The overall predicted 30 year trend for the district, considering all soil-landform, crop rotation and tillage options on a proportional basis, is a decrease of 3.1%. Looking at the different crop / management systems averaged over the district show variations from -6.1% to +1.8%. If, however, one considers specific site characteristics such as individual soil-landform, crop rotation and tillage options, then the differences range from 30% to +12 5. The reality is that even though an area in total might be maintaining soil quality, there could be some specific combinations of factors that might be very degrading. The practical implication is that it is possible to identify and concentrate conservation extension efforts on only those combinations of environment and management that lead to significant loss of soil quality. While the protocol developed in this study provides a single "index" for regional level assessments, it also identifies specific contributing combinations of factors. It is important to recognize that while the "scaling up" process involves a loss of detail, the questions or objectives should also be more general at regional levels.
It is the identification and assessment of individual contributing factors, that distinguishes the present protocol. Previous regional analyses using EPIC (Izaurralde et al. 1996, Rosenberg et al. 1997) selected "typical" farm situations to represent a region and then modeled soil characteristics and yield over time. This approach simplified the input requirements, represented real-life management and allowed for both regional assessment and inter-regional comparison. However, it is entirely dependent on the selection of the single typical farm and its response to selected management inputs. It does not represent the range of environmental conditions in region and is not able to identify or assess the contributions of the various soil and management combinations that may be found in a particular region. Based on the excellent experience available for site selection, the single value from a typical farm approach will often be very similar to the mean for an area. However, there are other situations where it will simply not be possible to truly represent the region. This will occur when the selected site, which might well be typical or represent the greatest area, is close to an extreme (cf. Figure 3, Ecodistrict 828). Or, if representing the mean, a single site will not identify dangerously degrading situations (cf. Table 5, CMY soil).
The procedure developed here has a high requirement for data input and data management. However, the data are available and, with some reasonable documented management assumptions, it is well suited for scaling objectives. The present study considered only natural resource and management components and specifically omitted economic aspects. The two tools used, LSRS and EPIC, performed adequately for the present objective, but other tools could be substituted for different objectives - limited only by data input requirements.
This information is provided by W.W. Pettapiece, K.L. Haugen-Kozyra and L.D. Watson. |
|