Impact of Soil Management Practices on the Net GHG Fluxes

As we have discussed, several agricultural practices have great potential to increase carbon sequestration, and in some cases, decrease the net emissions of GHG. However, to date, very few field studies have been made that simultaneously examine C sequestration together with the other GHG emissions. Hence, we often have to rely on model estimates. The following analysis relies extensively on the application of the soil carbon model CENTURY (Parton et al., 1993), and on the estimation of N2O emissions based on analysis with DNDC (Li et al., 1992 a, b). Simulations were carried out using soil and climate data from 5 locations in Canada (Table II) to estimate the influence that selected changes in agricultural management practices are likely to have on N2O and CO2 fluxes.

To calculate changes in soil C stock (0-20 cm depth) and stabilization of the soil organic C pools, the CENTURY model was run from 1910 to 2029. Crop production yields were calibrated for each run. Management practices, including tillage, fertilizer addition, etc., were taken from the work by Smith et al. (1997). The DNDC model was run for the period between 1970 and 2029. Weather data for the 1970-1999 period were obtained from weather stations close to each site. The same weather data were used to simulate the weather for the next 30-yr period (i.e., 2000 to 2029).

A description of the soils, fertilizer application and climate at the 5 sites (Table II) is as follows:

1. The Dark Brown Chernozem soil at the Lethbridge, Alberta site has a sandy clay loam texture changing to a clay loam below 30 cm depth. A rotation of wheat-wheat-fallow was simulated with a fertilizer application of 15 and 40 kg ha-1 N applied in the 1st and 2nd year of the rotation, respectively. Long-term annual precipitation averages just over 400 mm, moisture deficits (evapotranspiration minus precipitation) average 268 mm and the annual air temperature averages 5.5 0C (Campbell et al., 1990). Snow packs are discontinuous and wintertime soil surface (5 cm depth) temperatures can range between -20 0C and 10 0C, undergoing numerous freeze-thaw events during Chinook wind occurrences that are common to this region.

2. The Brown Chernozemic soil at the Swift Current, Saskatchewan site is primarily of loam texture. A wheat-wheat-fallow rotation was also run at this location with fertilizer application of 15 and 40 kg ha-1 N applied in the 1st and 2nd year of the rotation, respectively. Long-term precipitation

TABLE II

Modeled cumulative change, over the next 30 years, of the combined N2O and CO2 emissions due to a change in management practice from conventional tillage

150% 50% Reduced

Fertilizer Fertilizer Permanent summer Forage in application application No-tillage grassland fallow rotation

150% 50% Reduced

Fertilizer Fertilizer Permanent summer Forage in application application No-tillage grassland fallow rotation

Mg CO2

equivalent ha

i

\)N2O

Three hills

— 1.6

2.6

13.1

10.3

Lethbridge

—7.1

2.1

5.1

5.0

—0.8

Swift current

—9.4

2.1

6.5

6.5

0.9

Harrow

—24.0

17.4

15.9

15.5

6.5

St. Foye

—16.1

26.9

24.6

29.3

19.5

3)CO2

Three hills

—0.1

0.0

34.9

111

Lethbridge

5.6

—5.0

11.6

101

13.1

Swift current

5.7

—1.7

11.5

93

11.9

Harrow

1.7

—15.0

23.9

201

20.1

St. Foye

2.1

—16.4

22.1

202

19.2

:) Combined N2O & CO2

Three hills

—1.7

2.6

48.0

121

Lethbridge

—1.5

—2.9

16.7

106

12.3

Swift current

—3.7

0.4

18.0

100

12.8

Harrow

—22.3

2.4

39.8

217

26.6

St. Foye

— 14.0

10.5

46.7

231

38.7

Note. Plus denotes a reduction in net GHG emissions and minus denotes an increase.

Note. Plus denotes a reduction in net GHG emissions and minus denotes an increase.

averages 334 mm, moisture deficits average 395 mm and the annual air temperature averages 3.3 °C (Campbell et al., 1990). Winter snow packs are often discontinuous, thus surface soil (5 cm) temperatures can reach -20 °C.

3. The Three Hills, Alberta site is located in the Dark Brown-Black soil transitional area, having a soil with clay-loam over clay texture. A continuous wheat system was simulated with fertilizer applied at 70 kg N ha -1. Long-term precipitation averages 408 mm, moisture deficits average 250 mm, and the annual air temperature averages 3.4 °C.

4. The soil at the Harrow, Ontario site has a poorly drained clay-loam soil (Ortho-Humic Gleysol) on which we simulated a corn-corn-barley-barley rotation with 180 kg ha-1 fertilizer N applied to the corn and 70 kg ha-1 to the barley. Long-term precipitation in this region averages 827 mm, moisture deficits average 86 mm and the annual temperature averages 8.7 °C.

5. The research station near Ste Foye, Quebec is approximately 30 km south of Quebec city. The soil is a clay loam belonging to the Gleysolic soil group. We simulated a corn-corn-barley-barley rotation with 180 and 70 kg ha-1 fertilizer N applied to corn and barley, respectively. Growing season precipitation averages 640 mm, moisture deficits are less than 100 mm, and the

1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030

Year

Figure 1. Annual changes in soil C for various management changes for a wheat-wheat-fallow rotation in Lethbridge, Alberta (1980-2029).

1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030

Year

Figure 1. Annual changes in soil C for various management changes for a wheat-wheat-fallow rotation in Lethbridge, Alberta (1980-2029).

mean annual air temperature averages 4.10C. While soils are only frost-free for about 150 days, they remain only slightly below 00C for most of the winter as snow packs can reach depths of 1 m.

The following changes in management were simulated: conversion from conventional tillage to no-tillage, conversion from cultivated land to permanent grassland, elimination of summer fallow, introduction of forage in a rotation, and a decrease by 50% and an increase by 150% of N fertilizer addition. Figure 1 presents an example of the soil C change predicted using the CENTURY model. It shows the cumulative soil C change over 30 yr for some of these management practices for a wheat-wheat-fallow rotation located near the Lethbridge research station in Alberta. The largest changes in soil C were due to the elimination of summer fallow and a change from conventional tillage to no-tillage (Figure 1).

CENTURY and DNDC models were also used to calculate the cumulative changes over 30 yr in CO2 and N2O emissions, in CO2 equivalents (Table II). The global warming potentials as given by IPCC (1996) were used for these calculations. The introduction of permanent grassland to previous cropland resulted in the largest net reduction in GHG. The largest changes were at the Ontario and Quebec sites where rainfall is less limiting to crop production than at the prairie sites. It was estimated that conversion of cropland to grassland reduced the combined emissions of CO2 and N2O by over 200 Mg CO2 equivalent ha-1. Conversion from conventional to no-tillage, reduction of summer fallow and introduction of forage into rotations all reduced net CO2 and N2O emissions. The increase in C sequestration was predominant over increased N2O emissions. The DNDC model generally indicated less N2O emissions due to the introduction of no-tillage, reduction of summer fallow and introduction of forage in the crop rotation. Reduction of fertilizer-N application reduced N2O emissions but also resulted in less C sequestration, while increasing N application increased C sequestration but also N2O emissions. In general, the models suggest that changes in the amount of N fertilizer applied will not have much effect on the net CO2 equivalent flux at most locations. The wide range of numbers in Table II clearly show that large errors can result if one merely multiplies the percent of arable land use by a carbon change coefficient when estimating the carbon mitigation potential of various land management practices.

The combined cumulative fluxes of N2O and CO2 from year 2000 to 2029 clearly indicate that several practices have the potential to reduce GHG emissions by shifting from conventional tillage to other management practices (Figure 2). This figure shows the relative net changes in CO2 equivalents due to CO2 and N2O emissions as compared to the control run which was business as usual using conventional tillage for a normally fertilized fallow-wheat-wheat rotation. Note that, at this site in Lethbridge, Alberta, the introduction of no-till, elimination of summer fallowing, and increased fertilization continued to reduce GHG emissions right up to the year 2029. Though the SOC pools have stabilized to some extent,

2000 2005 2010 2015 2020 2025 2030

Year

Figure 2. Predicted cumulative flux (Mg CO2 equiv. ha-1) over the period between 2000-2029 due to a change from conventional tillage to various management practices in Lethbridge, Alberta.

2000 2005 2010 2015 2020 2025 2030

Year

Figure 2. Predicted cumulative flux (Mg CO2 equiv. ha-1) over the period between 2000-2029 due to a change from conventional tillage to various management practices in Lethbridge, Alberta.

less N2O emissions are predicted to occur, particularly from permanent grassland. The other sites also showed similar trends (Table II).

5. Impact of Climatic Variations

Research results aimed at determining the influence of various management practices on C change in agricultural soils may be affected by climatic variations. It is well known that drought and excess precipitation can have a negative effect on crop production and thus C sequestration. In many semi-arid regions, a moderate increase in temperature accompanied by increased precipitation would likely lead to an increase in crop production. It is then possible that favorable weather over a number of years may enhance crop growth and inflate C sequestration rates for a particular management practice, whereas the practice may produce less C storage under normal weather conditions. The CENTURY model was run for a wheat-fallow rotation under conventional tillage (CT) using soil and weather data from a site near the Lethbridge Research station in Alberta in order to discern the impact of climate variations on SOC storage. Two simulations were carried out. First, the model was run for the period 1910-1999 using 30-yr climate normals, and then from 1910 to 1969 with 30-yr climate normals followed by monthly weather data from 1970 to 1999. For this case, the model predicts substantially more C sequestration (3 Mg C/ha or about 6%) for the period between 1983 and 1999 using monthly weather data (Figure 3). The increased C sequestration is not necessarily due to greater

1985

Year

Figure 3. Climatic effects on predicted soil C change at Lethbridge, Alberta for a wheat-fallow rotation under conventional tillage (CT) for the period between 1970-1999.

yearly rainfall, but more to differences in rainfall distribution during the growing season and a slight increase in temperature. Several other simulations were carried out for the same period using actual weather data and in many cases no change in soil organic matter were predicted. These results confirm that climatic variations from year to year can influence carbon change coefficients as determined from field data. It demonstrates that over a short period, climatic variations can have almost as large an impact as management practices. We emphasize that these are model projections and that they must be interpreted with care, however, supportive field observations have been reported by Campbell et al. (2001b).

0 0

Post a comment