Potential impacts of agricultural land use on soil cover in response to bioenergy production in Canada

Land Use Policy 75 (2018) 33–42

Contents lists available at ScienceDirect

Land Use Policy journal homepage: www.elsevier.com/locate/landusepol

Potential impacts of agricultural land use on soil cover in response to bioenergy production in Canada ⁎

T

⁎

Jiangui Liu , Ted Huffman , Melodie Green Ottawa Research and Development Centre, Agriculture and Agri-Food Canada, Ottawa, Ontario, K1A 0C6, Canada

A R T I C LE I N FO

A B S T R A C T

Keywords: Marginal land Bioenergy Crop residue Land use Sustainable Soil cover

The introduction of a market for agricultural biomass to feed large-scale second generation bioenergy (cellulosic ethanol) or other bio-products has positive implications for food producers and bio-product industries but may impact soil quality. In order to assess the potential impact on Canada’s agricultural lands, we integrated land use and soil capability maps and land management information to introduce several scenarios of crop residue harvest rates and land use conversions. The implications for soil quality, as represented by soil cover, were assessed for each scenario. The results showed that average soil cover at the national scale would decrease by more than 1 day if 40% of annual crop residues (by mass) were harvested, but the negative impacts could be resolved by increased adoption of conservation tillage methods. Additional biomass could be produced by converting low quality agricultural land to perennial biomass crops, but this would result in increased intensity of food production on high quality land. The total area of high and low quality land within the agricultural region of Canada is roughly equal, and the amount of high quality land currently used for perennial crops is about the same as the amount of low quality land used for annual crops, and ‘balancing’ production with capability would result in a net increase in both food and biofuel feedstock, with little impact to soil quality. About 6.73 M ha of high quality land is covered by forest, shrub and grass, and conversion of this land to agricultural production would have a negative impact on soil quality. The study indicates considerable potential for production of both food and biofuel feedstock on Canada’s agricultural lands through careful land use planning. Our analysis using soil cover as an indicator of environmental sustainability also indicates that land use planning should be cautious to prevent soil degradation. Particularly, regional variability of land use and soil capability distribution requires region specific land use policy for sustainable biofuel feedstock production.

1. Introduction The consumption of fossil fuels has contributed to an increase in atmospheric CO2, which is directly related with global warming. The use of biomass for biofuel production could mitigate global warming by reducing the dependence on fossil fuels and decreasing CO2 emissions (Naik et al., 2010). Current global bioenergy production is about 10% of that produced by fossil fuels (IPCC, 2012), and a greater potential can be anticipated from the world’s vegetated lands outside denser forests, croplands, urban areas and wilderness (Haberl et al., 2013; Guo et al., 2015). Hence, many countries have developed biofuel policies to boost an economic sector and a market for bio-products (HLPE, 2013). Both federal and provincial governments in Canada are promoting biofuel industry to exploit the huge stock of biomass resources (Le Roy and Klein, 2012). From census data it has been estimated that average

yearly potential bioenergy production from crop residue in Canada could be about 81.8 million barrels of ethanol (Li et al., 2012). However, questions remain on how an expanding bioenergy sector will interact with other issues such as food production, biodiversity, soil degradation, environmental sustainability and carbon sequestration (Berndes et al., 2003). First generation biofuels, consisting of bioethanol from grain and sugar and biodiesel from oilseeds, are considered to compete with food supply and thus increase food prices (Mueller et al., 2011; Mitchell, 2008). The development of second generation biofuels such as cellulosic ethanol may mitigate the impact of biofuel production on food supply. Crop residue and perennial woody/herbaceous bioenergy crops from agricultural landscapes are two important bioenergy supply chains (Smith et al., 2015; Mabee and Saddler, 2010). Growing perennial energy crops on surplus and/or poor quality agricultural land can be a

⁎ Corresponding authors at: Global Health and Tropical Medicine – Instituto de Higiene e Medicina, Universidade Nova de Lisboa (GHMT-IHMT-UNL), Rua da Junqueira 100, 1349-008 Lisboa, Portugal UMR BIPAR, INRA, ANSES, Ecole Nationale Vétérinaire d’Alfort, Université Paris-Est, 94700 Maisons-Alfort, France...... In addition, note that the ONLY ONE asterisk should appear after both names 'Ana Domingos' AND 'Alejandro Cabezas-Cruz. E-mail addresses: [email protected] (J. Liu), ted.huff[email protected] (T. Huffman).

https://doi.org/10.1016/j.landusepol.2018.03.032 Received 11 December 2017; Received in revised form 14 March 2018; Accepted 14 March 2018 0264-8377/ Crown Copyright © 2018 Published by Elsevier Ltd. All rights reserved.

Land Use Policy 75 (2018) 33–42

J. Liu et al.

as Soil Cover Days (SCD). To report at different spatial scales, the SCD is calculated according to the proportions of different crops and management practices. For perennial crops this include the frequency and timing of harvest or grazing, and for annual crops it includes different tillage types such as conservation/conventional/no-tillage, crop growth and residue decomposition. Detailed information on the indicator can be found in Huffman and Liu (2016) and Huffman et al. (2015).

major contributor to bioenergy production (Hoogwijk et al., 2003; Chemento et al., 2016; Feng et al., 2017). Agricultural land use and management practices can impact soil quality and greenhouse gas emissions (King et al., 2004; Smith et al., 2010). Potential impacts on greenhouse gas emissions by converting marginal land to perennial energy crops has been assessed by Liu et al. (2017). Crop residues on the soil surface protect soil from the erosive forces of wind and rain, and provide the building blocks for soil organic matter (Johnson et al., 2010). Therefore, removing crop residue can cause both direct and indirect adverse impacts, over short and long terms (Allmaras et al., 2004; Wilts et al., 2004; Mann et al., 2002; Lindstrom, 1986). Determination of a sustainable residue removal amount requires an integrated approach (Muth et al., 2013; Muth and Bryden, 2013). Contrary to annual row crops that can contribute to soil carbon losses, the establishment of perennial crops for biofuel has the potential to increase soil carbon stocks and generate other ecosystem services such as wildlife habitat and runoff prevention (Awasthi et al., 2017; Kantola et al., 2017). In addition, incorporating perennial bioenergy crops into agricultural landscapes can make use of land that is marginal for annual crops, and hence alleviate competition with food crops (Awasthi et al., 2017; Feng et al., 2017). To support the development of sustainable bioenergy production from agricultural land, the availability of land resources and the impact on the environment need to be assessed. A review shows that estimated production capacity varies greatly among different studies because land availability and yield levels of energy crops are uncertain (Berndes et al., 2003). In the context of intensification of agricultural production and environmental conservation, agri-environmental indicators have been developed to evaluate and report on the status and trends of environmental quality impacted by agricultural production activities (e.g., Clearwater et al., 2016). Soil cover provides environment protection through diminished wind and water erosion, limited leaching and run-off, increased weed control and improved soil fertility (as references cited in Büchi et al., 2016). The Soil Cover Indicator was developed to evaluate the status of soil cover provided by vegetation, crop residues and snow at a regional scale in Canada (Huffman et al., 2012, 2015). It was further developed for field and farm level applications in Switzerland by integrating crop model simulations (Büchi et al., 2016). In this paper, a study was conducted mainly to understand the resource potentials for bioenergy production in Canada using several land use change scenarios and crop residue harvest levels, and to evaluate the impact on soil quality. Agricultural census data, a national land use map and a soil capability rating system were used as a base to estimate land resources, land use conditions and land management practices across Canada’s croplands. The impact of agricultural cellulosic biofuel on soil quality was assessed using the Soil Cover Indicator.

2.2. Soil and land use inventory In our scenarios we assumed that high quality soils would be used for annual food crop production, while poor quality soils would be exploited for perennial crop production. This trend is supported in a study on cropland change in Alberta, Canada (Zhang et al., 2014). In reality, economic benefit generally determines land use, but in the case of food versus fuel, we assume that legislation, incentives or competitive advantage would relegate biofuel crops to land less suited to annual cultivation due to low fertility, poor climate or physical limitations. In order to investigate the potential for land use changes involving food and biofuel crops on different land types, a national land use map at 30 m resolution was intersected with the Canada Land Inventory (CLI) Soil Capability for Agriculture maps at 1:250,000 scale. A circa 2000 digital land cover map for the agricultural regions of Canada developed by Agriculture and Agri-Food Canada (AAFC, 2009) was acquired and is referred to as Circa-2000 LC. The map consists of 13 general land cover types (Table 1), 5 of which were of interest in this study; Annual Cropland, Perennial Cropland, Forest, Shrubland, and Grassland. CLI is a comprehensive multi-disciplinary land inventory of rural Canada that categorizes agricultural land into Soil Capability classes based on characteristics of the soil as determined from detailed soil surveys (AAFC, 2013a). Table 2 provides definitions of the classes. Generally (and for this study) soils rated as Classes 1–3 are considered prime (or good) agricultural land, whereas Classes 4–6 are considered marginal (or poor) for agriculture. Class 7 has no capacity for arable agriculture or permanent pasture. A similar approach has been adopted to identify marginal land in USA (Gelfand et al., 2013; Feng et al., 2017). Digital versions of the appropriate CLI map sheets were downloaded from the AAFC website and georeferenced to Circa-2000 LC. All data manipulations were performed using ArcMap. The intersection of the land use map and the CLI maps was conducted in order to provide an inventory of the areas of different combinations of land-use and land-capability classes at regional, provincial and national levels. In particular, the area of annual crops, perennial crops and forest/shrub/grass on CLI soil classes 1–6 were of interest. Fig. 1 illustrates the national extent of the study at a generalized level.

2. Material and methods

2.3. Census data

2.1. The Soil Cover Indicator

The Census of Agriculture is conducted every 5 years in Canada, and collects a wide variety of land use, land management and economic information for every farm in the country (Statistics Canada, 2008). Census data interpolated to Soil Landscapes of Canada polygons (AAFC,

Soil Cover is included in the OECD list of farm management indicators (OECD, 2001). It is defined as “the equivalent number of days in a year that soil of agricultural land is covered with vegetation”. The concept was adopted in Canada and a model was developed to estimate the total equivalent number of days in a year that the land is covered by crop canopy, crop residue and snow (Huffman et al., 2012, 2015). It has been used as a tool for agri-environmental health assessment and reporting at national and regional scales, providing direct information on the impact of different crops and field activities on the risk of soil erosion. By integrating information on crop type, soil, climate and associated field activities, plant growth and residue decomposition are simulated using a crop calendar at a daily time step to quantify the fraction of soil under cover. Thus, the indicator can quantify not only the daily fraction cover but also the annual equivalent number of days the soil is covered for a given crop in a given region, which is referred to

Table 1 Circa-2000 land cover types; “*” denotes land use types considered in land use management scenarios in this study. Water Exposed Land Built-up Shrubland* Wetland Grassland* Annual Cropland* Perennial Cropland and Pasture* Forest/ Trees*

34

Land Use Policy 75 (2018) 33–42

J. Liu et al.

harvested areas AAi* and APi* as:

Table 2 Land capability classification for agriculture. Class

Description

1 2

Soils have no significant limitations for crops Soils have moderate limitations that restrict the range of crops or require moderate conservation practices Soils have moderately severe limitations that restrict the range of crops or require special conservation practices Soils have severe limitations that restrict the range of crops or require special conservation practices Soils have very severe limitations that restrict their capability in producing perennial forage crops, and improvement practices are feasible Soils are capable only of producing perennial forage crops, and improvement practices are not feasible Soils have no capacity for arable culture or permanent pasture Organic Soils (not placed in capability classes)

3 4 5 6 7 O

PAi * = AAi * / TA

(1)

PPi * = APi * / TP

(2)

where TA represents the total area of annual crops and TP represents the total area of perennial crops derived from the census data. Subscript “*” denotes no differentiation of the CLI classes. Using the proportions of areas calculated from the census data, and the total annual/perennial crop areas obtained from the land cover map, areas for a given annual (A′Aij) or perennial (A′Pij) crop i in a CLI class j (Section 2.2) can be calculated as:

A′Aij = PAi* × A′A*j

(3)

A′Pij = PPi* × A′P*j

(4)

A′A*j

A′P*j

and represent areas of annual and perennial crops in a where CLI class j, respectively, as derived from the land cover and the CLI maps.

2013b), was used in this study to estimate the distribution of specific crops and management practices required to model soil cover days using the soil cover indicator.

2.3.2. Tillage proportions Different levels of residue cover fractions are associated with different types of tillage practices, according to Kline (2000) and a crop management study carried out in Manitoba Department of Agriculture (1986). Information on the proportion of annual crop area prepared using each of three tillage types; conventional tillage, conservation tillage and no-till is collected as part of the agricultural census. For implementation of the soil cover model, we assumed that tillage practices were identical for all crops except for those that require tillage for harvest (potatoes) and those that are not tilled (pasture and hay.

2.3.1. Crop areas Area proportions of crops are needed in calculating the soil cover indicator at the regional scale, because different crops have different soil cover. While the Circa-2000 LC map provides the total area of annual and perennial crops, it does not differentiate specific crops such as hay, pasture, wheat or soybeans in a given spatial unit. However, harvested areas of both annual and perennial crops are included in the census data. Using the census data, area proportions of an annual (PAi* ) or a perennial crop (PPi* ) are calculated from their corresponding

Fig. 1. Agricultural region of Canada south of 60 °N represented by Circa-2000 LC, overlain with Canada Land Inventory (CLI) polygons and provincial boundaries.

35

Land Use Policy 75 (2018) 33–42

J. Liu et al.

in a variety of programs (Eilers et al., 2010). Ecoregions are components of Ecozones, and are the stratification level for which most of the input data for the soil cover indicator are available. We first scaled all land cover, CLI and Census data to the Ecoregion level following the methods introduced in Sections 2.2 and 2.3. Soil cover for each scenario was evaluated for each of the 85 ecoregions completely covering all agricultural areas in Canada:

2.3.3. Residue cover fraction at harvest Residue cover fraction after harvest is another input required for the soil cover indicator. It was estimated using crop yield (Statistics Canada, 2017) and crop-specific harvest indices. It is documented that crop grain yield y and aboveground residue mass M are related through a constant harvest index H using Eq. (5) (Bolinder et al., 2007), although variability in H can be induced by cultivars or environmental factors (Fan et al., 2017):

M = y (1−H )/ H

SCD = FA ∑ PAi * *(∑ t j * SCDij ) + FP ∑ PPi * * SCDi + (1−FA−FP )* 365

(5)

i

i

(7)

We adopted the harvest index found in Bolinder et al. (2007) to derive residue mass from census grain yields. Residue cover fraction Fr is related to residue mass M through the Beer-Lambert Law, as deduced by Gregory (1982):

Fr = 1−exp(−kM )

j

where FA and FP represent the proportions of annual and perennial crops in an ecoregion, respectively; PAi* and PPi* represent the proportions of annual or perennial crop area in crop type i , respectively; t j and SCDij represent the fraction of tillage type j and the correspondent annual soil cover days for crop type i . We assumed perennial crops are not tilled, and all annual crops have the same proportions of different tillage practices (except potatoes, for which there is no ‘no-till’ option) because no tillage information is available for individual crops. We also assumed that all land not used for crops (i.e. forest and shrub) have no exposed soil and therefore have 365 equivalent days of annual coverage. For presentation and discussion, results were rolled up to the provincial and national levels.

(6)

where k is the cover coefficient per unit residue mass, which has been estimated through regression using field data collection (Gregory, 1982; Steiner et al., 2000). Residues from different crops have different cover coefficients; for example cereal crops such as wheat and barley generally have higher cover coefficients than grain corn (Manitoba Department of Agriculture, 1986). In this study, we used 0.36 ha per ton (or 3.6 m2 per kg) for corn and 0.60 ha per ton (6.0 m2 per kg) for small grain cereal crops.

3. Results and discussion

2.4. Scenarios of change in land use and residue management

3.1. Inventory of land use conditions and land capability for agriculture

Several scenarios were designed to assess the impact on SCD of changes in land use and residue management practices induced by increasing markets for agricultural bio-products. For baseline conditions: 1) land use was based on the Circa-2000 LC; 2) areal proportions of crops and different tillage practices (no-till, conservation tillage and conventional tillage) were derived from census data; and 3) crop residue cover fractions after harvest were derived from reported yields. For residue management scenarios aboveground crop residue was removed at different percentages of dry weight. Crop residues were considered to be harvested only from buckwheat, flax, grain corn (maize), spring-seeded cereal grains (barley, wheat, oats) and winter wheat. The harvest index for these crops is 0.23, 0.26, 0.50, 0.41 and 0.41 respectively (Bolinder et al., 2007). Residue from other crops was not considered useful or profitable for biofuel feedstock. The land use scenarios, starting from baseline conditions and applying a 20% residue removal rate were considered as following:

Fig. 2 shows Census agricultural land use values every 5 years from 1981 to 2011. Average areas of “Total farm area”, “Land in crops”, “Summer Fallow”, “Tame or seeded pasture” and “Other land use” during this time period were about 67.06, 34.32, 6.09, 4.64 and 22.01 M ha, respectively. The area in crops increased steadily from 30.97 M ha to 35.35 M ha, whereas the area of summerfallow decreased from 9.07 M ha to 2.09 M ha. Generally, these land use categories were relatively stable or changed slowly, which may reveal actual variations in land use, or changes in definitions of categories and survey methods that have impacted the reported values or quality. However, in this study Census data were not used to quantify the absolute areas of land use, but were applied to the areas from the maps to quantify the relative proportions of different annual/perennial crops and different tillage methods. The land use planning scenarios took into consideration annual/ perennial crops, forest, shrubland and grassland. A national inventory of land use and land capability classes is given in Table 3. The total area of crops, forest, shrubland and grassland were approximately 50.08 (annual crops 34.05 and perennial crops 16.03), 55.42, 9.28 and 8.79 M ha, respectively. The total study area (124.29 M ha) is larger than that reported in the census data because the two maps (land use

1) LU-1: convert annual crops on CLI-456 soils (marginal land) to perennial energy crops. 2) LU-2: following LU-1, convert perennial crops on CLI-123 soils to annual crops. 3) LU-3: following LU-2, convert 1/3 of annual crops (all on CLI-123 soils) to perennial crops. 4) LU-4: following LU-3, convert equivalent amount of forest/shrub/ grass on CLI-123 soils to annual crops. 5) LU-5: starting from baseline conditions, convert all forest/shrub/ grass on CLI-123 soils to annual crops. 6) LU-6: starting from baseline conditions, convert 26% of corn area and 3% of spring wheat area to switchgrass. These two percentages correspond to the area of the two crops used for biofuel feedstock in Canada in 2010 (Tollefson and Madramootoo, 2012). 2.5. Quantification of soil cover A national eco-stratification framework has been developed in Canada to divide the entire landmass into a hierarchical classification system (Ecological Stratification Working Group, 1995; AAFC, 2013c). This framework has been used as a common basis to report on the state of the environment and the sustainability of agroecosystems in Canada

Fig. 2. Canada’s national agricultural land use condition based on census data (Census table 004-0002; once every five years).

36

Land Use Policy 75 (2018) 33–42

J. Liu et al.

where a decrease in SCD by more than 1.0 days still occurs. Compared with the situation of residue removal at 40% level without current tillage practices (Fig. 5c), promoting adoption of conservation tillage practices allows for more residue removal, although other consequences should also be considered, such as maintaining soil carbon content.

Table 3 Areas of different land capability classes (M ha), derived from Circa-2000 land cover map and the CLI maps. CLI class

Annual crops

Perennial crops

Forest

Shrubland

Grassland

Total

CLI 1–3 CLI 4–6 CLI 7 Organic

24.90 8.76 0.25 0.15

8.05 7.29 0.51 0.18

5.76 17.45 30.32 1.89

0.80 3.02 4.51 0.95

1.17 6.46 0.99 0.17

40.68 42.98 36.59 3.34

Total

34.05

16.03

55.42

9.28

8.79

123.58

3.3. Impact on soil cover by land use conversion Converting annual crops grown on marginal land to perennial energy crops (Scenario LU-1) would greatly increase regional soil cover. By converting about 8.76 M ha of these lands, the national annual SCD will increase by 7.7 days over the baseline conditions of 281.2 days. The largest increase is in the Mixed Grassland ecoregion in the Prairies (> 20 days; Fig. 6a), where the area of annual crops on marginal land consists of a larger proportion of the total crop area (∼46%). The proportion of total cropland consisting of annual crops on marginal land is below 10% in the ecoregions in the western portion of the Mixedwood Plains ecozone, where the increase of SCD is only around 5 days. Scenario LU-2 considers that the conversion of annual crops on marginal land to perennial bioenergy crops (LU-1) would increase the demand for annual crops elsewhere, resulting in the conversion of perennial crops on high quality soils (CLI 1–3) to annual crops. About 7.29 M ha of such land is available and even though the conversion to annual crop production would lower SCD values from LU-1 the cumulative effect of both scenarios would be an increase in national SCD of slightly more than 0.6 days over the baseline conditions. However, considerable regional variability is apparent (Fig. 6b); cumulative SCD increases noticeably in most areas in the Prairies whereas it decreases in most areas in eastern Canada, with exceptions along the northern fringes of the agricultural ecumene. This is primarily due to the higher proportion of perennial crops on high quality soil in the eastern regions, so that relatively more land is converted to annual crops. For instance, in the Mixed Grassland, Moist Mixed Grassland and the Aspen Parkland ecoregions in the Prairies, the proportion of total cropland represented by perennial crops on prime soil is about 2.1%, 5.5% and 11.6% respectively, whereas in the Manitoulin-Lake Simcoe and the Lake Erie Lowland ecoregions in Ontario, the proportion is about 35.4% and 31.1%, respectively. The combined effect of LU-1 and LU-2 on land use is that all annual crops are now grown on high quality soils and all perennial crops are on marginal land, and the total area of annual crops is about the same as the baseline conditions. Due to livestock and crop rotation needs, however, scenario LU-3 considers that 1/3 of the annual crop area determined under LU-1 and LU-2 would be converted to perennial crops. This would increase national SCD by 9.7 days over baseline conditions and the effect would be seen in all ecoregions. Noticeably, SCD in the eastern agricultural regions show relatively smaller changes, but remain close to baseline conditions with either an increase or decrease by no more than 5 days (Fig. 6c). Following and cumulative with the previous 3 scenarios, LU-4 considers that an equivalent amount of forest, shrubland and grassland on high quality soils in each region is converted to annual crops as compensation for the loss of annual crops in LU-3. The national SCD increases by about 0.8 days compared to the baseline, but agricultural regions in the east and in the northern fringe areas in the west show decreases in SCD (Fig. 6d) due to larger proportions of forest, shrub and grass on good soils in these regions. From the baseline conditions, if all prime soils currently covered by forest, shrub or grass were converted to annual crops (LU-5), the total area of annual crops would increase by about 7.73 M ha nationally, an increase of about 22.3%. This would induce a decrease of SCD by about 14.0 days nationally. The decrease in eastern Canada and the northern fringe area of western agricultural regions would be quite severe, while the change in the cropland area of the Prairies would be minor (Fig. 6e). It has been estimated that about 26% of corn and 3% of spring

and soil capability maps) also cover areas not reported by farmers. Prime agricultural soils (CLI classes 1–3) are primarily used for annual crops (∼61%), with lesser amounts under perennial crops (∼20%) and forest, shrubs and grass (∼19%). Marginal agricultural land (CLI 4–6) is predominantly (63%) under forest, shrubs and grass, with 20% under annual crops and 17% used for perennial crops. Ninetyeight percent of Class 7 soils were assessed as being covered by forest, shrub and grass, but the small area mapped as crops on Class 7 must, by definition, indicate error in the maps. The distribution of land use on soil capability generally reflects landowners tendency to use land for greatest profitability, but the relatively large area of low-intensity use of high quality soils also shows considerable potential for increased agricultural production. The relatively small area of organic soils (O) is not rated for agricultural capability and is not considered in this study. Fig. 3 shows the distribution of annual and perennial crops in agricultural ecozones across Canada. Ecozones in the three Prairie Provinces (Alberta (AB), Saskatchewan (SK) and Manitoba (MB)), and in the Mixedwood Plains ecozone of Ontario (ON) and Quebec (QC) have a high (> 40%) proportion of annual crops, while perennial crops or other land use types dominate in other zones.

3.2. Impact on soil cover by residue removal Removal of residues from crop fields can reduce soil cover, particularly during the period when residue is the major component covering the soil surface, such as after harvest and before significant crop development. Annual SCD for annual crops decreases regularly with increasing rates of residue removal. Fig. 4 shows the change in average national SCD with increasing residue removal rates. Average SCD over cropland (annual and perennial) across Canada was 281.5 days when residue was not removed, and decreased by about 0.4, 2.2 and 3.3 days when residue was removed at rates of 20%, 40% and 60% by weight, respectively. Regional changes in average SCD at these three removal rates are shown in Fig. 5a–c. The extent of SCD decrease is associated with the proportion of area in annual crops as depicted in Fig. 3a. The area proportion of annual crops in southern AB and SK is the highest in the country (> 70%), and SCD will decrease by more than 1 day even at only 20% residue removal rate. Another area with a relatively high proportion of annual crops is in the Mixedwood Plains ecozone, where the proportion of annual crops is ∼40–60% and a reduction of SCD will exceed 1 day at 40% residue removal rate. In more northern regions the proportion of annual crops is considerably lower and SCD reduction only exceeds 1 day when residue is removed by more than 60%. The above assessment is based on current (2006) tillage practices in each region, as derived from Census data. Conservation tillage and notill practices have been increasingly adopted for soil and environmental protection in Canada over the past 20–30 years. A scenario analysis was conducted to evaluate the combined effect of conservation tillage and residue removal on SCD; crop residue was simulated as being removed by 50%, whereas tillage was simulated as 50% conservation tillage and 50% no-till instead of the actual conditions as reported in the census. The result is shown in Fig. 5d, where it can be seen that SCD increases in most regions except for the semi-arid region of southern SK and AB 37

Land Use Policy 75 (2018) 33–42

J. Liu et al.

Fig. 3. Percent of annual (a) and perennial (b) crops respective to total agricultural area, using ecoregion as spatial unit.

wheat produced was used for biofuel in Canada in 2010 (Tollefson and Madramootoo, 2012). From baseline conditions, if this amount of land were used to grow perennial energy crops (LU-6), the national SCD would increase by about 0.8 days, regardless of the potential difference in bioenergy production. Most agricultural regions, particularly the Prairies and the northern regions in eastern Canada, would show a minor increase of SCD, while in southwestern Ontario and the StLawrence Lowlands ecoregion, SCD would undergo an increase of more than 5 days due to a large area increase of perennial crops, as corn is a major crop in this region. Fig. 4. Change of soil cover days (SCD) with percent of residue removal at the national scale.

3.4. Discussion This study evaluated the impact of developing bioenergy supply chains on soil protection, using the Soil Cover Indicator developed in Canada as a criterion. Several scenarios were designed to take into 38

Land Use Policy 75 (2018) 33–42

J. Liu et al.

Fig. 5. Changes of soil cover due to residue removal of 20% (a), 40% (b), 60% (c) with current tillage practices, and residue removal of 50% with 50% conservation tillage and 50% no-till (d); removal rate is based on mass.

consideration two agricultural bioenergy supply chains for second generation (biomass ethanol) bioenergy; harvest of annual crop residues and growing perennial energy crops on agricultural land,

including exploitation of lands currently considered marginal for profitable annual crop production. This assessment is directed toward supporting policy development to ensure a sustainable supply of food 39

Land Use Policy 75 (2018) 33–42

J. Liu et al.

Fig. 6. Changes of soil cover respective to baseline conditions for different land use scenarios*: a) LU-1; b) LU-2; c) LU-3; d) LU-4; e) LU-5; f) LU-6. *Scenarios 1–5 are cumulative; Scenario 6 is independent.

status in a region. This also demonstrates that through proper land use planning and residue management strategies, crop residue can be removed for bioenergy production without significantly impact soil cover status. Except for the specific residue harvest and land use scenarios presented in the study, other management practices which play important roles in soil cover were not considered. For example, under base conditions we assume that: 1) all residues are left in the field at grain harvest and the amount in weight and the cover fraction can be estimated from the grain yield; 2) the change of residue cover fraction occurs at the time of straw baling at typical dates for a region after grain harvest; and 3) tillage will reduce residue cover by a given fraction according to the method used. For instance, residue cover fraction will reduce to 10% or 50% of its original level using conventional or conservation tillage, respectively (Manitoba Department of Agriculture, 1986). A change in timing of field operations will also induce a change in soil cover. From Canada’s land use and land capability inventory analysis, the area for high quality soils (40.68 M ha) is larger than the area for annual crops (34.05 M ha), and about 20% of low quality soils (marginal land) are used for annual crops (Table 3). The implication from this information is that, annual crops on low quality soils can be exchanged with perennial crops on high quality soils without impacting regional soil cover and food production. Under the pressure of growing

and biofuel feedstock. The intention was to assess the land resources and residue biomass availability rather than to analyze the complete cycle of the bio-economy; therefore, biomass yields from perennial crops at different regions and the energy equivalence from biomass were not considered. Soil cover is an agri-environmental indicator incorporating climate conditions, crop and residue management practices in quantification of the annual equivalent days that soil is protected from exposure in a region. It is influenced by agricultural management practices (Huffman et al., 2015), such as crop types and their areal proportions, crop calendar and plant morphology, tillage and residue management methods. It is noted that reduction of equivalent soil cover days due to residue removal is compromised by three factors. First, soil cover in this study is evaluated at the regional scale rather than field scale. Removal of crop residue only reduces soil cover of annual crops, whereas perennial crops and other land use types are not affected; therefore the equivalent reduction of soil cover appears to be small. Second, residue only affects soil cover significantly during a relatively short period of time in a year, i.e., from crop harvest to snow fall in autumn, and from snow melt to crop rigorous growth in spring. Thirdly, the relationship between residue cover and residue mass follows an exponential equation (Eq. 6). Reduction of percent residue cover is lower than percent removal of residue by mass. These three factors jointly impact soil cover 40

Land Use Policy 75 (2018) 33–42

J. Liu et al.

Martin of AAFC is greatly appreciated.

population for annual crop area expansion, regional soil cover can only be maintained through improved residue and tillage management practices. In this study, area proportions of crops derived from census data were used to derive regional equivalent soil cover values. Crop proportion is important because different crops have different soil cover conditions. For example, the life cycle of soybeans in eastern Canada is generally shorter than that of corn and it retains much less biomass after senescence (Huffman et al., 2015; Liu et al., 2016), thereby resulting in a lower SCD than corn. Conservation tillage methods have been widely adopted to protect soils and the environment, and while this leads to an increase in SCD for many crops, intensive tillage is required for some conditions and types of crops, such as potatoes. Perennial herbaceous crops provide much higher SCD values than annual crops. SCD for an individual field in a single year is completely dependent on crop type and field management practices, but areal and temporal average SCD values for a farm or a region are highly dependent on longer-term farming practices and crop rotations. It is well accepted that crop residue cover is critical to protect soil from wind and water erosion, and in fact soil loss through erosion is directly related to residue cover fraction (Gregory, 1984; Bilbro et al., 1994). Under best management practices guidelines, it is recommended that a minimum of 30% soil cover be retained in order to achieve sufficient environmental protection (FAO, 2015). However, it is also acknowledged that the effect for protecting soil from wind and water is dependent on residue type and orientation (standing versus flat). For instance, the cover coefficient per unit mass of residue for small grain crops such as oats or wheat is generally larger than that of corn (Gregory, 1982; Manitoba Department of Agriculture, 1986). Maintaining soil carbon balance is another ecosystem service that crop residue can provide, and for this an integrated framework is needed to take into consideration soil properties, climate conditions and management practices in order to model the biochemistry, field operations and erosion factors (Muth and Bryden, 2013; Muth et al., 2013). It will be the orientation of our future research efforts.

References AAFC (Agriculture and Agri-Food Canada), 2009. Land Cover for Agricultural Regions of Canada, Circa 2000. Accessed on October 5, 2017. http://open.canada.ca/data/en/ dataset/16d2f828-96bb-468d-9b7d-1307c81e17b8. AAFC, 2013a. Canada Land Inventory. Accessed on October 5, 2017. http://sis.agr.gc. ca/cansis/nsdb/cli/index.html. AAFC, 2013b. Interpolated Census of Agriculture by Soil Landscapes of Canada. Accessed on October 5, 2017. http://open.canada.ca/data/en/dataset/9c285bb1-7919-426ab6c0-29a4d2edde48. AAFC, 2013c. A National Ecological Framework for Canada. Accessed on October 5, 2017. http://sis.agr.gc.ca/cansis/publications/manuals/1996/index.html. Allmaras, R.R., Schomberg, H.H., Douglas, Jr.C.L., Dao, T.H., 2004. Soil organic carbon sequestration potential of adopting conservation tillage in US croplands. J. Soil Water Conserv. 55, 365–373. Awasthi, A., Singh, K., Singh, R.P., 2017. A concept of diverse perennial cropping systems for integrated bioenergy production and ecological restoration of marginal lands in India. Ecol. Eng. 105, 58–65. Berndes, G., Hoogwijk, M., Van Den Broek, R., 2003. The contribution of biomass in the future global energy supply: a review of 17 studies. Biomass Bioenergy 25 (1), 1–28. Bilbro, J.D., Harris, B.L., Jones, O.R., 1994. Erosion Control With Sparse Residue. In, USDA ARS Conservation Report #37, Crop Residue Management to Reduce Erosion and Improve Soil Quality. Southern Great Plains. 84pp.. . Bolinder, M.A., Janzen, H.H., Gregorich, E.G., Angers, D.A., VandenBygaart, A.J., 2007. An approach for estimating net primary productivity and annual carbon inputs to soil for common agricultural crops in Canada. Agric. Ecosyst. Environ. 118 (1), 29–42. Büchi, L., Valsangiacomo, A., Burel, E., Charles, R., 2016. Integrating simulation data from a crop model in the development of an agri-environmental indicator for soil cover in Switzerland. Eur. J. Agron. 76, 149–159. Chemento, C., Almagro, M., Amaducci, S., 2016. Carbon sequestration potential in perennial bioenergy crops: the importance of organic matter inputs and its physical protection. Glob. Change Biol. Bioenergy 8, 111–121. Clearwater, R.L., Martin, T., Hoppe, T. (eds.), 2016. Environmental Sustainability of Canadian Agriculture. Agri-Environmental Indicators Report Series Report #4. ISBN No. 978-0-660-04855-0, 245pp. Ecological Stratification Working Group, 1995. A National Ecological Framework for Canada. 132 pp.. Accessed on July 26, 2017. http://sis.agr.gc.ca/cansis/ publications/ecostrat/index.html. Eilers, W., MacKay, R., Graham, L., Lefebvre, A., 2010. Environmental Sustainability of Canadian Agriculture. Agri-Environmental Indicator Report Series – Report #3. Agriculture and Agri-Food Canada, Ottawa, Ontario. Fan, J., McConkey, B., Janzen, H., Townley-Smith, L., Wang, H., 2017. Harvest index–yield relationship for estimating crop residue in cold continental climates. Field Crops Res. 204 (15), 153–157. FAO, 2015. Food and Agriculture Organization of the United Nations: Conservation Agriculture. Accessed July 25, 2017. http://www.fao.org/ag/ca. Feng, Q., Chaubey, I., Engel, B., Cibin, R., Sudheer, K.P., Volenec, J., 2017. Marginal land suitability for switchgrass, Miscanthus and hybrid poplar in the Upper Mississippi River Basin (UMRB). Environ. Modell. Softw. 93, 356–365. Gelfand, I., Sahajpal, R., Zhang, X., Izaurralde, R.C., Gross, K.L., Robertson, G.P., 2013. Sustainable bioenergy production from marginal lands in the US Midwest. Nature 493, 514–517. Gregory, J.M., 1982. Soil cover prediction with various amounts and types of crop residue. Trans. Am. Soc. Agric. Eng. 25 (5), 1333–1337. Gregory, J.M., 1984. Prediction of soil erosion by water and wind for various fractions of cover. Trans. Am. Soc. Agric. Eng. 27 (5), 1345–1350. Guo, M., Song, W., Buhain, J., 2015. Bioenergy and biofuels: history, status, and perspective. Renew. Sustain. Energy Rev. 42, 712–725. Haberl, H., Erb, K.H., Krausmann, F., Running, S., Searchinger, T.D., Smith, W.K., 2013. Bioenergy: how much can we expect for 2050? Environ. Res. Lett. 8 (3), 031004. HLPE, 2013. Biofuels and Food Security. A Report by the High Level Panel of Experts on Food Security and Nutrition of the Committee on World Food Security, Rome. 132pp.. . Hoogwijk, M., Faaij, A., Van der Broek, R., Berndes, G., Gielen, D., Turkenberg, W., 2003. Exploration of the ranges of the global potential of biomass energy. Biomass Bioenergy 25, 119–133. Huffman, E.C., Liu, J., 2016. "Soil cover". In: Clearwater, R.L., Martin, T., Hoppe, T. (Eds.), Environmental Sustainability of Canadian Agriculture: Agri-environmental Indicator Report Series - Report #4. Agriculture and Agri-Food Canada Chapter 6, pp. 53–63. Huffman, T., Coote, D.R., Green, M., 2012. Twenty-five years of changes in soil cover on Canadian Chernozemic (Mollisol) soils, and the impact on the risk of soil degradation. Can. J. Soil Sci. 92, 471–479. Huffman, T., Liu, J., Green, M., Coote, D., Li, Z., Liu, H., Liu, T., Zhang, X., Du, Y., 2015. Improving and evaluating the soil cover indicator for agricultural land in Canada. Ecol. Indic. 48, 272–281. IPCC, 2012. Special Report on Renewable Energy Sources and Climate Change Mitigation. Cambridge University Press, Cambridge 246pp. Johnson, J.M.F., Karlen, D.L., Andrews, S.S., 2010. Conservation considerations for sustainable bioenergy feedstock production: if, what, where, and how much? J. Soil Water Conserv. 65 (4), 88A–91A. Kantola, I.B., Masters, M.D., DeLucis, E.H., 2017. Soil particulate organic matter increases under perennial bioenergy crop agriculture. Soil Biol. Biochem. 113, 184–191.

4. Conclusions The intersection of soil resources and land use over Canada’s agricultural lands has been investigated using a land use map and a map of land capability for agriculture. Results showed that within the current agricultural extent, some high quality soils are not currently used for annual crop production, while about 8.76 M ha of annual crops are grown on low quality (marginal) soils. Based on our results, it appears that there is considerable potential to ensure ongoing food production as well as promote bioenergy production in Canada. Soils that are marginal for annual crops may be profitable for perennial energy crops and could be exploited for biofuel stock supply. However, the current system of private land ownership in Canada may be a deterrent to rational adjustment to a situation of ‘highest and best use’ of all classes of land, and some incentives and/or legislation that considers regional variability of soil capability and land use conditions may be required in order to maintain food production, increase bioenergy supplies and protect soils from degradation. Crop residues provide an important source of bioenergy feedstock, but they are also important for protecting soils from wind and water erosion and carbon loss. Land use change and residue harvest should be conducted in a sustainable manner, taking into consideration new technologies and approaches to soil conservation, crop rotation and crop management. Acknowledgements This study was supported by Agriculture and Agri-Food Canada (AAFC) under the Sustainability Metrics project (#J-000486), the Bioproducts Targeted Call project (#J-001598) and the Integrated Soil and Land Use data project (#J-001383). Revision and comments by Tim 41

Land Use Policy 75 (2018) 33–42

J. Liu et al.

Muth, D.J., Bryden, K.M., Nelson, R.G., 2013. Sustainable agricultural residue removal for bioenergy: a spatially comprehensive US national assessment. Appl. Energy 102, 403–417. Naik, S.N., Goud, V.V., Rout, P.K., Dalai, A.K., 2010. Production of first and second generation biofuels: a comprehensive review. Renew. Sustain. Energy Rev. 14, 578–597. OECD, 2001. Environmental indicators for agriculture. Methods and Results, vol. 3 OECD Publishing, Paris 400pp. Smith, C.T., Lattimore, B., Berndes, G., Bentsen, N.S., Dimitriou, I., Langeveld, J.W.A., Thiffault, E., (Eds.), 2015. Mobilizing Sustainable Bioenergy Supply Chains – InterTask Project Synthesis Report. IEA Bioenergy ExCo: 2015:04. ISBN 978-1-910154-205 (eBook electronic edition). 170pp. Smith, P., Bhogal, A., Edgington, P., Black, H., Lilly, A., Barraclough, D., Worrall, F., Hillier, J., Merrington, G., 2010. Consequences of feasible future agricultural land-use change on soil organic carbon stocks and greenhouse gas emissions in Great Britain. Soil Use Manag. 26 (4), 381–398. Statistics Canada, 2008. 2006 Census of Agriculture. Available at. http://www.statcan. gc.ca/pub/95-629-x/2007000/4182413-eng.htm. Statistics Canada, 2017. Estimated Areas, Yield and Production of Principal Field Crops by Small Area Data Regions, in Metric and Imperial Units, Annual, 1976 to 2016. http://www5.statcan.gc.ca/cansim/a26?lang=eng&id=10071. Steiner, J.L., Schomberg, H.H., Unger, P.W., Cresap, J., 2000. Biomass and residue cover relationships of fresh and decomposing small grain residue. Soil Sci. Soc. Am. J. 64, 2109–2114. Tollefson, L., Madramootoo, C., 2012. Irrigated Biofuel Production in Canada. Global Biofuel Production. Accessed July 25, 2017. http://ppts.icidonline.org/korea_2014/ korea_crop_1.pdf. Wilts, R., Reicosky, D.C., Allmaras, R.R., Clapp, C.E., 2004. Long-term corn residue effects: harvest alternatives, soil carbon turnover, and root-derived carbon. Soil Sci. Soc. Am. J. 68, 1342–1351. Zhang, X., Huffman, T., Liu, J., Liu, H., 2014. Soil capability as a predictor of cropland change in Alberta, Canada from 1988 to 2010. Soil Use Manag. 30, 403–413.

King, J.A., Bradley, R.I., Harrison, R., Carter, A.D., 2004. Carbon sequestration and saving potential associated with changes to the management of agricultural soils in England. Soil Use Manag. 20 (4), 394–402. Kline, R., 2000. Estimating crop residue cover for soil erosion control. The Soil Factsheet. Ministry of Agriculture and Food, British Columbia 4pp. Le Roy, D.G., Klein, K.K., 2012. The policy objectives of a biofuel industry in Canada: an assessment. Agriculture 2, 436–451. Li, X., Mupondwa, E., Panigrahi, S., Tabil, L., Sokhansanj, S., Stumborg, M., 2012. A review of agricultural crop residue supply in Canada for cellulosic ethanol production. Renew. Sustain. Energy Rev. 16 (5), 2954–2965. Lindstrom, M.J., 1986. Effects of residue harvesting on water runoff, soil erosion and nutrient loss. Agric. Ecosyst. Environ. 16, 103–112. Liu, J., Huffman, T., Shang, J., Qian, B., Dong, T., Zhang, Y., 2016. Identifying major crop types in Eastern Canada using a fuzzy decision tree classifier and phenological indicators derived from time series MODIS data. Can. J. Remote Sens. 42, 259–273. Liu, T., Huffman, T., Kulshreshtha, S., McConkey, B., Du, Y., Green, M., Liu, J., Shang, J., Geng, X., 2017. Bioenergy production on marginal land in Canada: potential, economic feasibility, and greenhouse gas emissions impacts. Appl. Energy 205, 477–485. Mabee, W.E., Saddler, J.N., 2010. Bioethanol from lignocellulosics: status and perspectives in Canada. Bioresour. Technol. 101, 4806–4813. Manitoba Department of Agriculture, 1986. Soil Facts: Crop Residue Management. Agdex No. 672. . Mann, L., Tolbert, V., Cushman, J., 2002. Potential environmental effects of corn stover removal with emphasis on soil organic matter and erosion: a review. Agric. Ecosyst. Environ. 89, 149–166. Mitchell, D., 2008. A Note on Rising Food Prices. World Bank, Washington D.C Accessed on July 25, 2017. http://hdl.handle.net/10986/6820. Mueller, S.A., Anderson, J.E., Wallington, T.J., 2011. Impact of biofuel production and other supply and demand factors on food price increases in 2008. Biomass Bioenergy 35, 1623–1632. Muth, D.J., Bryden, K.M., 2013. An integrated model for assessment of sustainable agricultural residue removal limits for bioenergy systems. Environ. Modell. Softw. 39, 50–69.

42