Materials and methods Directory UMM :Data Elmu:jurnal:A:Agriculture, Ecosystems and Environment:Vol81.Issue2.Oct2000:

114 E. Huffman et al. Agriculture, Ecosystems and Environment 81 2000 113–123 practices, has become immediate. The implications of this in terms of data, especially in the complex, interrelated and broad spatial context of ecosystems, is massive. In the absence of complete information about these relationships, the concept of a small number of indicators serving as proxies for a wider variety of parameters has been proposed. In addition, the use of indicators as expressions of information rather than as hard data serves as a means of deliver- ing information to decision-makers and the public in an understandable manner. In 1993, in response to recommendations by the Auditor General of Canada, the Canadian Agri-Food Research Council, the Federal-Provincial Agricul- ture Committee on Environmental Sustainability and the scientific community, Agriculture and Agri-Food Canada AAFC initiated the Agri-Environmental In- dicators AEI project. The expressed goal of AEI was to develop an information base that decision-makers could rely on for environmental conditions and trends in Canadian agriculture. A range of indicators are be- ing developed to assess the degree to which key envi- ronmental issues are being addressed. The objectives are to help identify areas and resources at risk of degra- dation, to help design and target remediation strategies and to facilitate communication. The project is guided by an advisory committee consisting of representatives of producer groups, non-governmental organizations, universities and federal and provincial governments. For the agricultural AEI project, a framework recog- nizing the cycle of cause and effect amongst farm management, resources and social pressures and mod- eled primarily on Australian and OECD ideas, was adopted McRae et al., 1995. Within this framework, selection criteria were established to ensure that each indicator was: 1 relevant to policy issues; 2 scientifically acceptable and defensible; 3 understan- dable to decision-makers; 4 reflective of time and or spatial change; and 5 feasible to obtain or develop. A review of international indicator initiatives and Canadian policy issues, stakeholder consultations and an assessment of data availability and scientific expertise resulted in an initial list of 49 potential agri- cultural indicators. Through an iterative consultation process these were integrated and condensed to six core indicators with various components Table 1. Indicators and their components are either under development tillage erosion, soil compaction, carbon Table 1 Canadian agri-environmental indicators and associated components Indicators Components Risk of soil degradation Water erosion Wind erosion Salinization Tillage erosion Soil compaction Organic carbon change Farm resource management Soil cover and land management Inputs management Risk of water contamination Nitrogen Phosphorus Pesticides Greenhouse gas balance Methane CH 4 Carbon dioxide CO 2 Nitrous oxide N 2 O Biodiversity change Species Habitat Input use efficiency Irrigation Chemicals and energy change, pesticide contamination, are being assessed in pilot areas phosphorus contamination, biodiversity change, input use efficiency or are being applied in all agricultural areas of the country. The remainder of this paper focuses on the soil cover, wind erosion and salinity indicators.

2. Materials and methods

The AEI project is designed to report on the state and trends in the impact of agriculture on the envi- ronment for the entire country at a scale appropriate for national program planning and policymaking. The nature of the project indicated that a biophysical rather than a cadastral or political spatial base was most appropriate. AAFC has recently completed a na- tional spatial stratification built on detailed soils data 1:50–1:100 km generalized to a scale of 1:1 m. This ‘Soil Landscapes of Canada’ SLC product serves as a unified national coverage of soillandformclimate information Shields et al., 1991. In addition, AAFC in cooperation with other federal, provincial and university partners has completed an updated ecolog- ical stratification Ecological Stratification Working E. Huffman et al. Agriculture, Ecosystems and Environment 81 2000 113–123 115 Fig. 1. Example of the ecostratification hierarchy in Canada. Group, 1995 of the country, and coordination of these two initiatives has resulted in a nested hierarchy of biophysically defined land units for Canada’s entire land mass. The smallest unit in the hierarchy is the soil landscape unit SLU, consisting in agricultural areas of an area of approximately 10,000–50,000 ha and encompassing 20–70 farms on average. These landscape units are relatively uniform with respect to parent material, soil development, landform and cli- mate and are nested within ecodistricts, which are in turn nested within ecoregions and ecozones Fig. 1. Each level at successively smaller scales is described by more generalized information. This ecostratification was selected as an appropriate framework for the AEI project as it allows reporting at a variety of scales, depending on data availabil- ity and validity. Also, in the absence of complete data, the biophysical basis of the hierarchy provides a reliable structure for modeling and ‘scaling up’ of site information to a spatial coverage. Although the SLCecostratification framework provided a good na- tional coverage of soil-related parameters suitable for the indicators project, it was also necessary to access a comparable database of land management, crop distribution and climatic conditions. The national ‘Census of Agriculture’ Statistics Canada, 1901–1991, was adopted as the basis for production information. The Census is conducted as a self-administered questionnaire on every farm every 5 years by Statistics Canada 1901–1991, and provides a wealth of information pertaining to agricultural production. The information can be subdivided into four sections: 1 farm structure; 2 crops and land use; 3 livestock and 4 economics. The structure component relates to farm size and ownership char- acteristics, the crops and land use section details the distribution and area of crops, pasture and woodland, the livestock portion relates to the type of animals and size of herd and the economic section covers capital investment levels and the monetary value of inputs and sales. Starting in 1991, questions pertain- ing to land management practices such as the use of conventional, conservation and no-till, summerfallow management and the use of conservation structures such as windbreaks and grassed waterways were also included. This national database has tremendous potential for analytical studies over large areas, but its use is some- what restricted by confidentiality provisions which prohibit the release of data that could be used to iden- tify individual farms. As a result, the database is avail- able only on the basis of enumeration areas EAs or larger census subdivisions or crop reporting districts. The EAs are roughly the size of SLUs and encompass approximately 50 farms each, but are defined on the basis of municipal boundaries and roads and thus have no relationship to natural landscape variation. In order to overcome this spatial discontinuity be- tween the biophysical and cadastral landscape, EA boundary files were input to GIS and an intersection routine termed ‘PARS’ polygon attribute reaggrega- tion system, Ballard and Schut, 1995 was developed to re-configure the Census data to SLUs. The resultant database covers the years 1981 and 1991 a similar dataset for 1996 is under construction and provides up to 90 socioeconomic variables for every agricul- 116 E. Huffman et al. Agriculture, Ecosystems and Environment 81 2000 113–123 tural SLU in the country Huffman and Unrau, 1995. This database of Census variables reconfigured to the SLC base is of significant importance to the project, as most of the indicators rely on it as a source of in- formation about agricultural activities. Climatic data in raw form is available as ‘station records’ from as many as 1800 weather stations across the country. For the AEI project, 1961–1990 monthly normals of selected parameters such as precipitation, temperature, growing-degree days, potential evapo- transpiration PET, dew point, solar radiation, sun hours and vapor pressure were ‘polygonized’ using Thiessen polygon or continuous surface methods and then reconfigured to ecodistricts Bootsma, 1997. For the indicators project, the ecodistrict values were applied to each constituent SLU. 2.1. Soil cover indicator One of the primary factors influencing changes in agricultural soil quality is the amount of cover pro- vided as a result of production activities Coote et al., 1981. Greater amounts of soil cover reduce erosion, increase soil organic carbon levels and provide habi- tat and nutritional benefits to a variety of organisms. Thus, measuring or estimating changes in soil cover over time can serve as an ‘indication’ of movement toward or away from environmental sustainability. An assessment of changes in soil cover must con- sider two factors: the area of crops which provide dif- ferent types and amounts of vegetative and residue cover and the tillage practices that are used to man- age the residue. For example, hay and other perennial sod crops provide almost complete soil cover over the year, whereas production of annual crops leaves soil bare for considerable periods of time. At the same time, however, conservation tillage or no-till maintains considerably more soil cover than conventional tillage using the moldboard plough. The essence of the soil cover indicator is to answer the question: What has been the cumulative effect of changes in crop area and residue management on soil cover over the past 15 years? Is the amount of soil cover increasing or decreasing, and at what rate? In order to provide a first ‘qualitative’ response to that question, agricultural Census data of crop areas was compiled at the provincial and national levels for every 10 years between 1901 and 1991. The 1991 Census also reports for the first time on the adop- tion of various soil conservation practices, including the area under conventional, conservation and no-till Dumanski et al., 1994. With the exception of soil conservation efforts adopted during the years of ‘dust bowl’ conditions in the 1930s and the general use of strip-cropping in the driest Brown soil zone; mainly Kastanozems areas of the prairies, tillage practices specifically oriented to maintenance of soil cover have not been widely used in Canada prior to the mid-1980s. For the purpose of assessing the combined effect of crop and tillage changes on soil cover, it was assumed that all tillage in 1981 and prior was ‘conventional’. For this level of application, a subjective rating of the amount of soil cover provided by each of the country’s major crops under three tillage scenarios was prepared from field data collected for soil erosion studies Shelton et al., 1991 and extended through ex- pert opinion Table 2. No interpretations concerning acceptable limits were attempted. From an analytical point of view, the combined effect of changes in crop areas and tillage practices can be cumulative or counteractive. For example, the adoption of conservation tillage on corn which in- creases cover may be accompanied by an areal shift from corn to soybeans that is sufficient to result in an overall decrease in cover. At the opposite extreme, if a reduction in the area of summerfallow is accom- panied by the adoption of chemical weed control it would show as a dramatic improvement in overall Table 2 Amount of soil cover provided by various crops and tillage prac- tices in Canada CNV a CNS b NT c Corn Zea mays L. Low Medium High Potato Solanum tuberosum L. Low Medium Medium Wheat Triticum spp. Medium High High Barley Hordeum vulgare L. Medium High High Oats Avena sativa L. Medium High High Hay High High High Fruit and berries Medium High High Tobacco Nicotiana tabacum L. Low Medium Medium Canola Brassica spp. Low Medium Medium Soybeans Glycine max L. Merr. Low Medium High Summerfallow Low Medium High a Conventional tillage. b Conservation tillage. c No-till. E. Huffman et al. Agriculture, Ecosystems and Environment 81 2000 113–123 117 soil cover. Although, this portion of the indicator study has not quantified the amount of soil cover more precisely than ‘low, medium and high’, it does take into consideration changes in both crop area and the adoption of conservation practices and the result provides an overview of the cumulative effect. 2.2. Wind erosion indicator Soil erosion by wind in Canada is considered to be a significant problem only in the prairie provinces of Manitoba, Saskatchewan and Alberta. Inherent wind erosion risk i.e. a dimensionless index assuming bare unprotected soil for this area was calculated using an equation developed from the work of Chepil 1945, 1956 and Chepil and Woodruff 1963 and defined as E = KCV 2 − γ W 2 1.5 1 where E is the maximum instantaneous soil movement by wind dimensionless, K the surface roughness and aggregation factor dimensionless, C the soil resis- tance to movement by wind dimensionless, V the drag velocity of the wind cm s − 1 , γ the soil moisture shear resistance dimensionless and W is the surface soil moisture content m 3 water m − 3 soil. For Alberta and Manitoba, data was taken from the extended legends accompanying wind erosion risk Table 3 Examples of estimated spring residue levels by ecoregion, crop sequence and tillage practice Ecoregion typical soil type a Crop sequence Tillage practice Spring residue of initial Mixed Grassland Brown Chernozem, aridic Kastanozem Crop after crop Conventional 50 Conservation 76 No-till 81 Crop after fallow Conventional 11 Conservation 22 No-till 36 Fallow after crop None 90 Parkland Black Chernozem, Chernozem Crop after crop Conventional 45 Conservation 60 No-till 81 Crop after fallow Conventional 6 Conservation 20 No-till 32 Fallow after crop None 90 a Canadian and WRB FAO systems Soil Classification Working Group, 1998. maps prepared in the 1980s Coote and Pettapiece, 1987; Coote et al., 1989, while in Saskatchewan erosion risks were calculated following the same procedure. For the AEI project, actual erosion risk was esti- mated by reducing the inherent erosion risk by a factor based on prevailing land use and management prac- tices. The erosion reduction factor is based on the amount of residue likely to be present and its effective- ness in controlling erosion. Residue at harvest was cal- culated using 10-year average crop yields, multiplied by a crop conversion factor based on crop-specific ra- tios of straw to grain yield. To calculate residue levels for the spring period of the following growing sea- son, initial residues were reduced according to crop- ping system, type and frequency of tillage and an over-winter decomposition factor. Residue reduction under conventional, conservation and zero tillage for the primary cropping sequences in each ecoregion, and the extent of tillage practices, was based on a survey and conservation specialists expert opinion. For ex- ample, land that is to be fallowed the following year, traditionally has no fall tillage, and thus, the residue reduction is due solely to decomposition. On the other hand, land that has been conventionally fallowed typi- cally has a very low amount of residue by fall. Table 3 presents several examples of the estimated proportion of initial residue remaining by spring. 118 E. Huffman et al. Agriculture, Ecosystems and Environment 81 2000 113–123 The residue reduction estimates were used to gen- erate a wind erosion reduction factor for each crop type and tillage practice and an area-weighted value was calculated for each SLU. For 1981, all tillage was assumed to be ‘conventional’, as at that time residue maintenance tillage practices were not widely prac- ticed and there is no data in the Census files. The estimated erosion rate for each polygon was then cal- culated using the equation NEROS = EROS1 − RED 2 where NEROS is the estimated erosion rate, EROS the inherent bare soil erosion rate and RED is the residue reduction factor. 2.3. Soil salinity indicator Soil salinization, or the accumulation of natural salts at or near the soil surface to the extent that it affects vegetative growth, is a land quality issue that affects agricultural sustainability in semi-arid environments. Currently, about 38 of the agricul- tural land in the Canadian prairies is affected to a Table 4 Description of the factors considered in calculating the risk of salinization for soil–landscape polygons Factor Description Present extent Px The extent of present salinity is based on published soil survey data, which includes soil analysis results and estimates of the extent of area affected. Ec e 8 dSm confirms a saline condition, and the extent of area affected, from 0.1 to 100, were assigned risk ratings from 1 to 10 on the basis of expert opinion. Topography Tp The risk of spread of salinity is inversely related to slope steepness, as changes in water table will affect larger areas on level topography. On this basis, soil–landscape polygons described as level slopes of 1–3, undulating 4–9 and hummocky 10 were assigned risk ratings of 5, 3, and 1, respectively. Soil drainage Dr Relationships amongst surface water, surface runoff, depth to water table, piezometric surface, soil drainage, infiltration, hydraulic conductivity, soil moisture storage and water quality are encompassed by the soil drainage class given in conventional soil survey reports. The soil drainage class provided by the SLC database was rated as follows for the SRI: well drained=1, imperfectly drained=8, poorly drained=10, very poorly drained=5 and surface water=3. Aridity Ar Salinity develops where evapotranspiration exceeds precipitation i.e. PE−P0. For the SRI, published maps showing iso-contour lines based on 30-year average aridity values for the growing season were superimposed on SLC maps and a value was determined for each polygon. Values of PE−P i.e.199, 200–249, 250–299, etc. were assigned risk ratings from 1 to 10 by experts. Surface cover Sc Cropping practices which enhance surface evaporation and facilitate deep percolation of precipitation tend to raise the water table. The practice of summerfallowing leaving a field bare over the growing season in order to store water can be considered to represent the highest risk of increasing soil salinity, while permanent cover forage, trees, etc. represents the lowest. A matrix developed by pedologists and conservation officers provides an Sc rating for each SLC polygon based on the distribution of summerfallow and permanent cover Table 5. moderate or severe degree by salt accumulation Eil- ers et al., 1995. Changes in the extent or severity of soil salinity is the subject of one component of AAFC Agri-Environmental Indicator ‘risk of soil degradation’. The goal of this initiative was to de- velop a salinity risk index SRI as a standardized, systematic approach to assessing the risk of change in soil salinity as a result of agricultural activities Eilers and Eilers, 1996. Development of this indicator is based on the as- sumption that landscapes with little or no salinity are unlikely to suddenly develop salinity, whereas land- scapes which have saline problems are the ones most likely to show change. The procedure for developing and applying an SRI consisted of five steps: 1 iden- tify the major components controlling salinity; 2 assign a relative risk rating to each component; 3 combine risk ratings; 4 group the calculated SRIs into classes; and 5 apply the SRI to the Canadian prairie region. Two aspects of soil salinity were considered; the physical conditions under which it develops and the land use activities and climatic conditions which in- fluence changes in salinity. Physical factors taken E. Huffman et al. Agriculture, Ecosystems and Environment 81 2000 113–123 119 Table 5 Matrix depicting the salinization ‘risk factor’ of different combi- nations of land cover types Permanent cover of polygon a Summerfallow of cropland 0–10 10–20 20–30 30–40 40–50 75–100 1 2 3 4 5 50–75 2 3 5 7 8 25–50 3 5 7 8 9 0–25 4 6 8 10 10 a Forest, hay and pasture. into consideration in developing the SRI were the present status of the extent and severity of saline conditions Px, topography Tp and soil drainage Dr. The dynamic factors of aridity Ar and surface cover Sc were incorporated to give: SRI=Px× Tp×Dr×Ar×Sc. A brief outline of each factor is presented in Table 4. The risk of salinization due to the distribution of summerfallow and permanent cover within a polygon is presented in Table 5.

3. Results and discussion