Accounting for ecosystem change in Canada
5.0 Appendices

A. Methodology and data limitations

Human Activity and the Environment 2021: Accounting for ecosystem change in Canada brings together data from many sources to provide accessible information on the state of Canada’s environment. It is a first effort to organize available data according to the new integrated and comprehensive statistical framework for ecosystem accounting described in the System of Environmental-Economic Accounting–Ecosystem Accounting (SEEA – EA)^Note that has been adopted by the United Nations Statistical Commission. While there are multiple components of these accounts, this report focuses on developing data for the core extent, condition and supply and use of ecosystem services accounts. The main data sources, methods and key data limitations are summarized below.

Extent and drivers of change

Terrestrial and freshwater ecosystems

The extent section compiles comprehensive terrestrial and freshwater area estimates that are comparable across the country at the ecoprovince level. Sources include data gathered through satellite imagery, ground plots and photo plots, soil surveys and respondent surveys. In some cases, multiple data sources have been combined to provide a more robust estimate. However, there are difficulties both in consistently delineating ecosystem types across the country and in accurately tracking change in ecosystem areas over time. Data are frequently inconsistent or unavailable across time and space, making time series comparisons difficult. Most land cover and land use maps currently available do not explicitly address changes over time. As a result, estimates of ecosystem extents may change in the future, reflecting not only actual changes on the ground over time, but also changes in data. For similar reasons, caution should be used when interpreting the changes in ecosystem extents over time that are presented in this report.

Table 2.1 (Part 1)

The total area for Canada includes land and water including the Canadian portion of the Great Lakes system. Ecoprovince boundaries are based on the Terrestrial Ecoprovinces of Canada, which exclude the Great Lakes.^Note

Freshwater areas are derived from the Canvec Series – Hydrographic Features at a scale of 1:50,000 using the layers for waterbodies and watercourses.^Note A 5 m buffer was applied to watercourses to estimate the area.^Note Land areas are calculated as the difference between the total ecoprovince area and the water area.

Mapping wetlands and measuring change in extent are difficult on a national and regional scale. Peatlands are organic wetland ecosystems with peat deposits that are at least 40 cm thick, including bogs and fens. Peatland areas can underlie various land covers—these areas overlap with the land and freshwater areas and with other ecosystem types (e.g., forest, tundra, other natural and semi-natural area). Peatland data are taken from the Peatlands of Canada Database (2011), which is based on Soil Landscapes of Canada (SLC) polygons.^Note Area is calculated by multiplying the percent of peatland by the area of the SLC and aggregating by ecoprovince. Map 2.1 displays soil landscape polygons where 50% or more of the area is classed as peatland. The development of updated spatial data on peatland areas is ongoing,^Note but these newer results have not been assessed fully for the purposes of this study.

Table 2.1 (Part 2)

Canada's official National Forest Inventory (NFI) is produced by Natural Resources Canada’s Canadian Forest Service (CFS) on a 10-year cycle and is based on a stratified sample of ground and photo plots to support reporting at the national and ecozone level.^Note The NFI uses the Food and Agriculture Organization of the United Nations (FAO) definition of forest land as "land spanning more than 0.5 hectares with trees higher than 5 metres and a canopy cover of more than 10% or trees able to reach these thresholds in situ," including areas that are temporarily unstocked (areas harvested or burned).^Note The National Deforestation Monitoring System tracks forest area lost to other land uses and, in conjunction with afforestation data, is used to estimate forest area between cycles of the NFI.^Note The extent of boreal and non-boreal forests was estimated with boundaries from Brandt (2009).^Note

Statistics Canada has produced an estimate of forest extent at the ecoprovince level for the purposes of ecosystem accounting. Estimates provided here were downscaled by benchmarking treed area data derived from medium resolution satellite imagery to the latest NFI estimates^Note This method increases uncertainty and error in the ecoprovince-level estimate. In addition, the use of treed area data excludes temporarily unstocked areas, which can be a significant component of forests in some regions, with estimates of about 10% of forest area temporarily unstocked a result of fire or harvest.^Note For these reasons, these estimates are not intended as official estimates of forest land, which are reported in the NFI and in The State of Canada’s Forests.

There is considerable difficulty in accurately classifying grassland areas using satellite imagery data because of the similarities between natural grasslands and tame or seeded pasture, both of which can be used for grazing livestock.^Note Assessed data sets showed considerable interannual fluctuations between these classes and between pasture and cropland classes (e.g., hay land, from which a hay crop may be harvested.) In order to minimize fluctuations and provide a more robust estimate, the grassland data are estimated using multiple data sets, with satellite imagery data averaged over several years. As a result, change over time is difficult to assess.

First, Agriculture and Agri-Food Canada’s (AAFC) space-based 30 m resolution Annual Crop Inventory^Note classes for grassland and pasture were averaged for 2014, 2015 and 2016. For areas located outside Canada’s agricultural ecumene, grassland was estimated with averaged 2010 and 2015 data from the temperate or subpolar grassland class from the Commission for Environmental Cooperation’s (CEC) Land Cover of North America at 30 m.^Note In mountainous ecoprovinces, where there was a high degree of variability in grassland results between the datasets, results were further adjusted using derived coefficients. Treed pasture was removed by overlaying the 2015 Land Cover of North America temperate or sub-polar needleleaf forest, temperate or sub-polar shrubland, temperate or sub-polar broadleaf deciduous forest and mixed forest classes over the area and results were aggregated by ecoprovince. Tame or seeded pasture areas from the Interpolated Census of Agriculture^Note were subtracted to avoid double counting, as these are reported as arable land in the intensive use area in Table 2.1 (Part 3). Final adjustments were made to ensure consistency with other classes.

Areas of arctic tundra were estimated using the tundra areas from Baldwin, et al. (2018)^Note and subtracting areas of permanent snow and ice from the 2015 Land Cover of North America snow and ice class. Note that this estimate of area north of the boreal zone includes freshwater areas and barrenland.

The other natural and semi-natural area category is calculated by subtracting all other ecosystem types from the total area. It may include, for example, woodland, shrubland, barrenland, wetland and lakes and rivers.

Table 2.1 (Part 3)

Arable land is reported as the sum of cropland, tame or seeded pasture and summerfallow from the Interpolated Census of Agriculture, which aggregates Census of Agriculture data by ecological and drainage units. Other land on farms (e.g., idle land, land occupied by farm buildings, wetlands and woodlands) and natural pasture are not included. Data for arable land do not indicate the amount of land that is potentially cultivable. There are differences between tabulations of Statistics Canada’s Census of Agriculture data by standard geographies and the interpolated data provided to AAFC. Specifically, confidentiality procedures are applied to the data in order to avoid the possibility of identifying any specific agricultural operation. This involves the suppression of selected data. As well, the interpolated data are based on the Census Geographic Component Database, in which splits of selected key farms have been reallocated to specific geolocations, rather than to the location of the farm headquarters.

Settlements and human infrastructure are represented with land cover data on built-up and artificial surfaces (BUAS), defined as areas that are predominantly built-up or developed including road surfaces, railway surfaces, buildings and paved surfaces, urban areas, industrial sites, mine structures and golf courses, as well the vegetated areas associated with these land covers. These estimates are based on the settlement (built-up and urban) and roads classes from AAFC’s Land Use, 2000 and 2010^Note at 30 m resolution for areas of Canada south of 60° N. For areas north of 60° N, where no AAFC land use data was available, estimates were generated for 2000 based on road lengths from Statistics Canada’s Road Network File.^Note For 2010, the Land Cover of North America urban and built-up class was used to estimate the BUAS for northern areas.

The 2015 national BUAS extent was estimated by adding the area converted to settlement from Environment and Climate Change Canada’s National Inventory Report (NIR) for the years from 2011 to 2015 to the 2010 national BUAS estimate. The NIR reports supplementary data in the common reporting format (CRF) tables, with Table 4.1 providing data on the area of land converted from forest, cropland and grassland to settlements. Settlements as defined in the NIR include all roads and transportation infrastructure; rights-of-way for power transmission and pipeline corridors; residential, recreational, commercial and industrial lands in urban and rural settings; and land used for resource extraction other than forestry. Land converted from wetlands and ‘other land’ are not included in the conversion. A detailed methodology of data development is available in Part 2 of the NIR.

A quality assessment of the change in BUAS areas in a buffer area surrounding 20 population centres was conducted for the 2000 to 2010 time period. These population centres were selected in different regions of the country. It found that the change accuracy was greater than 90% for large polygons greater than 10 ha, which contributed 90% of the change, and above 80% for medium sized polygons between 5 to 10 ha, which contributed 7% of change. Accuracy was lower for smaller polygons, but these polygons contributed less than 5% of the change. Additional but more general assessments were also done using satellite imagery.

Marine and coastal ecosystems

As a first step a hexagon grid was built for the entire area of Canada’s exclusive economic zone. These 1 km² hexagons are used as the basic spatial unit for all of the marine data in this report.

Data for bathymetry, slope and the terrain ruggedness index were calculated from the General Bathymetric Chart of the Oceans (GEBCO) data (version 20150318). Depth classes were assigned by averaging bathymetry data over each hexagon.

Slope was calculated using the tool in ArcGIS, resampled at 25 m, and then the mean for each hexagon was calculated using zonal statistics.

Terrain ruggedness was calculated as a raster layer using the method described in Riley et al. (1999).^Note It was then resampled at 25 m and the mean for each hexagon of the grid was calculated using zonal statistics.

Data for seagrass meadows, kelp forest and cold water coral were taken from a mix of polygon and point sources. For seagrass meadows, both polygon and point data were used. Polygon data that overlapped were treated as a single patch. Where point data did not overlap existing polygon data, the patch size was assumed to be equivalent to the ecoregion’s average seagrass patch size. These derived polygons were then merged with the initial polygon data.

A similar methodology was used for cold water coral; however, the applied mean patch size was calculated using only coral areas less than 100 km², since the inclusion of larger patches would have skewed the mean and overestimated coral areas. Many of these larger patches were estimated using kernel density methods and may also overestimate coral area.

For kelp forests, only polygon data was used, as point data on the East Coast was scarce and clearly underestimated kelp area.

Salt marsh estimates were based on United Nations Environment Programme (UNEP)^Note polygon data. Although there are other known areas of salt marsh in Canada, it was not possible to obtain area estimates for this analysis.

Seagrass, kelp and salt marsh ecosystems all occur in coastal areas, including in a number of small bays and inlets on the borders between land and sea—they were included in the extent even if in areas that were assigned to land. Furthermore there were some areas of overlap between ecosystem types. These areas have been assigned to both ecosystems rather than creating joint ecosystem types. As such, totals in this table will not match totals in other marine tables in the publication. As detailed marine data is costly to acquire, there are many gaps in the extent accounts, some of which could be filled by modelling exercises or potentially satellite data. Accounting for change over time of marine ecosystems is likely to prove particularly difficult.

Climate

Average annual and seasonal air temperature changes from 1948 to 2016 are produced by Environment and Climate Change Canada (ECCC) based on gridded temperature data interpolated from weather stations.^Note The long-term temperature trend (1948 to 2016) refers to the linear trend of temperature departures from the 1961 to 1990 climate normal. The data are tabulated by ecoprovince and for different ecosystem types and areas by ecoprovince. Caution should be exercised when analyzing change results in the North because of lower climate station densities. Significance levels are not available.

To calculate the temperature change occurring in forest, freshwater, peatland, agricultural and built-up and artificial surface areas (Table 2.9), the temperature change data were overlayed on each class layer and an average was generated by ecoprovince. Each class is treated independently from the others and overlap exists between classes. Some classes differ from the areas reported in Table 2.1. Specifically, the forest area layer uses treed area from Beaudoin (2017), while agricultural area is based on the total farm area (TFAREA) variable by soil landscape polygon from the Interpolated Census of Agriculture (2011).^Note This variable represents all areas operated by farms including cropland, summerfallow, tame or seeded pasture, improved pasture and other areas on farms (woodlands, wetlands, idle land and farm buildings including barns, greenhouses, mushroom houses and dwellings).

Average annual, maximum and minimum monthly temperature, precipitation, evapotranspiration and potential evapotranspiration, as well as average annual and seasonal change in precipitation, evapotranspiration and potential evapotranspiration from 1979 to 2016 are based on the Ecological Assimilation of Land and Climate Observations (EALCO) model used by the Canada Centre for Mapping and Earth Observation, Natural Resources Canada.^Note

These variables were estimated from models that used a combination of climate and satellite data. Caution should be used when interpreting these results and in particular trend results. There can be higher levels of uncertainty in some areas because of a scarcity of data, for example, in northern and mountainous regions.^Note These variables are useful indicators for identifying where ecosystem changes might be occurring or may have occurred. For other types of research, such as water budget or climate change analysis, broader considerations and more validation are recommended. The time series data from the EALCO model were tested for the presence of serial correlation and for anomalous observations (outliers). The ARIMA function in the R statistical program was used to compute the overall trend. The ARIMA function produces a linear trend and the associated significance level is adjusted for any existing serial correlation and anomalous observations. Statistically significant linear trends at the 90% confidence level or above are indicated.

The average climate variables are defined as the mean of the reference period. The months used to calculate each season are as follows: spring (March, April, May); summer (June, July, August); fall (September, October, November); and winter (December, January, February).

Condition characteristics

Total water storage change and water yield

Total water storage change is a coarse resolution estimate of changes in the amount of water stored in the environment above and below the Earth’s surface including groundwater, soil moisture, snow, ice and surface water. Change in total water storage is indicative of changing climate conditions and is useful for understanding potential influences on ecosystems. Results should be interpreted with caution, in context with supporting data.

Total water storage change data are based on Wang and Li (2016)^Note who used monthly data from the Gravity Recovery and Climate Experiment (GRACE) satellite mission to assess terrestrial water storage climatology for 2002 to 2016. This data has been resampled from the GRACE data to a resolution of 5 km using the EALCO model.^Note

Time series results were generated monthly at the ecoprovince level. Data were suppressed for smaller ecoprovinces using a threshold of approximately 90,000 km² based on recommendations from Wang et al. (2014).^Note The time series data were tested for the presence of serial correlation and for anomalous observations (outliers). The ARIMA function in the R statistical program was used to compute the overall trend. The ARIMA function produces a linear trend and the associated significance level is adjusted for any existing serial correlation and anomalous observations. Statistically significant linear trends at the 90% confidence level or above are indicated.

Water yield is an estimate of freshwater runoff, derived from data on the unregulated flow of water in rivers and streams in Canada. Data were suppressed at the ecoprovince level for the North, but are included in the Canada-level estimate. The national average annual water yield is area-weighted based on ecoprovince areas. The methodology for water yield is described in Statistics Canada, 2017, “Freshwater in Canada,” Human Activity and the Environment, Catalogue no. 16-201-X and Statistics Canada, 2009, “The Water Yield for Canada as a Thirty-year Average (1971 to 2000): Concepts, Methodology and Initial Results,” Environment Accounts and Statistics Analytical and Technical Paper Series, Catalogue no. 16-001-M, no. 7.

Forest condition

The National Forestry Database is Canada’s main source of provincial and territorial data on forest management and impacts on forest resources.^Note These data differ from the spatial data on timber harvesting and burned areas mapped and reported by ecoprovince in this report in Table 3.3 and Map 3.3. Fire data were taken from the National Burned Area Composite (NBAC), which calculates the area of forest burned on a national scale for each year since 1986. The NBAC is part of the Fire Monitoring, Accounting and Reporting System and is based on the integration of data from fine and coarse resolution satellite data from Natural Resources Canada and Provincial, Territorial and Parks Canada Agencies. Time series harvest data (1985 to 2015) were identified by 30 m Landsat remote sensing, as part of the Canada Landsat Disturbance (CanLaD), by Guindon, et al. (2017).^Note

Urban greenness

This analysis provides a synoptic view of urban greenness in summer for three reference years over an 18-year period as a measure of urban condition. It used the normalized difference vegetation index (NDVI) generated from moderate resolution imaging spectroradiometer (MODIS) at a spatial resolution of 230 m to estimate urban greenness for 996 of 1,010 population centres (i.e., those located south of 60° latitude).

The analysis used scaled NDVI, where values from 0 to 1 correspond to a vegetation gradient of non-vegetated (0) to highly vegetated (1). Population centre pixels were classified as either ‘green’ or ‘grey.’ The urban green class corresponds to areas with an NDVI greater than or equal to 0.5, representing urban areas with more vegetation. Areas with lower values are considered grey and are largely non-vegetated, though patches of grass, shrubs or crops, or other unhealthy/poor condition vegetation will be included (Figure A.1). Water areas were excluded from the analysis.

Greenness was assessed for the reference years 2001, 2011 and 2019 for the same physical area using the 2016 population centre boundary to ensure consistency. The mean NDVI was calculated using mean weekly best-quality maximum-NDVI corrected for cloud and other atmospheric residual contaminants^Note for nine weeks from June 25 to August 26 for each year.

This assessment of greenness has several limitations associated with the use of NDVI to represent greenness, including the coarse resolution of the MODIS data and the selection of the 0.5 NDVI threshold to classify green or grey pixels. As well, no distinction was made between greenness resulting from publicly accessible or private green spaces. For trend analysis, assessment of additional time series data are required, while higher resolution data are needed for the identification of detailed urban green spaces. A next step for this work will be the assessment of green space extent and greenness condition using more spatially-detailed datasets and additional time periods.

For more information, see Grenier, M., et al., 2021, “Urban greenness, 2001, 2011 and 2019,” EnviroStats, Statistics Canada Catalogue no. 16-002-X.

Figure A.1
Examples of urban pixels classed as green or grey

Description for Figure A.1

The purpose of this image is to visually display the levels of greenness between varying land covers. The green or grey class is based on the MODIS NDVI value.

The MODIS pixels are represented by a box with a white outline and are overlaid on high resolution imagery provided by Google to visualize what is present on the ground. The image compares twelve MODIS pixels organized into two horizontal rows. The top row displays decreasing greenness levels in green pixels and the bottom row displays decreasing greenness levels in grey pixels. In between the two rows there is an arrow spanning across the page from left to right, shaded in a gradient of green to white to illustrate decreasing greenness.

The first six MODIS pixel images are classed as green because their MODIS NDVI value is greater than or equal to 0.5, with the greenest pixel on the left and the least green pixel on the right. The last six MODIS pixel images are classed as grey because their MODIS NDVI value is less than 0.5, with the least grey pixel on the left and the greyest pixel on the right. From left to the right, the first green pixel represents mostly treed area with a few buildings present and a stream and road passing through it, the second represents a golf course with mostly grass, the third represents an agricultural area with part of the pixel overlaid on residential area, the fourth represents a city park with a baseball diamond, grass, and a parking lot, the fifth represents a residential area with large lots and trees, and the sixth represents a residential area with smaller lots and some trees. From left to right, the first grey pixel represents a residential area with few trees and grass and mostly artificial surfaces, the second grey pixel represents a residential area with no trees, some grass and mostly artificial surfaces, the third pixel represents an area with some large buildings, mostly paved surfaces and some grass, the fourth pixel represents an area with mostly large buildings, paved surfaces and some grass, the fifth pixel represents a new residential development with individual houses, paved surfaces, bare ground and no vegetation, and the sixth pixel represents an entirely paved surface.

Landscape fragmentation, human landscape modification index (HLMI) and human freshwater influences index (HFII)

This report measures ecosystem degradation and human impacts on landscapes in several ways including estimating directly and indirectly modified areas, linear feature density, natural and semi-natural patch size, distance to natural and semi-natural patch, the human landscape modification index (HLMI) and the human freshwater influences index (HFII).

Directly modified land (circa 2016) includes areas used for agriculture (e.g., cropland, pasture and summerfallow) from the Interpolated Census of Agriculture, recent forest harvest (1986 to 2015) from CanLaD and built-up and artificial surfaces using the datasets described above in the extent section. Indirectly modified areas include all other terrestrial and freshwater extent.

Data on natural and semi-natural patch size and distance to patch (circa 2016) are calculated based on spatial data files from AAFC’s Land Use, 2015 beta,^Note Statistics Canada’s Road Network File and from CanLaD. Natural and semi-natural patches include all land classes except settlements (built-up and urban), roads, cropland (annual and perennial), harvested forest (from 2001 to 2015) and managed grassland (natural grass and shrubs used for cattle grazing). A single patch of natural and semi-natural land has a minimum size of 9 pixels (at 30 m x 30 m) with an area of 8,100 m². Distance to patch is the average distance from any location in the ecoprovince to the nearest patch of natural and semi-natural land. Average natural and semi-natural patch size should be interpreted with distance to natural and semi-natural patch to get a more complete understanding of fragmentation. Note that the presence of islands can lower the average natural and semi-natural patch size.

Linear feature density (circa 2016) is calculated based on the length of linear features including roads from Statistics Canada’s Road Network File, and rail lines, cutlines and electrical transmission lines from Natural Resources Canada’s Topographic Data of Canada.^Note Linear feature density excludes other types of infrastructure, such as pipelines, and is represented in metres per square kilometre of the total ecoprovince area.

The HLMI is a composite index of the above variables calculated at the pixel level by Statistics Canada. It represents human modifications circa 2011 using the datasets described above with the exception of forest harvest data, which is for the period from 2001 to 2011. It aggregates three measures of human modifications and provides a score to indicate how much the land area has been modified from its natural state. This tool enables comparisons of the level of human modification or use. The index is based on three principles: the degree to which an area has been modified—from a natural or semi-natural state to the most modified state of built-up and artificial surfaces; the distance of an area to the nearest patch of natural and semi-natural land and the size of that patch; and the distance of an area to the nearest linear feature and the density of those linear features. The formula for calculating the HLMI is below.

• Linear feature (LF) fragmentation index (LFFI) = (LF Density * 0.5) + (LF Distance *0.5)
• Natural patch fragmentation index (NPFI) = (Size of closest natural and semi-natural patch *0.5) + (Distance to nearest natural and semi-natural patch *0.5)
• Fragmentation index (FI) = (LFFI *0.5) + (NPFI *0.5)
• Green index (GI) = Natural and semi-natural pixels (*1) + Forest harvest and managed grassland areas (*2) + Cropland (*3) + Urban and artificial surfaces (*4)
• HLMI = (GI *0.5) + (FI *0.5)
• Pixel values are re-scaled from 0 to 100.

Areas with lower scores are generally more intact and therefore potentially able to supply ecosystem services such as water filtration, climate regulation, habitat maintenance and pollination that would be more in line with their natural condition. Areas with higher scores represent progressively more altered or intensively-used ecosystems. Canada totals are an area-weighted average based on ecoprovince areas. Other indicators similar to the HLMI exist at a global scale, for example, the Human Footprint, the Global Human Modification of Terrestrial Systems and the Forest Landscape Integrity Index. These indicators differ in the variables used and methodology applied, but all represent the influence of human activities on terrestrial areas.^Note

The HFII presents an aggregated ranking by drainage region of the individual rankings of 13 variables and indicators. These variables were selected because they are associated with various anthropogenic influences on freshwater ecosystems. They include climate change, population density, the HLMI, water crossing density (e.g., bridges and culverts), dams, freshwater intake and nutrient emissions from industrial plants, farms and wastewater treatment plants (Table A.1). Higher ranked drainage regions (i.e., 1^st, 2^nd, 3^rd, etc.) are subjected to a higher number of direct and indirect human influences on their freshwater ecosystems.

Water crossing density is calculated using Statistics Canada’s Road Network File^Note and represents the number of bridges or culverts crossing a water body or water course per square kilometre. Dam density results were compiled using Natural Resource Canada’s CanVec dam data integrated with data from an inventory of large dams produced by the Canadian Dam Association.^Note Where dams were coincident (within 1 km) between the two datasets they were counted once. Caution should be used when analyzing dam data as there are differences in coverage across provincial jurisdictions.

Marine and coastal condition

Sea surface temperature and salinity data were taken from the World Ocean Atlas data selector,^Note projected to the Canadian Albers equal area projection and assigned to the underlying hexagonal grid using zonal statistics in ArcGIS Pro. For marine ecoregion and depth class averages, the data were averaged across the geography. As there are relatively few partial hexagons in the grid, weighting the averages by area made an insignificant difference. Stock sustainability data were obtained from the Fisheries and Oceans website^Note and were assigned to species group and regions using the stocks’ fishing area available in the downloadable file for 2019.

Sea ice extent data were estimated using annual shapefiles from the National Snow and Ice Data Center^Note for the months of September, February and March, projected to the Canadian Albers equal area projection. These were then intersected with the marine ecoregion geography to estimate extent by marine ecoregion. For the Atlantic ecoregions the maximum value of February or March extent was taken for each year. These maximum extents were then averaged to produce decade averages for the Atlantic ecoregions while the September minimum values were averaged by decade for Arctic ecoregions.

Data on the area of aquaculture sites were taken or estimated from numerous sources, as referenced in Table 3.11 and Map 3.11. The area of sites was calculated directly from the polygon files used for British Columbia fish sites, and for sites in Prince Edward Island and Nova Scotia.^Note

Point data were found for Newfoundland aquaculture along with an estimate of total aquaculture area.^Note The area of sites for different regions around the island was estimated by dividing the total area by the number of sites.

For New Brunswick, coordinates and area of aquaculture sites were available and were used to create the map and table.

British Columbia shellfish aquaculture sites were mapped using a PDF map and a list of licenses by fisheries region. The area was estimated using an average area per farm of 8.3 hectares, taken from DFO (2017).^Note

The total area and location of sites for aquaculture in Quebec were taken from the ministère de l’Agriculture, des Pêcheries et de l’Alimentation (MAPAQ).^Note

Data for oil licenses were mapped using shapefiles from the Canada–Newfoundland and Labrador Offshore Petroleum Board, Canada–Nova Scotia Offshore Petroleum Board and Crown–Indigenous Relations and Northern Affairs Canada.^Note

Ecosystem services supply and use

Provisioning services

The System of Environmental Economic Accounting – Ecosystem Accounts (SEEA – EA) includes several categories of biomass provisioning services. The intent is to recognize the ecosystem contribution of provisioning services, though where this contribution is difficult to distinguish, gross biomass harvested is recognized as an adequate proxy measure. Each service is defined such that there is no double-counting of the ecosystem contribution of individual services. For example, the production of cultivated livestock is not included as a provisioning service where the biomass provision of fodder crops and grazed biomass are counted. Similarly, aquaculture production that relies on wild-caught fish or harvested crops as feed is excluded, since inclusion would result in double-counting, while production that requires no feed inputs (e.g., oysters, mussels) can be included. Note that the data reported in Table 4.1 includes different moisture contents.

Agricultural production is reported as a proxy of the provisioning service provided by agricultural ecosystems. Estimates of crop, honey and maple production include the majority of grain, oilseed, pulse, corn for silage, tame hay (alfalfa, other tame hay and forage seed), potato, vegetable, fruit, honey and maple production (as syrup). Estimates of fodder corn production are calculated using a standard percentage moisture content of 70%. Estimates of production of hay are based on a standard dry matter content of 90%. Estimates for fruit are for marketed production. The estimates may include some data assessed at data quality standard E (use with caution). Provincial data that are unavailable or data that are suppressed to meet confidentiality or data quality standards are not included in the total. The estimates exclude greenhouse vegetable, mushroom, tobacco, cannabis, nursery, sod, or Christmas tree production, as well as grazing on crop residue.

Forage production estimates for tame or seeded pasture and natural land for pasture (rangeland) are based on the areas reported on the Census of Agriculture multiplied by estimates of the average provincial animal unit month (AUM) taken from Yungblut (2012).^Note Biomass provisioning estimates exclude the production of meat, dairy, egg, wool and fur-bearing animals.

The volume of harvested timber reported in the National Forestry Database was converted to tonnage weight by adjusting for wood density.^Note The green weight with bark was estimated using forest product conversion factors for the United States for conifer and non-conifer saw/veneer and pulpwood/fuelwood logs. Non-commercial harvests (e.g., for residential firewood) are excluded.

Fisheries landings are defined as the part of the commercial catch that is put ashore. Seafisheries include groundfish, pelagic and other finfish and shellfish. The data may include some farmed shellfish production (e.g., Atlantic oysters). Freshwater landings data exclude Newfoundland and Labrador, Prince Edward Island, Nova Scotia, British Columbia and Yukon Territory.

Total aquaculture production of shellfish is reported and includes some wild production. It excludes aquaculture production of finfish, restocking of lakes and freshwater fisheries. Data are collected from each of the provincial departments responsible for aquaculture. Provinces and territories with data not available are not included in the total.

Economic data including GDP and employment by sector are available following the North American Industry Classification System (NAICS). An effort was made to align economic statistics on the sectors benefiting from ecosystem provisioning services (i.e., agriculture, forestry, fishing, hunting and trapping etc.); however, in some cases data were aggregated at a higher level. The report grouped the sectors according to previously defined industry groupings. For this reason, there are some differences in the treatment of sector-level aggregation. For example, GDP and employment data for the forest sector includes manufacturing activities, while the primary agriculture sector excludes manufacturing and aquaculture was omitted. GDP and employment data are available for fishing, hunting and trapping industries, to which aquaculture and fish processing were added. However, data on the contribution of the fishing and seafood sector to census subdivision (CSD) employment income was based on an aggregation of data from the 2016 Census, and includes only fishing, aquaculture (finfish and shellfish) and fish processing.

Regulating services

According to the SEEA – EA, ecosystem contributions to global climate regulation services include measurement of carbon sequestration and retention of carbon in ecosystems.^Note Under this standard, carbon sequestration reflects the ability of ecosystems to remove carbon from the atmosphere and store it for long periods of time. Net ecosystem carbon balance is considered an appropriate metric. Where net carbon sequestration is zero or negative, the service supplied by the ecosystem is zero. Carbon retention supplies a service through avoided carbon emissions and includes carbon in above ground and below ground biomass (including in the seabed) and soil organic carbon (including peatlands to a maximum of 2 m depth). It excludes inorganic carbon in freshwater, marine and subterranean ecosystems, fossil fuel deposits, as well as harvested wood products (carbon stored in produced assets) and stocks of crops or livestock (short-term storage). Measurement of carbon retention is a focus particularly for ecosystems where the stock of carbon is at risk of emission, e.g., as a result of fire, deforestation or peatland draining.

This report makes use of data related to carbon sequestration and retention that are produced by other departments to meet their existing reporting requirements; ^Note however, there are some differences in focus and gaps exist, e.g., for marine and coastal ecosystems and unmanaged forests. For example, carbon stock change and fluxes of CO₂ to the atmosphere are the subject of reporting requirements for Land use, Land-Use Change and Forestry as part of the National Inventory Report (NIR) to the United Nations Framework Convention on Climate Change (UNFCCC). Carbon stock changes and fluxes are also modeled by NRCan CFS to meet other reporting frameworks such as the Montréal Process. These data are based upon models and are subject to limitations and uncertainty, as detailed in the original documentation.

Carbon sequestration service supply for arable and urban ecosystems is based on data reported in the common reporting format (CRF) tables as part of the NIR 2020. Values reported in CO₂ equivalents have been converted to carbon. Sequestration attributed to built-up and artificial surfaces is based on the net carbon stock change in living biomass for settlements remaining settlements as reported in CRF Table 4E. This estimate of carbon removals for urban ecosystems accounts for net carbon uptake by urban trees, in which biomass decay is implicit. It includes 69 population centres in Canada with a population over 30,000 (of 947 population centres on the 2011 Census), capturing major Canadian cities representing 62% of 1990 urban area and 79% of 1990 population.^Note It does not include emissions related to urban expansion. Sequestration attributed to arable land (cropland, summerfallow and improved pasture) is based on reporting of net removals for cropland remaining cropland as reported in the CRF Table 4B. This estimate is based on a gains-loss approach to carbon stock estimation using land management changes reported on the Census of Agriculture. Emissions associated with land conversions to cropland were not included in Table 4.4.

Carbon ecosystem indicators are modeled by the CFS for managed forests using the Carbon Budget Model of the Canadian Forest Sector (CBM-CFS3) and national compilation of forest inventories, forest growth and yield tables, large-scale disturbances and forest management activity data, as part of the National Forest Carbon Monitoring, Accounting and Reporting System (NFCMARS) supporting reporting in the NIR 2020.^Note The definition of forests is based on the definition in the Marrakesh Accords (i.e., 25% crown closure, with potential tree height at maturity of 5 m or greater and covering an area of 1 ha or greater). Forests are classified as managed or unmanaged based on the occurrence of management activities for timber or non-timber and on the level of protection against disturbances.^Note Note that this definition of forest area differs from that used for the National Forest Inventory, which is based on the FAO definition (i.e., 10% canopy cover, 5 m height, 0.5 ha).^Note

Table 4.3 reports carbon stocks by Intergovernmental Panel on Climate Change (IPCC) carbon pool. Indicators reported in Table 4.4 include net primary productivity (NPP), net ecosystem productivity (includes heterotrophic respiration) and net biome productivity (NBP), which includes harvesting transfers and disturbance emissions, for managed forest land. Negative NBP represents a loss of carbon and therefore is not a carbon sequestration service. Note however that harvesting transfers include wood products that can store carbon in the built environment over the long term, in addition to wood products that are more quickly oxidized. The species-weighted density of commodities used to estimate carbon in harvested wood products (oven dry tonne wood material/m³)^Note differs from the densities used for estimating the green weight of commercial timber harvests as a biomass provisioning service in Table 4.1. In addition, these two estimates use different data sources for fuelwood and firewood.

Kelp forests, seagrass meadows and salt marshes in coastal area have an important role in carbon cycling. Studies on Canadian seagrass meadows have currently produced a wide range of estimates for carbon accumulation in sediment at different sites and mapping of these ecosystems is ongoing. Section 2.1 provides an estimated minimum known extent.

B. Glossary

Abyssalpelagic: refers to waters at depths of greater than 4 km depth. The abyssalpelagic class includes areas where the maximum water depths exceeds 4 km.

Agriculture sector (primary): includes the following North American Industry Classification System (NAICS) codes: 111 – crop production, 112A – animal production (excluding aquaculture), 1151 – support activities for crop production and 1152 – support activities for animal production. Excludes agricultural input suppliers, food product manufacturing, wholesale and retail, and food services.

Ambient air concentrations: amount of air pollutants in outdoor air expressed in parts per billion (ppb) by volume for gases and micrograms per cubic metre (µg/m³) for particulate matter. Average indicators are used to capture prolonged or repeated exposures over long periods or chronic exposure, while peak indicators are used to capture immediate or acute short-term exposures. The average and peak air quality definitions vary by pollutant. For more information, see Air quality

Ambient air quality: refers to outdoor air quality. The Canadian Ambient Air Quality Standards set limits, based on human health considerations for outdoor air concentrations of fine particulate matter (PM_2.5), ozone (0₃), sulphur dioxide (SO₂) and nitrogen dioxide (NO₂).^Note

Arable land: reported as the sum of cropland, tame or seeded pasture and summerfallow from the Interpolated Census of Agriculture. Other land on farms (e.g., idle land, land occupied by farm buildings, wetlands and woodlands) and natural pasture are not included.

Bathypelagic: refers to waters at depths of 1 km to 4 km. The bathypelagic class includes areas where maximum water depths are within this range.

Built-up and artificial surfaces: include areas that are predominantly built-up or developed including road surfaces, railway surfaces, buildings and paved surfaces, urban areas, industrial sites, mine structures and golf courses, as well the vegetated areas associated with these land covers, at a resolution of 30 m or greater.

Canadian Ambient Air Quality Standards (CAAQS): the CAAQS were developed by the Canadian Council of Ministers of the Environment, under the Canadian Environmental Protection Act, as outdoor air quality targets to drive air quality management across Canada. Standards exist for fine particulate matter (PM_2.5), ozone (0₃), sulphur dioxide (SO₂) and nitrogen dioxide (NO₂) with an averaging time (e.g., 24-hour, annual, 8-hour, 1-hour), numerical value and statistical form. For more information, see Air Quality Report.

Carbon sequestration: the removal and long-term storage of carbon from the atmosphere. The carbon sequestration service focuses on the amount of biocarbon accumulated in ecosystems net of respiration, decomposition and combustion in an accounting period.

Carbon retention: carbon is stored in ecosystems in living and dead biomass and soil organic carbon. The carbon retention service relates to the ability of ecosystems to retain carbon, thus avoiding emissions of carbon to the atmosphere. It is quantified as the stock of carbon retained (stored) in ecosystems including forests, wetlands (e.g., peatlands) and agricultural, coastal and marine areas at the beginning of the accounting period.

Climate normal: a three-decade average of a climatological variable such as temperature or precipitation.

Cropland: area used for growing crops. Total area of cropland from the Interpolated Census of Agriculture includes reported areas of hay and field crops, vegetables, nursery products, fruit, berries, grapes and nuts. Total area of cropland in the AAFC Land Use spatial data product includes annual and perennial crops.

Cutlines: a line cut through an area to facilitate cadastral or seismic surveys or create fire breaks. E.g., seismic cutlines are narrow corridors up to about 10 m wide cut through the landscape including forest, peatland and tundra. Often used to transport survey equipment for natural resource exploration.

Distance to natural and semi-natural patch: the average distance from any location in the ecoprovince to the nearest patch of natural and semi-natural land.

Degree of modification: the degree to which a landscape has been modified from a natural state. Heavily built areas with a large proportion of artificial surfaces and lower proportion of natural and semi-natural areas can be considered the most modified or the least intact ecosystems.

Drainage region: grouping of sub-sub drainage areas as defined by the Standard Drainage Area Classification, Statistics Canada’s official classification of drainage areas. For more information, see Standard Drainage Area Classification (SDAC) 2003.

Ecoprovince: the second level (under ecozones) of the Ecological Land Classification, a hierarchical classification of ecological areas in Canada. For more information, see Ecological Land Classification.

Ecosystem: defined in Article 2 of the Convention of Biological Diversity as a dynamic complex of plant, animal and micro-organism communities and their non-living environment interacting as a functional unit.^Note

Ecosystem condition: the quality of an ecosystem measured in terms of its abiotic, biotic and landscape and seascape-level characteristics across a range of temporal and spatial scales.^Note

Ecosystem functions: the physical, chemical and biological processes (e.g., nutrient cycling, carbon cycling, etc.) that occur in ecosystems.

Ecosystem services: a wide range of services that flow from ecosystems and provide benefits to people, often grouped into three categories: provisioning services (e.g., supply of food, fuel, fibre and water), regulating services (e.g., filtration, purification, regulation and maintenance of air, water, soil, habitat and climate) and cultural services (e.g., nature-based recreation or education).

Ecozone: the first level of the Ecological Land Classification, a hierarchical classification of ecological areas in Canada. For more information, see Ecological Land Classification.

Epipelagic: refers to waters from the water surface to 200 m depth. The epipelagic class includes areas where the maximum water depth reaches 200 m. This region is further divided, with the coastal epipelagic class having a maximum depth of 50 m.

Evapotranspiration (ET): the process of evaporation from land surfaces and transpiration from plants. It is controlled by surface water availability and by meteorological variables such as net solar radiation, air temperature, humidity and wind speed.

Exclusive Economic Zone (EEZ): refers to the area of the sea in which a country has rights to the exploration and use of marine resources. The area covers from the coast to 200 nautical miles offshore.

Fishing and seafood sector: includes the following North American Industry Classification System (NAICS) codes: 1141 – fishing; 1125 – aquaculture and 3117 seafood product preparation and packaging.

Forest: ecosystems dominated by trees, including temperate forest and boreal forest. Canada’s National Forest Inventory uses the FAO definition of forest, “land spanning more than 0.5 hectares where the tree canopy covers more than 10% of the total land area and the trees can grow to a height of more than 5 metres. It does not include land that is predominantly urban or used for agricultural purposes.”^Note Canada’s National Inventory Report to the UNFCCC uses the Marrakesh Accords definition of forest (25% crown closure, with potential tree height at maturity of 5 m or greater and covering an area of 1 ha or greater).^Note

Forest sector: includes the following North American Industry Classification System (NAICS) codes: 113 – forestry and logging; 1153 – support activities for forestry and logging (e.g., forest conservation services, forest fire fighting services, forestry maintenance, log hauling, pest control and timber cruising and valuation); 321- wood product manufacturing and 322 – paper manufacturing.

Grassland, natural pasture and rangeland: areas dominated by grasses or grass-like plants including natural grasslands or native rangelands of the Canadian Prairies used for grazing, as well as other areas dominated by grassy vegetation (e.g., wetland, alpine meadows).

Groundfish: fish living near the bottom of the ocean. Also known as demersal or bottomfish.

Human landscape modification index (HLMI): a composite index used to measure direct human modifications to the landscape, based on the degree that an area has been modified from a natural or semi-natural state, the relationship of an area to the nearest patch of natural and semi-natural land and the size of that patch, and the relationship of an area to the nearest linear feature and the density of those linear features. Values range from 0 to 100, with higher scores indicating more intensively-used ecosystems and lower scores representing more intact ecosystems.

Human freshwater influences index (HFII): a ranked index of selected variables and indicators that reflect anthropogenic influences on freshwater ecosystems including population density, the HLMI, water crossings, dams, temperature change and nutrient emissions from industrial plants, farms and wastewater treatment plants.

Landscape fragmentation: the breaking up of areas of natural and semi-natural landscapes into smaller and more disconnected or isolated patches.

Linear feature density: a measure of linear features that cut across a landscape, calculated based on the length of roads, rail lines, cutlines and electrical transmission lines per unit area.

Managed forest: for the purposes of reporting on greenhouse gas emissions to the United Nations Framework Convention on Climate Change (UNFCCC), forest areas are classified as managed or unmanaged based on the occurrence of management activities for timber and non-timber resources (including parks) and on the level of protection against disturbances.^Note

Mesopelagic: refers to waters at depths of 200 m to 1,000 m. The mesopelagic class includes areas where maximum water depths are within this range.

Natural and semi-natural patch: includes all land classes except built-up and artificial surfaces (settlements and roads), cropland (annual and perennial), harvested forest (from 2001 to 2015) and managed grassland (grass and shrubs used for cattle grazing). A single patch of natural and semi-natural land has a minimum size of 9 pixels (at 30 m x 30 m) or 8,100 m².

Net primary productivity: gross primary productivity (the rate at which photosynthetic plants and bacteria use sunlight to covert CO₂ and water to carbon compounds used to fuel growth (biomass)) less cellular respiration.^Note

Net ecosystem productivity: net primary productivity less heterotrophic respiration.

Net biome productivity: net ecosystem productivity less disturbance losses. Equivalent to net ecosystem carbon balance at the landscape level.

Nitrogen dioxide (NO₂): important in the formation of ozone in the atmosphere and a precursor to fine particulate matter.

Ocean acidification: the process by which the pH level of the ocean decreases mainly because of the absorption of CO₂.

Parkland: ecosystems characterized by tall herbaceous vegetation, scrub and scattered large trees. This ecosystem type occurs for example in the Parkland Prairies ecoprovince, including the Aspen Parkland ecoregion, which is a transitional area between the prairie grassland and the boreal forest.

Peatland: wetland ecosystems with peat deposits that are at least 40 cm thick.

Pelagic fish: fish that live in the water column and that are not dependent on the ocean floor or shoreline for habitat.

Population centre: has a population of at least 1,000 and a population density of 400 persons or more per square kilometre, based on population counts from the current Census of Population. All areas outside population centres are classified as rural areas. Population centres are classified into three groups: small (population between 1,000 and 29,999), medium (population between 30,000 and 99,999) and large urban (100,000 or more).^Note

Potential evapotranspiration (PET): represents the evapotranspiration that would occur without limitations on water supply. PET is therefore linked to the amount of energy available to generate ET in a specific area and is independent of water supply.

Salinity: estimate of the amount of salt dissolved in water.

Sea knoll: an isolated small elevation on the deep seafloor.

Seamount: an elevation of the seafloor 1,000 m or higher, either flat-topped or peaked.

Shellfish: aquatic shelled animals including molluscs (e.g., oysters and mussels) and crustaceans (e.g., crabs and lobsters).

Species abundance and richness: species abundance refers to the total number of individuals of a species in an area, population or community. Species richness refers to the total number of species in an area.

Stratification: the existence or formation of distinct layers in a body of water identified by thermal or salinity characteristics or by differences in oxygen or nutrient content.

Sulphur dioxide (SO₂): emitted when a fuel or raw material containing sulphur is burned or used in industrial processes such as metal ore smelting.

Summerfallow: total area of summerfallow from the Interpolated Census of Agriculture. Summerfallow is a land management practice that involves leaving land fallow in summer to conserve water or manage weeds.

Tame or seeded pasture: total area of tame or seeded pasture from the Interpolated Census of Agriculture. This category represents managed grazing lands supporting introduced forage species and receiving periodic cultural treatment, such as tillage, fertilization, mowing and irrigation.

Thermokarst topography: an uneven landscape in permafrost regions marked by small depressions and lakes, formed by subsidence and following melting of ground ice.

Total water storage change: an estimate of the change in the amount of water stored in the environment above and below the Earth’s surface including groundwater, soil moisture, snow, ice and surface water. To convert water measured in depth in mm to a volume per area in m³/m², divide by 1000.

Tundra: treeless areas of dwarf shrub and other low-lying sedges, mosses and lichen that develop north of the boreal zone and in alpine regions at higher elevation.

Turbidity: a measure of the relative clarity or cloudiness of a liquid, caused by suspended particles (e.g., clay, silt, metals, organic matter, microorganisms), that is measured in nephelometric turbidity units (NTU).

Urban densification: increase in the number of people or residential units within an established area.

Urban greenness: a biotic condition variable for urban areas that classifies population centre areas as green or grey based on the normalized difference vegetation index (NDVI) from moderate resolution imaging spectroradiometer (MODIS). This assessment does not correspond to the actual area of vegetated or unvegetated land cover in cities; rather it measures whether a given pixel (230 m x 230 m) meets a set vegetation threshold (NDVI ≥ 0.5).

Water cycle: the natural cycle in which water evaporates from the Earth’s surface including the oceans to the atmosphere and returns to the earth as precipitation.

Water yield: an estimate of freshwater runoff that provides information on Canada’s renewable freshwater supply. It is derived from data on the unregulated flow of water in rivers and streams. Although water yield provides an estimate of renewable freshwater, it can include some water that is considered non-renewable (e.g., melt water from receding glaciers). To convert water measured in depth in mm to a volume per area in m³/m², divide by 1000.

Wetland: areas that are permanently or temporarily saturated with water for periods long enough to promote wetland or aquatic processes, as indicated by poorly drained soils, hydrophytic vegetation and various kinds of biological activity adapted to a wet environment. Wetlands can be categorized as organic wetlands (more commonly referred to as peatlands) or mineral wetlands and are classified as bog, fen, swamp, marsh or shallow water (< 2 m).^Note

Windthrow: natural disturbance in forests caused by wind, leading to stem breakage or root system failure. May occur at the individual tree or stand level. Also known as blowdown.^Note

C. Acknowledgements

Human Activity and the Environment 2021 was prepared by the Environment and Energy Statistics Division under the direction of Carolyn Cahill (Director), Jeff Fritzsche (Assistant Director), François Soulard (Chief) and Jennie Wang (Editor).

This report is dedicated to the memory of Giuseppe Filoso and Michael Bordt.

Giuseppe was an important member of the HAE team, having worked on the publication for over three decades up till his passing in May of 2019. His camaraderie and sense of humour, as well as his considerable technical GIS expertise, are sorely missed by his numerous friends and colleagues. Joe, we all feel so lucky to have known you and we share countless fond memories of you.

Michael was one of the inaugural members of the HAE team, involved in the first edition produced in the 1970s and many following editions, until he left Statistics Canada in 2012. Friend and mentor, we all benefitted from his boundless imagination and world-renowned energy, gently encouraging us to become better versions of ourselves. Environmental statisticians in Canada and abroad owe a debt of gratitude to Michael, and this edition of HAE in particular builds squarely on his legacy. Michael passed away in August, 2021.

Rest in Peace.

Analysis and reporting were performed by:

Lauren Allen, Jessica Andrews, Ann Helen Jean-Baptiste, Mark Henry, Marcelle Grenier, Roxane Jaffray, François Soulard, Katharine Strong and Jennie Wang

Recognition is given to the following people and areas for their support in data development and production, map and infographic production, factchecking and review of this report:

Lauren Allen, Jessica Andrews, Merinah Buller, Giuseppe Filoso, Ann Helen Jean-Baptiste, Nick Lantz, Hugo Larocque, Eleen Marzook, Yasmina Seddiki, Katharine Strong and Myriam Venasse.

The support and co-operation of the following individuals and federal departments with regards to data provision and review is also gratefully acknowledged, including:

Dr. Alain Pietroniro, P. Eng, Professor and Chair in Sustainable Water Systems in a Changing Climate, Schulich School of Engineering, Department of Civil Engineering, University of Calgary

Agriculture and Agri-Food Canada: Andrew Davidson, Bahram Daneshfar, Xiaoyuan Geng and Melodie Green.

Fisheries and Oceans Canada: Messan Agbaglah, Zeba Ali, Michael Bordt, Gisele Magnusson, Andrea Moore and Andrea Niemi.

Environment and Climate Change Canada: Elizabeth Bush, Jason Duffe, Vincent Cheng, Douglas MacDonald, Brett Painter, Jon Pasher and Susan Preston

Natural Resources Canada: Luc Guindon, Darren Janzen, Shusen Wang and the Canadian Forest Service.