Appendix G: Thousand Islands National Park case study methodology

Case study area

This case study focused on several distinct areas for different components of the study (Map 4.1). The boundaries of these different areas are described below:

1. Land cover: The Canadian portion of the Thousand Islands Ecosystem boundary delimits the zone analyzed for the land cover information.
2. Pressures: In order to better depict the external influences on the park, a 100 km buffer around the Thousand Islands Ecosystem was created and served as a boundary for the pressures analysis, namely for the population and agriculture data. Any census consolidated subdivision (CCS) ¹  that touched the 100 km buffer, or was found within it, was included in the pressures analysis. ²  Data in Table 4.1 Population and agriculture are presented for this 100 km buffer boundary and the Canadian portion of the Thousand Islands Ecosystem boundary.
3. Valuation: The valuation component of the case study focused solely on the Thousand Islands National Park land fragments, a total of 22.3 km².

Methodology for the analysis of land cover

Several land cover compilation sources, each with varying levels of image resolution and spatial coverage, were compared to determine the land cover of both the Thousand Islands National Park and Thousand Islands Ecosystem.

Differences in the land cover by image source are presented in Tables 1 and 2 (Appendix G) at the 12-class land cover level, ³  which includes detailed forest land cover classes. Land cover classes were determined based on the commonality of the classes between the different image sources.

Instead of Canada Centre for Remote Sensing (CCRS) land cover, the reference land cover product for the Measuring Ecosystem Goods and Services (MEGS) project, Parks Canada LANDSAT-TM land cover was chosen to represent the land cover of the Thousand Islands National Park and Thousand Islands Ecosystem because:

1. wetlands are better represented in Parks Canada LANDSAT-TM than in CCRS land cover
2. the finer resolution is better suited for small area analysis
3. Parks Canada LANDSAT-TM is the image source used by Parks Canada for their analyses
4. Parks Canada LANDSAT-TM would also permit future analysis of parks in other areas of the country
5. Parks Canada LANDSAT-TM land cover classes are a good representation of Canada’s land cover; however, no time-series exist for Parks Canada LANDSAT-TM.

Valuation of ecosystem goods and services by land cover type

The overall value of ecosystem services by land cover was estimated using existing monetary values of ecosystem goods and services (EGS) taken from a 2009 report by Troy and Bagstad, Estimating Ecosystem Services in Southern Ontario. Dollar values per hectare were weighted based on land cover areas for each of the data sources used in the analysis of land cover. The use of different land cover sources resulted in significant variations in the estimates of the value of the park’s EGS, highlighting the sensitivity of the valuations to the resolution of land cover data.

Methodology for applying monetary values from Troy and Bagstad (2009) to the Thousand Islands National Park

The GIS layers used include the Thousand Islands National Park protected areas layer (excluding islets, Mainduck and Yorkshire Islands) and the Troy and Bagstad land cover and monetary valuation layer. The case study took the values per hectare from Troy and Bagstad and extracted the monetary data found for the park area, using the polygons for the park confirmed by Parks Canada. For each satellite data source, the Thousand Islands National Park protected areas layer was used to identify the land cover present in the park.

In order to compute monetary values for all land cover compilations used in the case study, a weighted average valuation method was applied. The Troy and Bagstad land cover classifications were organized and grouped to the MEGS eight-class land cover, the highest level of detail possible from this source. For each land cover class, the various Troy and Bagstad categories were weighted according to the land area that they covered in the park. For each land cover data source, the land cover classes were rolled-up to match the MEGS eight-category classification. Each category was then multiplied by the weighted average generated from the original study (dollars per hectare).

Valuation of individual ecosystem goods and services in the Thousand Islands National Park, by land cover type

Selecting the study sites

To facilitate comparisons between the Thousand Islands ‘policy site’ and potential study sites for benefit transfer, original valuation studies from across Canada were gathered, geo-spatially referenced, and incorporated into the MEGS geospatial database. ⁴  This geospatial framework links each study site with a variety of data that describe its physical land cover characteristics and socio-economic context, including proximity to settled areas. As noted in Appendix A, characteristics of land cover ecosystem units (LCEU), such as terrain ruggedness and land cover, are closely tied with the types of ecosystem services that that land area may have the potential to provide. Establishing these links for each study site helps analysts decide which study sites have the potential to be used as ‘donors’ of ecosystem service values for the policy site valuation exercise. In Map 1 (Appendix G) selected geo-spatially referenced study sites for the southern Ontario region are overlaid with deciduous forest land cover and the settlements footprint map layers.

Ecosystem services values within a site are a function of underlying biophysical characteristics. However, they are also influenced by the type of beneficiaries (e.g., number, income and preferences) of the ecosystem’s services, as well as by the availability of substitute and complementary sites and services. ⁵  These aspects are taken into account when selecting potential sites from which to transfer benefit values, using information provided by the geospatial database, including spatially referenced socio-economic data in conjunction with information provided within the studies themselves, to conduct analysis of potential sites.

Transferring the unit values

Having chosen study sites that aligned as closely as possible with the policy site, the next step was to transfer ecosystem service values from the study sites to the Thousand Island National Park policy site, based on specific targets. Depending on the type of value being transferred, the targets were expressed in dollars per household per unit area (hectares) per year, as in the case of option, bequest and existence values, ⁶  or as dollars per person per year, as in the case of recreation values. ⁷ 

Adjusting the unit values

While a close alignment between study and policy sites is a key requirement of any unit transfer exercise, transferred values can be further adjusted for socio-economic differences that exist between sites. ⁸  Individuals’ willingness to pay for any good or service is dependent upon their socio-economic circumstances in addition to many other factors, such as values and beliefs of the beneficiaries of services. Regarding socio-economic circumstances, a number of studies ⁹  have shown that the willingness to pay for ecosystem services rises along with increases in household income. For this reason, transferred values for the Thousand Islands National Park were adjusted to account for inter-site differences in income.

Adjusting unit values requires statistically sound socio-economic information gathered and reported on a spatial basis for both the original study site and the policy site. Use of the MEGS geospatial framework facilitated the adjustments.

The Thousand Island National Park has the potential to provide a wide variety of services to beneficiaries in and around the park. The experimental valuation exercise considered a sub-set of these services: recreation and option, bequest and existence values.

Next appendix | Previous appendix