Methodology
How was the data generated?
In order to properly convey the possible habitats for Barn Swallows, several GIS techniques were required to achieve each objective. Most of these techniques were through ArcMap.
The project went through several stages to show habitat suitability for Barn Swallows. The first stage was to collect land use data for the Metro Vancouver district. This was supplied by Michael Coombes through the Land Cover Classification Public project report, supplying land cover classification of 5m by 5m resolution. After collecting this data and importing it into ArcMap, the question was, which classification would Barn Swallows be found under? The answer was the “urban mix” classification. Under this classification is included “ß Mixed fabric of ground covers corresponding to detached housing, trees, lawns, landscaping vegetation, streets, sidewalks and driveways” (Coombes, 2013). Most importantly under this was the detached housing, which Barn Swallows use exclusively for breeding. To display areas matching this classification, all polygons that were categorized as “urban mixed” were selected through the attributes table and a new layer was created from this collection, becoming a layer of exclusively urban mix classification.
Next, it was necessary to find insect abundance within Metro Vancouver, as this would show the best feeding grounds for Barn Swallows. How this was done was similar to the previous task, using the land cover classification. Using these classifications, they were ranked by our clients expert opinion (O. Lansdorp, personal communication, March 25, 2014) on a scale from 1 to 5, 1 denoting the lowest abundance and 5 denoting the highest. After the ranks were supplied, the polygons were converted into raster format using the “feature to raster” tool. Each cell had a 5mx5m resolution, in keeping with the original data. After rasterized, the data was then reclassified as follows; Bare soil was given a rank of 2, Grass/herb was given a value of 4, shrub was given a value of 5, mixed tree was given a value of 5, coniferous a value of 4, deciduous was given a value of 5, built environment was given a value of 1, urban mix a value of 1, barren was also 1, mountain shadow and urban/ mountain shadow were given a value of 3, water was 5, barren-modified was given a value of 1 and roads were 0. All other values were given a value of 0. After the data was reclassified it was set aside for the next step in the project.
The method used to classify feeding habitats for Barn Swallows was using the data collected on insect abundance and combining that with the possible nesting grounds in urban mixed environments, 500m study areas were created. In ArcMap, a shapefile with random points was created using the “create random point” tool and using the urban mix shapefile as the constraining feature in creation, forcing all random points created to only appear within urban mix polygons, where Barn Swallow nest would appear. After creating a randomized number of points, 295777 in all, these polygons were made into individual study areas, 500m in diameter, as per the client’s specifications. To create the 500m study areas, the “create hexagon tessellation tool was used. Using this tool, interlocking hexagons 500m wide were created around many of the points, though not overlapping. Only 8025 of these hexagons were created, as that was all that was necessary to cover the entire urban mix classification of Metro Vancouver. Using these hexagon tessellations and the reclassified raster of insect abundance in Vancouver, it was then possible to use the “zonal statistics” tool to create a sum of all the ranks found in each study area, creating a total score ranging from as low as 353 to as high as 32514.
For an additional analysis, the insect abundance objective, the eleven nesting locations supplied by the client were used. Using the GPS coordinates supplied, points were created within ArcMap at the precise GPS coordinates, and processed in the same fashion as the random points in urban mix.
Finally, to achieve the objective of finding livestock historical changes, it was necessary to go through several steps. First, municipality shapefiles were required, which was obtained from the SIS Data Warehouse at Simon Fraser University. Second, livestock statistics were necessary, but could only be found in .pdf format from the Department of Agriculture of British Columbia. After collecting this data for each municipality, it was necessary to enter new fields into the attribute table of the municipality data, creating fields ranging from 1986 to 2006 for cows, sheep, pigs and horses. After these fields were created, the data given by the statistical reports of livestock and agriculture were manually added into each field using the field calculator.