{"id":11,"date":"2016-11-29T11:49:00","date_gmt":"2016-11-29T18:49:00","guid":{"rendered":"https:\/\/blogs.ubc.ca\/snappingturtle\/?page_id=11"},"modified":"2016-12-02T09:55:37","modified_gmt":"2016-12-02T16:55:37","slug":"methodology","status":"publish","type":"page","link":"https:\/\/blogs.ubc.ca\/snappingturtle\/methodology\/","title":{"rendered":"Methodology"},"content":{"rendered":"

1.\u00a0DATA GATHERING and\u00a0PREPARATION<\/strong><\/p>\n

Based on research of habitat preferences, we selected important habitat qualifications for both snapping turtle hibernation and nesting sites. We then obtained corresponding spatial data from a variety of sources, including the Region of Waterloo’s open data site, the Land Information Ontario database, and Geogratis. As referential data, we\u00a0found Waterloo Regional boundaries, creeks and rivers, wetlands, land cover type, roads, soil types, elevation and elevation points.<\/p>\n

We were also hoping to find a dataset of snapping turtle sightings in southern Ontario in order to evaluate the efficacy of our spatial analysis. However, once we downloaded the only existing dataset, we found that the data was too sparse and did not cover most of our analysis area.\u00a0<\/em><\/p>\n

After gathering our data, we formatted it to serve the purposes of our analysis, including re-projecting the shapefiles, clipping them to the study area, and joining\/editing attribute tables to make sure that they were ready to be used. For example, our land use shapefile had an attribute table with approximately 20 different land use types! For the purposes of our analysis we needed to simplify them by combining categories into\u00a0more workable, relevant groups\u00a0(for example,\u00a0Treed Swamp, Marsh, Bog, Fen, and Thicket Swamp became reclassified as Wetland, universally conducive to snapping turtle habitat).<\/p>\n

We also had to join attribute tables to various soil layer files until we could access the ‘dominant sand fraction’ class on the appropriate soil layer. We chose to use the ‘dominant sand fraction’ class in our analysis because it displayed soils based on the grain size of the dominant sand fraction: coarse, medium, etc. (Snapping turtles typically prefer coarse sand and even gravel for their nesting sites).
\nIn order to include snapping turtle’s preference for sunny nesting areas, we needed to create an aspect map. We obtained a DEM from the Region of Waterloo’s open data catalogue, created a TIN from the elevation points, and converted it to a raster layer. We then converted this raster layer using the Surface > Aspect tool to create an aspect raster, clipped it to the project boundaries and reclassified it into five directional classes. This way, we could compile the south and southwest facing aspects (the most preferable) together.<\/p>\n

\"screen-shot-2016-12-01-at-11-22-28-pm\"<\/p>\n

\"screen-shot-2016-12-01-at-11-59-53-pm\"<\/p>\n

 <\/p>\n

 <\/p>\n

 <\/p>\n

 <\/p>\n

 <\/p>\n

 <\/p>\n

 <\/p>\n

2.\u00a0CREATING EUCLIDEAN DISTANCE RASTERS<\/strong><\/p>\n

\"screen-shot-2016-12-01-at-5-56-59-pm\"To ensure that the distance to bodies of water and the distance away from\u00a0roads was considered in our analysis, we needed to create layers that distinguished between different distances from features.
\nAlso recognizing that our river layer only included major rivers, and that snapping turtles can hibernate even in\u00a0small bodies of water, we realized the necessity of including a creek layer as well. From these two layers, we\u00a0created Euclidean Distance rasters. We also created Euclidean Distance rasters with our road layer by separating roads into three categories: high use, medium use, and low use, and creating separate layers for each category. Using these new layers, we created Euclidean Distance rasters to symbolize the distance from different types of roads.<\/p>\n

3. DETERMINING WEIGHTS for the MULTICRITERIA\u00a0EVALUATION (MCE)<\/strong><\/p>\n

We then used an analytical hierarchy process (123AHP.com) to\u00a0rank the importance of each factor for habitat suitability. We did this process twice – once for hibernation habitat, and again for nesting habitat. These values were added to our habitat suitability tables.<\/p>\n

Hibernation\u00a0Habitat Factor Weights:<\/p>\n