In this lab we experimented with different methods of calculating a home range for coyote in Utah given coyote location information.

Lab mastery and skills

• Using two separate methods to calculate home range

  • Minimal Convex Polygons; based on the outermost extent that the animal travels and generating a polygon around all other collected points

    • This is accomplished by generating a polygon by animal that is able to contain all other points generated by the animal. This was easily able to be collaborated with the original data by symbolizing the points by animal.

  • Fixed Kernel Density Estimate; generates a polygon based off of a single animals track points and using a probability of home range centered around the most densely frequented areas

    • By providing a dome like shape that covers each vector point with the value highest at the cross at the (x,y) point and decreasing in value towards the outer extent of the dome. Essentially by calculating the area under the dome and adding the values where kernels overlap. This generated a map whose highest values represented the greatest density.

  • Then, running calculations to generate two separate data sets that have separate thresholds to include 50% of the highest values and subsequently 95% of the highest values. Lastly, converting the rasters to polygons to help better communicate the information.

• Utilizing python script to run multiple inputs and generate a home range for each animal

  • In order to avoid tediously running Kernel polygons through Arcpro for each animals, using a python script can accelerate this process. Just like what was done manually priorly in arc, with a script written by Chris Garrand and Tyler Hatch, python evaluates the attribute table of the input data, designates between the animal fields, runs kernel densities, is able to also account for 50% and then 95% point density functions, and eventually saves the data.

• Utilizing polygons to extract a mask of vegetation cover and discover preference ratios by land cover

  • In addition to generating polygons for each individual animals python also generated a density point polygons for the entire dataset, this is what will be used to extract a layer for the vegetation input data. This layer appears identically to the polygon for the kernel data set. Then, I calculated a frequency table that counts the number of time a vegetation raster type occurs within this polygon. And additionally extracting raster values by coyote point to help better understand vegetation type that is actually being occupied.