Coyote Home-Range Analysis

Evaluation of Coyote Home Range and Core Range Data

Data Overview

The coyote radio-collar tracking data, provided by Dr. Eric Gese at Utah State University, was used to determine the home and core ranges of coyotes within a specific study region. The data provided is a subset of a much larger dataset collected by Bryan Kluever of Utah State University during the period 2010-2014 near Dugway, Utah.

Home Range Calculation Methods

Two primary methods were used to calculate the coyote ranges: Kernel Density Estimation (KDE) and Minimum Convex Polygon (MCP). MCP was used to define the total space used by the animal, as it creates a polygon around every single point in the data set. However, this method has several drawbacks, such as sample size effects, sensitivity to outliers, and often including areas never used by the animal.

KDE was used to estimate the probability of use areas, with 95% and 50% quantiles being the typical areas computed. The area in the 95% output represents an area with a 95% probability that the animal is inside that area, considered the general or home range. The 50% area is considered the core area of activity.

Batch Processing and Python Integration

Batch processing played a crucial role in handling multiple datasets simultaneously, automating repetitive tasks to increase efficiency and reduce errors. Basic scripts in ArcPy (Python) were utilized to loop and automate repetitive parts of the process, enhancing the efficiency of the analysis. These scripts were written by Chris Garrard and Tyler Hatch and were integral to this process as they provided a much more efficient processing time.

Editing and Labelling

As a result of processing the density tools, each of the ranges had multiple distinct polygons. To accurately portray the total area of the entire range, I used the 'Edit' and 'Merge' tools to combine these multiple polygons from the same coyote into a single cohesive polygon. This approach ensured an accurate total acreage calculation by consolidating all separate polygons that represented individual coyote ranges into one unified area.

In the process of visualizing and labeling the acreages for the ranges, Arcade expressions within the labelling properties were utilized to ensure clarity and precision in the map displays. Even when the attribute field data was manually set to 'float' and 'numeric' with no decimal points, the labeling within ArcGIS Pro was malfunctioning. To bypass this issue and only display acres as whole numbers, I used the formula Round($feature.Acres) + " " + "Acres" to round the acreage values and append the unit of measurement, "Acres." This approach helped avoid the display of insignificant decimal points, thereby making the labels more readable and professional.

Extracting Vegetation Values and Determining Preference Ratio

To analyze the vegetation types available within the coyote ranges, the 'Extract by Mask' tool was used to clip the LANDFIRE vegetation types to the home range area. Then, the 'Extract Values to Points' tool was used to determine which vegetation type each coyote point intersected, providing important data for understanding the actual use of vegetation types. Furthermore, the 'Frequency' tool was used to understand how many times coyote points intersected a specific vegetation type. This information was essential for developing the preference ratio, indicating coyote habitat preferences. To calculate the preference ratio, I compared the landcover types available within the general home range to those actually utilized by the coyotes. This involved comparing the frequency of coyote points on each landcover type to the expected frequency, which was determined by multiplying the total number of coyote points by the proportional fraction of each landcover type within the general home range. The preference ratio was then calculated by dividing the observed frequency (utilized) by the expected frequency (available), thereby identifying landcover types that coyotes differentially preferred. 

Results and Observations

Figure 1: The first map on the left illustrates the core range (8,375 acres), general range (71,902 acres), and minimum bounding area (180,805 acres) for coyote home ranges. The core range, highlighted in red, indicates areas of highest use, while the general range in blue encompasses areas of moderate use. The yellow minimum bounding area shows the total extent of coyote movement. The second map on the right displays the home ranges of individual coyotes, with each range color-coded and labeled with its size in acres. This map reveals significant variability in home range sizes, reflecting different territorial behaviors among the coyotes.

The accompanying Table 1, provides detailed data on various vegetation classes, showing counts, proportions, expected counts, actual counts, and preference ratios. Vegetation types were determined using a 30-meter LANDFIRE raster layer, with the count field representing the number of 30m x 30m pixels of each vegetation type encountered. Preference ratios, with values greater than one indicating a preference, highlight coyote habitat preferences. Notable preferences with higher sample points were observed for "Greasewood Flats" and "Big Sagebrush Shrublands" areas.

Geospatial Observations

The spatial distribution analysis revealed clear preferences for certain vegetation types, likely providing optimal conditions for foraging and denning. High preference ratios for specific vegetation classes suggest these areas are crucial for coyote habitat management and conservation efforts. Methodologically, KDE was effective for identifying areas of high use but could be influenced by outliers, while MCP provided a comprehensive view of the total area used but lacked the detail of KDE. Integrating both methods offered a balanced approach, capturing both core and peripheral areas of coyote activity.