Labs‎ > ‎

Lab 10 - Habitat Modeling

Introduction

Lab 10 in Advanced GIS and Spatial Analysis (WATS 6920) was an opportunity to apply GIS tools in habitat modeling.  Using 3 inputs, Vegetation Classifications, Slope, and Distance to Water, we developed a framework to predict the probability of ungulate utilization across the landscape of the Escalante Watershed in south-central Utah.

Objectives

As defined, our objectives were to:

  • Gain experience running fuzzy logic model using common and familiar spatial datasets
  • Exposure to Matlab
  • Articulate a critical evaluation of model, methodology, and model outputs
  • Background

    Domestic ungulates (in our example, cows) are released to graze during the summer months on public lands across the West under the federal grazing allotment permit system.  This seasonal influx of high densities of large herbivores can have negative consequences on the landscape including: increasing erosion, altering plant community composition, increasing competition with native herbivores, influencing predator population and spatial dynamics, and disease transmission implications.  In this exercise we ultimately produced a map, presumably with predictive potential in determining habitats cattle will use in the absence of fencing or other active management.   

    Context Map - Escalante Watershed

    Click on image below to enlarge:


    For an Adobe PDF of above Context Map - Escalante Watershed click here.

    Input 1 - Vegetation Input

    Taking pity upon us, Joe and Shannon provided us with the finalized inputs that were eventually incorporated into our ungulate habitat occupancy model.  The first input, Ungulate Foraging Preference, was derived from LANDFIRE land cover data (2008).  Existing cover types were classified as follows:

    I simply added the provided file input_veg.asc, and altered the number and color code of categories to match the above schematic.

    Ungulate Foraging Preference Map - Escalante Watershed

    Click on image below to enlarge:


    For an Adobe PDF of above Ungulate Forage Preference Map - Escalante Watershed click here.

    Input 2 - Percent Slope

    As before, the percent slope .asc file was provided, and only required symbology manipulation.  This input was derived using the Slope Tool, Percent Option, performed on a 10 meter DEM.

    Percent Slope Map - Escalante Watershed

    Click on image below to enlarge:


    For an Adobe PDF of above Percent Slope - Escalante Watershed click here.

    Input 3 - Distance to Water

    The map below was derived from the provided input_distWater.asc file.  Procedures are described in detail here.

    Euclidean Distance to Water Buffers Map - Escalante Watershed

    Click on image below to enlarge:


    For an Adobe PDF of above Distance to Water Map - Escalante Watershed click here.

    Final Product - Probability of Ungulate Occupancy Model - Escalante Watershed

    As described in our instructions, I used Matlab software and ran the GrazingProb_3input.fis code.  As prompted, I defined the inputs (provided).  The FIS calculation took approximately 5 minutes, as did the next step, saving the FIS Grid to File.  After bringing the .asc file into ArcMap, I used the ASCII to Raster command (I needed to disable the run in the background command for it to complete).  I then defined our projection (UTM Zone 12N) and Calculated Statistics such that I could manipulate the display.  Following the 6 C’s of Cartography I prepared the map below.

    Click on image below to enlarge:


    For an Adobe PDF of above Probability of Ungulate Occupancy Model click here.

    Discussion

    My results spatially represent the probability of ungulate occurrence within the Escalante Watershed.  It is important to note that in spatial models, as with any modeling exercise, our resultant models are only as good as our inputs.  Implicit within our models are assumptions about vegetation types that cattle select for, as well as topographical limitations based on assumptions about the ability of cattle to occupy areas of differing slopes and distances to water.  As one who spent almost four years studying and modeling the predator-prey dynamics of the Blackfoot Valley, Montana, I note that an important input that may be missing from our analyses is the probability of predation risk.  Ungulates may not use cells with preferable forage, desirable slopes, and in close proximity to water if these areas make them more susceptible to apex predators.  However these inputs represent a parsimonious start in predicting ungulate occupancy.    

    References

    Joe Wheaton.  2014.  Advanced GIS Courses.  Accessed 03/21/2014  

    ESRI Forum ArcGIS.  Accessed 03/21/2014  


    Ċ
    Jarod Raithel,
    Mar 21, 2014, 3:05 PM
    Ċ
    Jarod Raithel,
    Mar 21, 2014, 4:11 PM
    Ċ
    Jarod Raithel,
    Mar 21, 2014, 3:35 PM
    Ċ
    Jarod Raithel,
    Mar 21, 2014, 4:31 PM
    Ċ
    Jarod Raithel,
    Mar 21, 2014, 3:49 PM
    Comments