Kansas State University
205 Leasure Hall
Manhattan, KS 66506-3501
Phone: (785)532-6070
Fax: (785)532-7159
kscfwru@ksu.edu

Homepage

About the Unit
Cooperators
Mission Statement
History
Unit News
Courses
Publications

Facilities

People
Faculty and Staff
Graduate Students

Research
Current Projects
Completed Projects
Technical Assistance
Research Experience for Undergraduates (REU)

Resources and Employment
Graduate School
Graduate Applications

Forms and Manuals
Assistantships and Positions
Careers

Verifying Ground-based Habitat Quality Monitoring and Micro-Habitat Selection by Lesser Prairie-Chickens (Tympanuchus pallidicinctus) with Remote Sensing Technology

Investigators:
Stephane Manes
Matt Bain
Kevin Price

Project Supervisor:
Dr. David Haukos

Funding:
Play Lakes Joint Venture
U.S. Geological Survey
The Nature Conservancy
CommonGround Capital

Cooperators:
AgPixel

Location: Throughout Kansas

Completion:  July 2016

Status: Completed

Objectives:
Test the feasibility of modeling vegetation characteristics associated with Lesser Prairie Chicken nesting habitat using image data captured by airborne sensors at multiple resolutions.
Compare the efficacy of image data collected via UAS platforms to image data collected by manned aircraft.

Results:
Vegetation characteristics associated with Lesser Prairie-Chicken nesting habitat, such as plant height, can be successfully modeled using image data captured by airborne sensors at multiple resolutions. This study shows that field data points with similar vegetation parameter measurements can be successfully clustered (i.e., classified) using NDVI spatial variability extracted from image data. For Gardiner Ranch, statistically significant differences were found among the clusters for ten of 13 field parameters, with seven at p < 0.05, and three at p < 0.1. For the Hoeme Ranch, statistically significant differences at p < 0.05, were found among the clusters for nine of 12 field parameters for which data were available. Although the results for Hashknife Ranch were not on par with Gardiner and Hoeme, statistically significant differences at p < 0.5 were found for two of 13 parameters and at p < 0.10 for four of 13 parameters. It is worth noting that for all three ranches, Robel Pole 100% data, Highest cm, and Point Center PC Vegetation Height, were statistically significant at p < 0.10, with most tests resulting in statistical significance at p < 0.5. This indicates that the methodology developed successfully classifies field data points with regard to vegetation height parameters. For all study areas for which 2015 nest data were available, nest sites were found to be significantly different statistically from random points, in each case at p < 0.001. Additionally, for nest sites, the amount of spatial variability, as measured by NDVI variance, generally fell within a relatively narrow range of values. NDVI variance patterns for nest sites suggest that the birds prefer somewhat, but not completely homogeneous, vegetation conditions at nest sites. Ultra-fine resolution image data collected via manned aircraft platforms appears to be optimal for modeling LPC nesting habitat. The methodology developed was successful using imagery with 5.0 to 10.0 cm pixel resolution. The results suggest that currently, for the purposes outlined here, there is little to be gained by collecting UAS data at resolutions finer than 5.0 cm. Although the efficacy of 3 D modeling of vegetation characteristics within the study areas was hampered by the vegetation characteristics and limits of the technology, as technological advances are made, continued research will likely yield better results.

The logical next step in this research will be to adapt and refine the methodology developed here to map larger study areas with regard to vegetation characteristics associated with preferred Lesser Prairie- Chicken habitat. To accomplish this, a formal aerial and field data collection protocol can be developed jointly by the research partners. Adequate lead time and good planning will save resources and will insure that adequate data resolution and optimal field parameter measurements are collected concurrently and in a manner that optimizes the accuracy and success of the research undertaken.

Additionally, analyzing nest data and field parameters, not just by discrete study areas (i.e., by ranch), but also as a consolidated dataset may yield additional useful information. Locating additional confirmed nest sites would be helpful in building confidence in the ability to map preferred nesting habitat. It might also be informative to analyze nest data to determine if survival is greater at sites exhibiting consistent NDVI variance patterns.

In summary, analyses of ultra-high spatial resolution image data, based on the methodology described in this report, can greatly assist in meeting the challenges associated with preservation of Lesser Prairie-Chicken habitat.


Gardiner

Products:

Report:
Price, K.P., L. Brien, D. Burchfield, and J. Bryant. 2015. Lesser prairie-chicken habitat mapping project: final report. Submitted to the Playa Lakes Joint Venture, Lafayette, Colorado.