Master's Theses

Department

Biology

Degree Name

Master of Science (MS)

Abstract

Quivira National Wildlife Refuge in Kansas, United States partnered with Fort Hays State University Hays, KS in 2014 to begin a collaborative research project that aimed to develop a long-term monitoring protocol guided by the Comprehensive Conservation Plan for the refuge published in 2013. This plan identified specific wildlife taxa underrepresented in management impact assessments throughout the property. As a result of this plan, surveys were established to monitor interactions between upland breeding birds and the vegetation community. I conducted point count surveys in 2016, 2017, and 2018 for 122 observation points across four transects. I measured seventeen vegetation variables at each observation point between 13-26 July 2016, 5-13 June 2017, 24-27 July 2017, and 18-22 June 2018. I obtained multi-spectral imagery for June 2017 from GeoEye-1 satellite operated by Satellite Imaging Corporation to compare the 17 vegetation variables with remotely-sensed vegetation data. I used reflectance signatures of five unique vegetation classes to generate five vegetation cover types by using supervised Maximum Likelihood Classification in ArcGIS. I modeled single-season occupancy by using traditional and remote-sensed vegetation variables as covariates for Bell’s vireo (Vireo bellii), grasshopper sparrow (Ammodramus savannarum), upland sandpiper (Bartramia longicauda), warbling vireo (Vireo gilvus), and western kingbird (Tyrannus verticalis). Covariates derived from multi-spectral imagery consistently performed equal to or better than comparable field-measured covariates for four of the five species. I then applied the multi-spectral imagery classification technique to imagery of proposed wilderness area at Crescent Lake National Wildlife Refuge in Nebraska, United States captured 27 June 2018 to assess translatability of these methods. I identified five habitat classes sensitive to vegetative productivity and exposed bare ground that potentially could be reassessed multiple times through the 15 year lifespan of the Comprehensive Conservation Plan to determine vegetation changes across the 9,915 hectares. These assessments promote an adaptive management approach to plant community dynamics on federal properties by allowing for annual assessments that better mimic real world dynamics but require a fraction of the resources.

Keywords

habitat identification, remote sensing, grassland, wetland, US Fish and Wildlife Service

Advisor

Dr. Mitchell Greer

Date of Award

Fall 2018

Document Type

Thesis

Rights

© The Author(s)

Comments

For questions contact ScholarsRepository@fhsu.edu

Download


Share

COinS