Publication Date

8-2018

Date of Final Oral Examination (Defense)

6-13-2018

Type of Culminating Activity

Thesis

Degree Title

Master of Science in Biology

Department

Biology

Supervisory Committee Chair

Jodi Brandt, Ph.D.

Supervisory Committee Member

Ian C. Robertson, Ph.D.

Supervisory Committee Member

Stephen Novak, Ph.D.

Abstract

Great Lakes coastal wetlands are some of the most crucial ecosystems to biodiversity in the Great Lakes Basin, yet suffer increasing degradation due to invasive plants. Wetland plant invasions can be controlled in their initial stages, but early detection of invasive plants using field surveys are often untenable due to budget constraints. Remote sensing techniques offer solutions to management objectives during the early stages of invasion on a landscape scale due to their ability to cheaply create spatially explicit information about plant distributions. Some invasive plants, such as Typha x. glauca, are conspicuous on a large scale, and can be mapped to their current extent using new satellite and modeling techniques. Inconspicuous invasive plants however, such as Hydrocharis morsus-ranae, may be undetectable by remote sensing sources and require predictive strategies. In this thesis I explored the use of remote sensing in the management of a conspicuous and inconspicuous invasive wetland plants in the St. Mary’s River, MI. I successfully classified the current extent of conspicuous Typha x. glauca and other wetland vegetation types to provide spatially explicit maps for early detection and management and examined methods that can be adapted for use in emergent wetlands worldwide. The habitat suitability of inconspicuous Hydrocharis morsus-ranae was also determined using novel fine-scale habitat covariates determined from lidar and radar. Habitat covariates derived from these sources should see wider use in species distribution modeling, particularly with wetland plants, to create better predictions of invasive plant expansions. Implementation of new and upcoming remote sensing data sources and methods will allow for better invasive wetland plant management at greater spatial and temporal scales than field studies alone.

DOI

10.18122/td/1451/boisestate

Download

Share

COinS