Integrated Stewardship of Great Lakes Coastal Wetlands

Linking concepts of remote sensing with plant invasions in a Great Lakes coastal wetlands complex.

Team surveying wetlands

Monitoring wetlands using UAVs (drones)

One of the ecosystems most vulnerable to invasive species are wetlands. Despite occupying only 6% of the terrestrial and freshwater surface of the earth, wetlands harbor nearly a quarter of the world’s most prolific invasive plants.  These plants can change the physical and chemical functions of a wetland, impacting its ecological value. The dominance of a single plant species that wetlands often suffer after an invasion have a resounding impact on native plant community diversity and structure.   To mediate the impacts of these introductions and the ones to come, better management methods are needed.  We ultimately seek to utilize concepts in remote sensing and link them to these ecological issues to provide better monitoring and management priorities. 

Location of the St. Mary’s River in the Great Lakes.

The region chosen for this study is the St. Mary’s River, the sole waterway connecting Lake Superior to the southern Great Lakes (Fig. 1). The St. Mary’s River forms the U.S. – Canada border in Northeastern Michigan.  This river experiences high shipping traffic and recreational use, making it a conduit for invasive species and their spread between the Great Lakes.  The river has two ecologically distinct invasive plants that threaten the quality of its coastal wetlands. The first, Typha x. glauca (hybrid cat-tail) decreases native plant biodiversity and alters community structure.  To control this plant, treatment efforts must focus on smaller or newer populations, as certain harvest methods have proven effective in removing cat-tail and increasing native plant diversity.  The second invasive plant of interest, Hydrocharis morsus-ranae (European frogbit), has been recently introduced and has begun rapidly spreading through the St. Mary’s. A true floating plant, its overgrown mats are known to greatly alter macroinvertebrate communities and submerged vegetation by shading them out. Each plant presents different ecological management questions and require different techniques to accurately determine their spread and impact. 

Classification of wetland communities in the St. Mary’s River using RapidEye and ESA c-band Synthetic APerture Radar imagery

The purpose of the Typha mapping effort is to provide resources and methodology to provide these low cost, relatively high resolution wetland maps that may be performed on an annual basis. This research in particular focuses on wetland mapping in the Northern Great Lakes, but the methodology and focus on classification of ecologically functional ecosystem types over one broader categories can be used in a variety of coastal environments, especially in areas where reliable ecological knowledge and field data can be utilized by the mapper (fig. 3).  To achieve that goal, we will 1) assess the contribution of multiple temporal radar and red-edge imagery into coastal wetland classification accuracy and 2) compare and contrast how classification methods map overall accuracy of Typha and certain vegetation classes of interest.

Predicted habitat suitability of European Frogbit in Munuscong Bay of the St. Mary’s River

Effective management of European frogbit (Hydrocharis morsus-ranae) in Michigan’s waters requires precise and timely control and prevention efforts.  Our second objective is to conduct athree-tiered risk analysis of frogbit in the St. Mary’s River using species distribution models (SDMs).  SDMs use known locations of species occurrence and environmental variables to predict the probability of species occurrence in an area.   SDMs are commonly used in invasive species management to forecast areas suitable for the establishment of new invasive populations. However, most SDMs are limited in that they typically use broad climate variables or elevation data that exist at a much coarser scale than the organisms or their populations.  In addition, traditional SDMs do not take into account species interactions, which we have observed to play a large role in the biogeography of European frogbit.  Our SDM methods will integrate ecological field data such as fine-scale UAV imagery (Fig. 4), fine-scale environmental variables, human facilitated dispersal,  sophisticated remote sensing, and machine learning techniques to create products that will enable our partners to forecast and manage frogbit in some of the most pristine wetlands left in the Great Lakes. 

Overall, our project hopes to identify integrated approaches to steer changing ecosystems such as Great Lakes coastal wetlands onto sustainable positive trajectories.

HES Researchers

Other Research Collaborators

Eric Clark, Sault Ste Marie Tribe of Chippewa Indians
Joseph Lautenbach, Sault Ste Marie Tribe of Chippewa Indians
Nicholas Reo, Dartmouth College
Shane Lishawa, LUC
Brendan Carson, LUC
Jason Tallant, University of Michigan
Dennis Albert, Oregon State
Beth Lawrence, University of Connecticut

Products

Funding Sources