Semester of Graduation

May 2024

Degree

Master of Science (MS)

Department

Renewable Natural Resources

Document Type

Thesis

Abstract

Wetland loss occurs at an alarming pace globally, with extremely high rates along the northern Gulf of Mexico. Louisiana loses a football field of wetland every 100 minutes: that is 77,000 m2 of wetland bird habitat lost daily. In Louisiana, marsh creation projects combat wetland loss, and while wildlife habitat is often used as a justification for restoration, wildlife receives little to no consideration during and after construction. Habitat characteristics such as site-specific hydrology, vegetation composition, and habitat structure affect the abundance of wetland birds and understanding these features is crucial to creating habitat that will benefit birds. My study compares bird species abundance, vegetation, and site-specific hydrology between natural and created marshes across southeastern Louisiana. I conducted point counts and vegetation surveys at six created and six natural marshes in 2021 and at 10 created and 10 natural marshes in 2022 and 2023. All species of birds seen and heard were recorded and call-back surveys were performed to increase detections of secretive marsh birds. At each created site, I used water-level recording devices to quantify differences in water levels and flooding frequency among sites. My results suggest that all sites vary widely in hydrologic regimes and vegetation communities. I completed 766 bird surveys (created = 413, natural = 353) and identified 9,650 individual birds of 110 different bird species including 88 species at natural marshes and 91 species at created marshes. I classified all observed birds into five habitat guilds: 1) Marsh Specialist, 2) Beach and Estuary, 3) Wetlands Generalist, 4) Habitat Mosaic Generalist, and 5) Forest. Additionally, five focal species were selected: Common Gallinule, King/Clapper Rail, Least Bittern, Seaside Sparrow, and Red-winged Blackbird. To determine habitat factors affecting bird abundance across created and natural sites, I built generalized linear mixed models in a Bayesian framework using the brms package in R. For each guild and focal species, I developed a candidate model set based on four a priori hypotheses on the drivers of bird abundance including a hydrology model, a habitat composition model, a habitat structure model, and a combination model. I ranked models using approximate Leave-One-Out (LOO) cross-validation in the loo package. For most guild and focal species, the combination model was ranked highest, indicating that hydrology, composition, and structure all affect bird abundance. The main drivers of marsh bird abundance were water level variability and plant composition. Vegetation community was a driver of abundance for the marsh specialist guild, the wetland generalist guild, and all focal species. Water depth variables positively affected abundance of the wetland generalist guild, the marsh specialist guild, Common Gallinule, Least Bittern, and Red-winged Blackbird and negatively affected the forest guild. The results of my study indicate that whether a marsh is created or natural is not the driver of bird abundance. Rather, depending on the guild and species, the drivers of avian abundance include hydrologic variation, vegetation communities, and/or vegetation structure. Management implications of my study include building marsh creation sites within the tidal range, as well as the inclusion of certain habitat features, such as tidal creeks and ponds, that can promote hydrologic connectivity and emergent vegetation communities. This study is one of the first to connect physical processes, such as hydrology, to vegetation and wildlife responses on created marsh sites in Louisiana and these types of studies must persist to inform and improve future restoration.

Date

1-22-2024

Committee Chair

King, Sammy

Download

Share

COinS