Date of Award

1-1-2025

Degree Type

Dissertation

Degree Name

Doctor of Philosophy in Biological and Environmental Sciences

Department

Natural Resources Science

First Advisor

Scott R. McWilliams

Second Advisor

Peter Paton

Abstract

Waterfowl that occupy seasonally available habitats cope with annual fluctuations in resource availability by undertaking regular annual migrations between winter and breeding sites. Across the annual cycle, migratory waterfowl are subject to environmental pressures and energetic tradeoffs that can have carryover effects from one season to another. Investigating the movement ecology and space use of waterfowl across the annual cycle can reveal patterns in waterfowl behavior, phenology, and distributions to inform conservation and management strategies. Further, comprehensive approaches that incorporate multiple data sources improve our ability to address basic and applied questions to inform waterfowl management. Here, I integrated data from individual-borne tracking devices, spatiotemporally extensive surveys, high resolution remotely sensed environmental data, and implicit isotopic signatures to investigate the movement ecology and space use of 3 species of waterfowl (American Black Duck [Anas rubripes], Atlantic Brant [Branta bernicla hrota], and Greater Scaup [Aythya marila]) that share a common estuarine wintering area in Rhode Island, USA.

In Chapter 1, I investigated how proximity to shellfish aquaculture influenced habitat selection and movement patterns of American Black Ducks during the non-breeding season (~ Oct - May) in Rhode Island, USA by fitting integrated step selection functions. While shellfish aquaculture is a globally expanding industry, including in urban estuaries that support non-breeding waterfowl, its effects on the spatial distribution of non-breeding waterfowl are poorly understood. I found the extent to which proximity to aquaculture influenced habitat selection of American Black Ducks depended on factors specific to individuals’ primary non-breeding sites. Further, proximity to aquaculture did not have biologically meaningful influences on American Black Duck movement rates across all sites. American Black Ducks across sites consistently selected for areas better suited for aquaculture (i.e., areas of potential future development) relative to areas poorly suited for aquaculture (i.e., areas less likely to be developed). The continued expansion of aquaculture into preferred American Black Duck habitats will increase American Black Duck interactions with aquaculture and therefore needs to be considered in the decision-making process for siting future aquaculture leases. Further research into American Black Ducks’ fine-scale interactions with aquaculture as it expands in preferred coastal habitats will provide evidence for the extent to which continued aquaculture expansion will influence non-breeding American Black Ducks.

Species distribution models (SDMs) are a key tool in avian ecology for predicting species' spatial variation and guiding spatially explicit management, though outcomes may vary based on the type of data collected. In Chapter 2, I used land-based survey and GPS telemetry data to compare the predictive performance and precision of SDMs fit using hierarchical generalized additive models for 3 species of waterfowl, Atlantic Brant, American Black Duck, and Greater Scaup, in Rhode Island, USA. I found predictive performance was high for both land-based survey and GPS telemetry models, but precision was better for survey than GPS telemetry models. I demonstrated that differences in data sampling schemes and space use patterns among species affected the predictive performance of SDMs, especially for those based on GPS telemetry data. Moreover, species-specific differences in SDMs were consistent with species’ differing foraging ecologies. Finally, integrating GPS telemetry model predictions as a covariate into an updated land-based survey model did not improve predictive performance or precision for any study species relative to the original survey model. I demonstrated that estimates of precision can substantially influence confidence in predicted SDMs and help determine which SDMs to use for making management decisions. Therefore, I urge researchers and managers to consider measures of precision (e.g., coefficient of variation) in predicted species distributions when developing management and conservation plans.

In Chapter 3, I quantified the movement ecology during spring migration of 3 species of waterfowl that shared a common wintering area. Migratory birds face considerable temporal and energetic constraints during spring migration, and for waterfowl in particular, spring migration phenology is linked to subsequent breeding season success. To balance energetic demands, waterfowl may optimize spring migration by adopting migratory strategies that either favor minimizing time or energy spent on migration and are related to species’ life history characteristics. I used GPS telemetry to quantify inter- and intraspecific variation in spring migration strategies of waterfowl that differ in life history traits: American Black Ducks, Atlantic Brant, and Greater Scaup. American Black Ducks demonstrated the most time-minimizing spring migration strategy compared to Atlantic Brant and Greater Scaup that demonstrated more energy-minimizing spring migration strategies. Atlantic Brant and Greater Scaup also had lower variation in metrics of migratory stopover behavior than American Black Ducks, highlighting the importance of stopover habitat for long-distance migrating species. I found limited evidence of carryover effects of winter movement and space use patterns on spring migration behavior across all 3 species. My findings demonstrate that species can have life history characteristics associated with both time- and energy-minimizing spring migration strategies, supporting the concept that optimal spring migration exists on a continuum between the two strategies.

Finally, determining the breeding season origins of migratory waterfowl that occupy high-latitude, remote areas is challenging yet key to effective management. Therefore, in Chapter 4, I integrated telemetry and stable-isotope (δ2H) datasets to estimate breeding season origins for Greater Scaup, an Arctic- and subarctic-breeding diving duck and declining species for which limited information is available on breeding areas and phenology. To illustrate this approach, I deployed satellite transmitters on non-breeding birds in Rhode Island, USA and the lower Laurentian Great Lakes of North America and collected feathers from these same individuals plus others captured or harvested at both locations. Using these data, I 1) compared breeding season origins and phenology between these 2 non-breeding sites, 2) demonstrated that telemetry-derived breeding season locales can be used to refine likely breeding and molt origin assignments derived from feather stable isotopes, 3) identified differences in likely breeding origins for the two non-breeding sites, and 4) compared likely breeding versus molting origins for the Rhode Island non-breeding site. The breeding season phenology and distributions of Greater Scaup that I documented do not align well with the timing or stratification of the U.S. Fish and Wildlife Service and Canadian Wildlife Service Waterfowl Breeding Population and Habitat Survey, specifically in northern Quebec and Alaska. Further, for waterfowl species that breed in remote habitats that are logistically challenging to access, the integration of telemetry and feather isotope information as I outlined provides a promising approach for delineating breeding and molting origins for distinct non-breeding subpopulations to quantify migratory connectivity. Thus, my approach, when applied to larger samples, could effectively inform waterfowl monitoring programs and address conservation challenges for high latitude breeding migratory birds.

Download

Available for download on Thursday, May 27, 2027

Share

COinS