Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.

Author ORCID Identifier

https://orcid.org/0000-0002-5931-6144

AccessType

Open Access Dissertation

Document Type

dissertation

Degree Name

Doctor of Philosophy (PhD)

Degree Program

Environmental Conservation

Year Degree Awarded

2023

Month Degree Awarded

February

First Advisor

Chris Sutherland

Second Advisor

Adrian Jordaan

Third Advisor

Graziella DiRenzo

Fourth Advisor

Sean Sterrett

Subject Categories

Ecology and Evolutionary Biology | Population Biology | Terrestrial and Aquatic Ecology

Abstract

This dissertation research focuses on the population ecology of the northern diamondback terrapin (Malaclemys terrapin terrapin) in Wellfleet Bay, MA. The northern diamondback terrapin is a Massachusetts-threatened turtle species restricted to estuarine environments. With a range from Cape Hatteras, North Carolina to Massachusetts, Cape Cod Bay is the northernmost part of the subspecies range. Mass Audubon’s Wellfleet Bay Wildlife Sanctuary has been using capture-mark-recapture (CMR) methods since 1980 marking over 3,000 individuals; however, low detection rates and variable search effort have resulted in unreliable population estimates not suitable for informing conservation practices within the bay. Low sample sizes, female biased captures, small study areas and use of traditional CMR methods has challenged the reliability of estimates in early studies throughout their range. Moreover, the seasonal movement ecology of this species suggests spatiotemporal dynamism in both detection and abundance, which is not formally understood in this system. Motivated by these critical knowledge gaps, my research aimed at estimating site-specific density and distribution patterns of the northern diamondback terrapin in Wellfleet Bay by developing two standardized population monitoring survey protocols, assessing environmental drivers of spatiotemporal variation in density, predict changes in density and distribution over the season and evaluate the performance of both population monitoring methods through a qualitative comparison of estimates. My research will contribute towards achieving much-needed baseline population estimates, and ultimately inform future conservation efforts at local and regional scales.

In Chapter 2, I address the challenges of traditional CMR methods by developing the first standardized spatial capture-recapture protocol for diamondback terrapins. The protocol was tested in my first field season (2018) and refined due to lack of captures for my second season (2019). I analyzed capture-recapture data on northern diamondback terrapins at two core monitoring sites within Wellfleet Bay during 2019 using the spatial capture-recapture survey protocol refined from 2018 to generate estimates of detectability, density, space use and sex structure using a hierarchical spatial capture-recapture (SCR) model. Detection was positively associated with survey effort at both sites. Detection was also influenced by day of season, tide cycle, the time of tide, survey time relative to the tide, cloud cover and windspeed. Results suggest an estimated density of 9 and 59 individuals per hectare and space use parameter of 309 m and 107m within the two sites. Sex structure was female-biased, with a sex ratio of 0.34 and 0.18 males within both sites.

In Chapter 3, I address the challenges of low sample size and spatially restricted study areas by developing the first standardized visual head count protocol for diamondback terrapins. I analyzed the repeated point count data generated from the visual head count protocol at 38 sites within Wellfleet Bay during season one to generate estimates of detectability and spatially explicit local abundance using a hierarchical N-mixture model. Detection was positively associated with air temperature and negatively associated with wind speed. Estimates of local abundance were generated for each site during each survey period. Local abundance declined over the sampling season but was higher on average in sheltered sites compared to sites that were exposed to the larger bay. Visual head count surveys increased the number of sample sites from two (using SCR), to 38 sites. Moreover, the seasonal change coupled with spatial differences between sheltered and exposed sites in local abundance provides evidence of spatiotemporal variation in local abundance within Wellfleet Bay.

In chapter 4, I build on the findings in Chapter 2 to investigate suitable nesting habitat as a hypothesized driver of spatiotemporal variation in near-shore distribution and abundance. I analyzed northern diamondback terrapin nest location data within Wellfleet Bay using nest data collected by Mass Audubon to 1) to predict areas of suitable nesting habitat and create an index of nest suitability (NSI) covariate using a scale selection Resource Selection Function and NCLD landcover data and 2) test the effect of NSI and day of season on spatiotemporal patterns of relative abundance using a generalized linear mixed model (GLMM). Results suggest selection for saltmarsh and beach habitat and avoidance of developed areas and open water. Expected relative abundance was influenced by the interaction between NSI and day of season with greater expected relative abundance within high NSI areas during the nesting season (2.30 individuals) compared to areas of low NSI (1.99 individuals). My results suggest proximity to nesting habitat has some influence on spatiotemporal variation in relative abundance and should be further investigated in true abundance models.

Finally, in chapter 5, I leverage the head count survey methodology developed in Chapter 2 and predicted suitable nest habitat generated in Chapter 3 to test hypotheses on species-habitat associations and how these relationships change through time within a season to predict spatiotemporal distribution and abundance of diamondback terrapins in Wellfleet Bay. I increased the visual survey intensity at each of the 38 sites from once per month to approximately every 10 days for 151 days to achieve a finer scale temporal resolution. I was unable to produce estimates of true abundance due to evidence of non-independence in the count data and thus had to change my analytical approach to occupancy modelling. I found suitable nest habitat to have no effect on occurrence patterns but found saltmarsh habitat and sheltered sites to positively influence occurrence. I also found a strong temporal influence, where occupancy probability at any given site within the bay was lowest in May and October and highest in July. I took the independent spatial and temporal effects one step further and created a 2-dimensional spatiotemporal predictive surface. From here, occupancy can be predicted at any point in space and in any point in time, which provides a more informative lens on how diamondback terrapin occupancy shifts through space and time in Wellfleet Bay.

In summary, my dissertation provides much needed standardized, methodological improvements to population monitoring of a species of high conservation concern range-wide. The SCR and visual head count protocols generated the first density estimates of local diamondback terrapin populations in Wellfleet Bay. The strategies employed to capture individuals or count surfaced heads are not restricted to what was used in my research so long as explicit survey effort and spatial encounter information are standardized and recorded, which allow flexibility of these methods to the mosaic of saltmarsh systems diamondback terrapins inhabit. Although more intensive and spatially restrictive, SCR methods provide valuable information on diamondback terrapin demographics including sex and age structure, as well as movement within the study area. Visual head count methods cannot include demographic data, but are efficient and able to reach larger spatial extents. Moreover, visual head count methods provide a natural framework for investigating habitat associations and distribution patterns that can be easily tracked through space and time. The methods and analytical frameworks developed here contribute significantly toward closing the knowledge gap on range-wide population status assessments, a primary goal in diamondback terrapin conservation and management.

DOI

https://doi.org/10.7275/32490975

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.

Share

COinS