Department of Fisheries Oceanography PhD Dissertation Defense by Megan Winton. Titled: Integrating telemetry data to improve abundance estimates and management advice for highly migratory marine species
When: Monday,
April 1, 2024
1:00 PM
-
2:00 PM
Where: > See description for location
Description: SMAST East Rooms 101-103
836 S. Rodney French Blvd, New Bedford
and via Zoom
Advisor:
Gavin Fay
Committee Members:
Diego Bernal, Steven Cadrin, Heather Haas, Gregory Skomal
Abstract
Telemetry-integrated models have long been heralded as a potential tool to overcome data gaps and improve distribution and abundance estimates for highly migratory marine species, which remain unavailable for many populations. However, conceptual and practical (i.e. computational) difficulties have limited their development and implementation to date. In this dissertation, I extend existing spatial modeling approaches for line-transect and capture-recapture survey data to allow for the direct integration of telemetry data into population models, with the aim of improving estimates of abundance and other population parameters. To do so, in Chapters I and II, I develop hierarchical models for satellite and acoustic telemetry data that are compatible with spatial approaches for estimating abundance from line transect and capture-recapture data. In Chapter III, I develop and evaluate the predictive performance of a framework for inferring the distribution and relative density of loggerhead sea turtles (Caretta caretta) from line transect data that integrated satellite telemetry and incidental catch data. Finally, in Chapter IV, I develop and test an acoustic telemetry-integrated spatial framework for estimating the abundance and dynamics of white sharks (Carcharodon carcharias) from capture-recapture surveys conducted at seasonal aggregation sites.
Conceptually, the hierarchical spatial structure of the models presented in each chapter provide a logically consistent way of approaching disparate datasets, which makes it possible to translate different data types across ecological subfields and accommodate several sources of bias common to both survey types as well as tagging studies. In so doing, this dissertation provides a straightforward, intuitive framework for integrating telemetry data into abundance and distribution estimates from standardized survey data, which has the potential to improve the reliability of science-based advice for highly migratory marine species, as illustrated by the results of each chapter. The results of Chapter I indicate that the model developed for satellite tagging data outperforms conventional estimators when the number of tag transmissions changes over time, a common source of bias in satellite telemetry studies that is rarely addressed. When applied to data collected from 271 satellite tagged loggerhead sea turtles by six different research programs, the new model suggests that tagged loggerheads inhabit the continental shelf along the U.S. Atlantic from Florida to North Carolina year-round but also highlight the importance of summer foraging areas on the mid-Atlantic shelf. In Chapter II, the model developed to estimate individual centers of activity for acoustic telemetry data successfully accounted for variation in receiver detection ranges, revealing fine-scale movements that were not apparent when conventional estimators assuming a constant detection range were applied. The integrated model in Chapter III jointly estimated the distribution of loggerhead sea turtles in the US mid-Atlantic from aerial survey, satellite telemetry, and incidental catch records in relation to sea surface temperature (SST) and depth and better predicted the distribution of the species in comparison with models based on individual data sources. Finally, the results of Chapter IV illustrate that conventional capture-recapture models do not adequately represent the migratory behavior of white sharks and can produce biased estimates of abundance that would be misleading if used as the basis for management advice. Because it directly links changes in abundance over time to the demographic processes underpinning them, the model described provides a more mechanistic understanding of the dynamics of white shark aggregations and improves the applied relevance of the results for the conservation and management of the species.
By providing a unified spatial modeling framework for electronic tagging and structured survey data that is computationally feasible and efficient to fit, the work presented in this dissertation will hopefully make these methods more accessible to ecologists and allow for their more widespread adoption. While much work remains to be done, integrated approaches are likely to become increasingly important for answering conservation and management-related questions as the availability of new data types grows, which will be especially valuable when dealing with rare, elusive highly migratory marine species, many of which are threatened or endangered. Though the focus here was to determine how various types of telemetry data could be integrated into models for standardized survey data and to evaluate how their integration impacted distribution and abundance estimates...
836 S. Rodney French Blvd, New Bedford
and via Zoom
Advisor:
Gavin Fay
Committee Members:
Diego Bernal, Steven Cadrin, Heather Haas, Gregory Skomal
Abstract
Telemetry-integrated models have long been heralded as a potential tool to overcome data gaps and improve distribution and abundance estimates for highly migratory marine species, which remain unavailable for many populations. However, conceptual and practical (i.e. computational) difficulties have limited their development and implementation to date. In this dissertation, I extend existing spatial modeling approaches for line-transect and capture-recapture survey data to allow for the direct integration of telemetry data into population models, with the aim of improving estimates of abundance and other population parameters. To do so, in Chapters I and II, I develop hierarchical models for satellite and acoustic telemetry data that are compatible with spatial approaches for estimating abundance from line transect and capture-recapture data. In Chapter III, I develop and evaluate the predictive performance of a framework for inferring the distribution and relative density of loggerhead sea turtles (Caretta caretta) from line transect data that integrated satellite telemetry and incidental catch data. Finally, in Chapter IV, I develop and test an acoustic telemetry-integrated spatial framework for estimating the abundance and dynamics of white sharks (Carcharodon carcharias) from capture-recapture surveys conducted at seasonal aggregation sites.
Conceptually, the hierarchical spatial structure of the models presented in each chapter provide a logically consistent way of approaching disparate datasets, which makes it possible to translate different data types across ecological subfields and accommodate several sources of bias common to both survey types as well as tagging studies. In so doing, this dissertation provides a straightforward, intuitive framework for integrating telemetry data into abundance and distribution estimates from standardized survey data, which has the potential to improve the reliability of science-based advice for highly migratory marine species, as illustrated by the results of each chapter. The results of Chapter I indicate that the model developed for satellite tagging data outperforms conventional estimators when the number of tag transmissions changes over time, a common source of bias in satellite telemetry studies that is rarely addressed. When applied to data collected from 271 satellite tagged loggerhead sea turtles by six different research programs, the new model suggests that tagged loggerheads inhabit the continental shelf along the U.S. Atlantic from Florida to North Carolina year-round but also highlight the importance of summer foraging areas on the mid-Atlantic shelf. In Chapter II, the model developed to estimate individual centers of activity for acoustic telemetry data successfully accounted for variation in receiver detection ranges, revealing fine-scale movements that were not apparent when conventional estimators assuming a constant detection range were applied. The integrated model in Chapter III jointly estimated the distribution of loggerhead sea turtles in the US mid-Atlantic from aerial survey, satellite telemetry, and incidental catch records in relation to sea surface temperature (SST) and depth and better predicted the distribution of the species in comparison with models based on individual data sources. Finally, the results of Chapter IV illustrate that conventional capture-recapture models do not adequately represent the migratory behavior of white sharks and can produce biased estimates of abundance that would be misleading if used as the basis for management advice. Because it directly links changes in abundance over time to the demographic processes underpinning them, the model described provides a more mechanistic understanding of the dynamics of white shark aggregations and improves the applied relevance of the results for the conservation and management of the species.
By providing a unified spatial modeling framework for electronic tagging and structured survey data that is computationally feasible and efficient to fit, the work presented in this dissertation will hopefully make these methods more accessible to ecologists and allow for their more widespread adoption. While much work remains to be done, integrated approaches are likely to become increasingly important for answering conservation and management-related questions as the availability of new data types grows, which will be especially valuable when dealing with rare, elusive highly migratory marine species, many of which are threatened or endangered. Though the focus here was to determine how various types of telemetry data could be integrated into models for standardized survey data and to evaluate how their integration impacted distribution and abundance estimates...
Contact: > See Description for contact information
Topical Areas: SMAST, Students, Graduate