Quantitative Tools for Monitoring Strategy Evaluation and Assessment of Sea Turtle Populations Public Deposited

http://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/np193d613

Descriptions

Attribute NameValues
Creator
Abstract or Summary
  • Green sea turtles, Chelonia mydas, have endangered and threatened populations globally, but several populations show signs of population recovery. In Hawaii, nesting female green turtles have increased 5.7% year⁻¹ since 1973, but wide fluctuations in census counts of nesting females make recovery diagnosis difficult. For effective management planning, it is critical to have the best information possible on vital rates, and to determine the best tools and practices for incorporating vital rate information, particularly variability, into population models to assess population size and status. Process and observation errors, compounded by late maturity, obscure the relationship between trends on the nesting beach and the entire population. Using sea turtle nesting beach surveys as a population index for assessment is problematic, yet often pragmatic because this is the only population index that is easily accessible. It is advantageous to use a modelling approach that estimates interannual variability in life history traits, accounts for uncertainty from individual-level variability, and allows for population dynamics to emerge from individual behaviors. To this end, I analyzed a long-term data set of marked green sea turtles to determine the degree of temporal variability in key life history traits. From this analysis, I built an agent-based model (ABM) to form the basis of a new assessment tool -- Monitoring Strategy Evaluation. In Chapter 2, I evaluated annual changes in demographic indicators (DIs) of 3,677 nesting green turtles from a 38-year tagging program in the Hawaiian Islands to determine if key life history traits are changing over time and in response to nester density. I used linear mixed models and multistate open robust design models to estimate several DIs and correlated them with nesting female counts. Mean nester carapace length and breeding probability were highly variable over time, suggesting shifts in age structure that could be due to recruitment. The top-ranked model predicted constant female survival over time. A significant positive relationship between the nesting population and breeding probability was evident, and breeding probability shows promise as an indicator of population recovery. This work contributes to a growing set of studies evaluating sea turtle demography for temporal variability and is the first for Hawaiian green turtles. In Chapter 3, I develop the Green Sea Turtle Agent-Based Model (GSTABM) to evaluate how recovery processes differ across disturbance types. The GSTABM incorporates individually variable age-at-maturity, clutch frequency and clutch size, annually variable breeding probability, environmental stochasticity and density dependence in hatchling production. The GSTABM simulates the process of population impact and recovery and the monitoring process, with observation error, on the nesting beach. The GSTABM captures the emergent patterns of interannual nesting variation, nester recruitment, and realistic population growth rates. Changes in survival rates of the nearshore age-stage classes directly affected adult and nester abundance, population growth rate and nester recruitment more than any of the other input parameters. In simulating 100 years of recovery, experimentally disturbed populations began to increase but did not fully return to pre-disturbance levels in adult and nester abundance, population growth or nester recruitment. In simulations with different levels of monitoring effort, adult abundance was poorly estimated, was influenced by population trajectory and disturbance type, and showed marginal improvements in accuracy with increased detection probability. Estimating recruitment showed improvements with increasing detection levels. In the GSTABM, variability in the nesting beach does not mirror variability in adult abundance. The GSTABM is an important tool to determine relationships with monitoring, population assessment, and the underlying biological processes driving changes in the population, and especially, changes on the nesting beach. In Chapter 4, I develop a new simulation-based tool: Monitoring Strategy Evaluation (MoSE) to determine which data source yields the most useful information for population assessments. The MoSE has three main components: the simulated "true" operating, observation, and estimation models. To explore this first use of MoSE, I apply different treatments of disturbance, sampling, and detection to the virtual "true" population, and then sample the nests or nesters, with observation error, to test if the observation "data" accurately diagnose population status indicators. Based on the observed data, I estimated adult abundance, nester recruitment, and population trend and compare them to the known values. I tested the accuracy of the estimated abundance when annually varying inputs of breeding probability, detection and clutch frequency were used instead of constants. I also explored the improvement of trend accuracy with increased study duration. Disturbance type and severity can have important and persistent effects on the accuracy of population assessments drawn from monitoring rookeries. Accuracy in abundance estimates may be most improved by avoiding clutch frequency bias in sampling and including annually varying inputs in the estimation model. Accuracy of nester recruitment may be most improved by increasing detection level and avoiding age-bias in sampling. The accuracy of estimating population trend is most influenced by the underlying population trajectory, disturbance type and disturbance severity. At least 10 years of monitoring data are necessary to accurately estimate population trend, and longer if juvenile age classes were disturbed and trend estimates occur during the recovery phase. The MoSE is an important tool for sea turtle biologists and conservation managers and allows biologists to make informed decisions regarding the best monitoring strategies to employ for sea turtles. This modeling framework is designed to provide an evaluation of monitoring program effectiveness to assist in planning future programs for sea turtles. Altogether, my research suggests certain life history traits of green sea turtles have important temporal variation that should be accounted for in population models, understanding the relationships between nesting and the total population is essential, and basing population assessments from nesting beach data alone is likely to result in incorrect or biased estimates of status indicators. The quantitative tools employed here can be applied to other sea turtle populations and will improve monitoring, and result in better estimates of current population trends and enhance conservation for all species of sea turtles.
Resource Type
Date Available
Date Copyright
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Advisor
Committee Member
Academic Affiliation
Non-Academic Affiliation
Keyword
Subject
Rights Statement
Peer Reviewed
Language
Replaces
Additional Information
  • description.provenance : Made available in DSpace on 2016-06-13T22:53:09Z (GMT). No. of bitstreams: 2 license_rdf: 1223 bytes, checksum: d127a3413712d6c6e962d5d436c463fc (MD5) PiacenzaSusanEH2016.pdf: 10048435 bytes, checksum: 9c2f26ffa7cc8d5ca50f88d8897c88a5 (MD5) Previous issue date: 2016-06-01
  • description.provenance : Submitted by Susan Piacenza` (susan.hilber@oregonstate.edu) on 2016-06-08T20:58:17Z No. of bitstreams: 2 license_rdf: 1223 bytes, checksum: d127a3413712d6c6e962d5d436c463fc (MD5) PiacenzaSusanEH2016.pdf: 10048435 bytes, checksum: 9c2f26ffa7cc8d5ca50f88d8897c88a5 (MD5)
  • description.provenance : Approved for entry into archive by Julie Kurtz(julie.kurtz@oregonstate.edu) on 2016-06-13T16:21:36Z (GMT) No. of bitstreams: 2 license_rdf: 1223 bytes, checksum: d127a3413712d6c6e962d5d436c463fc (MD5) PiacenzaSusanEH2016.pdf: 10048435 bytes, checksum: 9c2f26ffa7cc8d5ca50f88d8897c88a5 (MD5)
  • description.provenance : Approved for entry into archive by Laura Wilson(laura.wilson@oregonstate.edu) on 2016-06-13T22:53:09Z (GMT) No. of bitstreams: 2 license_rdf: 1223 bytes, checksum: d127a3413712d6c6e962d5d436c463fc (MD5) PiacenzaSusanEH2016.pdf: 10048435 bytes, checksum: 9c2f26ffa7cc8d5ca50f88d8897c88a5 (MD5)

Relationships

In Administrative Set:
Last modified: 08/22/2017

Downloadable Content

Download PDF
Citations:

EndNote | Zotero | Mendeley

Items