AbstractThe life history characteristics of sea turtles - including their long life spans, late age-at-maturity, migratory behaviour and widespread habitat usage - impede population monitoring of each demographic life stage within the population. Monitoring sea turtle populations has mostly focused on the adult females at the nesting beach and less commonly on juvenile and sub-adult turtles in their foraging habitats. Despite the limitations of only measuring one demographic state, the greater and more abundant and widespread information on nesting turtles make it an important source of information for management and conservation. Despite this dominance, survey errors, power and confidence of abundance estimates and trend detection in nesting turtle populations have been given little focus until recently.
In this study, I investigated efficiency of sampling programs for monitoring nesting sea turtles where conducting a full-time census is not feasible. Sampling regimes for nesting turtles ranged from five to 100% data coverage and were investigated for five species of turtles. I examined the impact of sampling errors and environmental correlates on trend detection of the nesting populations for 51 populations of sea turtles, comprising five species. These results were then applied to a nesting population with unknown seasonality and abundance, and identified it as one of the world's largest populations of nesting flatback sea turtles.
The high inter-annual variability in nesting turtle abundance means that several decades are often required to detect trends for raw data. This thesis shows that climatic and environmental predictors can be used effectively to predict nesting abundance and dramatically increase the power in trend detection. A model for estimating annual abundance for green turtles using the year, previous year's nesting numbers, Southern Oscillation Index, and specific humidity reduced the number of years needed to detect trends with the same power by 43%. In addition, sampling errors associated with low survey effort of only two weeks had a low impact on the power of detecting trends in the population, indicating that for abundance estimates, annual accuracy in abundance estimates is overshadowed by the large inter-annual variation in abundance. This thesis presents a comprehensive investigation of survey errors for sampling nesting turtle populations.
|Date of Award||Dec 2010|
|Supervisor||Colin Limpus (Supervisor) & Milani Chaloupka (Supervisor)|
Sampling efficiency for monitoring nesting sea turtle populations
Whiting, A. U. (Author). Dec 2010
Student thesis: Doctor of Philosophy (PhD) - CDU