Sea Turtle In-Water Study
Project Overview

This study was designed to provide fundamental data for the in-water population of turtles along the southeast coast of the United States, with particular emphasis on measuring abundance. Much of our knowledge of sea turtles has come from data collected on nesting beaches or opportunistically during regulation of fisheries. Through an in-water survey we hoped to provide a better understanding of turtle abundance, spatial variability and population characteristics (e.g., size, sex, life history, genetic composition, health).

The turtle research community has long recognized the need for such data. In 1998 the Turtle Expert Working Group gave voice to the need in calling for further in-water studies. Despite the value of data such a project might produce, the concept of a large-scale in-water survey was regarded as somewhat risky. These concerns were not unfounded. In the past, trawl surveys of similar design caught few turtles. In contrast, recent anecdotal information suggested that numbers of turtles in the water were increasing. To address concern over the possibility of low catches and to assess interactions of turtles with fishermen, the fishery independent sampling design was limited to depths between 15 and 40 feet in order to saturate the area with sampling stations. Additionally a fishery-dependent sampling component using commercial trawlers was added to the project.

Concern over low catch of turtles was quickly dispelled. It appears that the loggerhead turtle population in this study area, as reflected by our data, is much larger than it was in the 1970s and early 1980s. Our catch rates are much higher than those reported for fishery-dependent surveys carried out on commercial shrimp trawlers. Differences in gear and towing speed may account for these higher catch rates, but it appears that loggerheads, at least the juveniles, are indeed more abundant now. Perhaps, this increase in abundance is due, in part, to the mandatory use of turtle excluder devices (TEDs) beginning in 1988 in South Carolina and region-wide in 1990. It may also reflect the rapid growth in nest number for loggerheads on south Florida beaches, in which case, perceptions among shrimp fishermen of an increasing turtle population may be misleading for the northern subpopulation.

Development of a scientifically valid index of abundance for loggerhead turtles was the primary goal of this study. We believe we have been successful in establishing a useful regional index of abundance. The values for the four years of this study range from 0.48 to 0.59 loggerhead turtles per 30.5m-net-hour. Although a majority of our stations (75%) produced no turtle catch, the mean catch rate has been remarkably similar each year giving us confidence that the methods and sampling effort have been adequate to establish a reasonable index of abundance.

As we began collecting data, it became clear to us that one simple annual index of abundance may be useful in examining long-term trends in overall turtle population status on a regional basis, but a number of inherent temporal, spatial, and perhaps environmental factors can affect turtle catch rates. We have seen that loggerhead abundance increases at lower latitudes. Inclement weather, for whatever reason, seems to reduce catch rates. These factors need to be recognized when a regional index of abundance is developed.

Over the four years of this study, a disturbing trend of reduced catch rates in the smaller size classes was noted. Examination of annual length frequency plots indicated that growth could account for a shift to larger size classes, but the observed decline in percentages of turtles in the smallest size classes may indicate a recruitment failure, perhaps related to declining nesting activity or an increase in natural mortality rates of smaller juveniles. However, little is known about the process of recruitment from the oceanic to the neritic juvenile stage and therefore numerous alternative explanations are possible. Regardless the reason, this pattern bears continued observation.

It is also clear a mix of individuals from several subpopulations of loggerheads occurs over the range of this study. Given that abundance trends for different subpopulations are possible, it is imperative to segregate turtle catch data by subpopulation. This, however, is no simple matter. Analysis of mitochondrial DNA is ideal for tracing offspring to nesting females and natal beaches; however, there is overlap of at least one haplotype that occurs on nesting beaches throughout the east coast and into the Gulf of Mexico. Therefore, for turtles of that haplotype, one must make assumptions and apply those to a probability analysis when assessing subpopulation trends. These assumptions reduce the robustness of the subpopulation data analysis and leave questions regarding the true population status, particularly for the northern subpopulation. Acknowledging these questions, analysis of DNA data indicated that natal origin for loggerhead turtles captured in this study was 19% (range 14-25%) from the northern subpopulation and 66% (range 60-70%) from the southern subpopulation.

Juvenile turtles exhibited some noteworthy patterns in spatial distribution. We have observed that juveniles may be more closely associated with inlets, perhaps because of more abundant prey, while adults may be more evenly distributed throughout the near-shore coastal area. We have also observed that juveniles, regardless of genetic haplotype, appear to have strong feeding site fidelity as demonstrated by inter-annual tag recaptures that were typically made near the initial tagging and release sites. This feeding site fidelity may underscore the importance of the prey base found in the near shore areas of the Carolinas and Georgia and is probably a critical aspect of the life history of loggerheads for both east coast subpopulations and perhaps others.

This project significantly improves understanding of turtle health. We provide values for blood chemistry of healthy and sick turtles as a reference for individuals charged with caring for sick turtles. Turtles that were deemed “sick” routinely exhibited blood chemistry values consistent with those of stressed or ill animals. A spin-off study that was facilitated by project-provided blood and scute samples indicated that methymercury can be relatively high is sea turtles (Day, 2003). Given that this area of the coast is known to be high in methylation rates of mercury and methyl mercury is common in prey items, the use of local feeding sites may jeopardize the health of migratory juveniles. Though sample sizes in the initial study were small, mercury levels in stranded turtles on SC beaches were found to be significantly higher than those for turtles capture at-sea live. Mercury may impair nervous systems and perhaps alter turtle behavior making those turtle more vulnerable to predators or interactions with man. Analysis performed North Carolina State University confirmed the presence of fibropapilloma in tissue samples of two loggerhead turtles collected in Georgia waters. Additionally 5-13% percent (depending upon year) of the turtles found in this study had evidence of significant trauma from boat propellers or sharks. Although turtle mortality to shrimp trawlers may be greatly reduced now because of the latest advancements in TEDs, it is clear that juveniles and adults will continue to be directly and indirectly affected by man.