Spatial ecology and conservation of sea turtles in coastal foraging habitat

Spatial ecology investigates mechanisms in nature by examining spatial patterns. Developing our knowledge of spatial ecology will improve our approaches to the conservation of threatened species. Our understanding of spatial ecology is limited for marine species such as sea turtles, due to the complexity and methodological challenges involved in the investigation at-sea behaviour. Most sea turtle species are conservation-dependent due to historical and current anthropogenic threats, but a lack of ecological knowledge in the marine environment may hinder or prevent appropriate actions being taken by conservation practitioners. Important knowledge gaps exist in our understanding of spatial and temporal movement of sea turtles in coastal foraging habitats where anthropogenic threats are high. These turtles may be relocated from their home habitats due to injury (e.g. from fisheries and boat strikes), following disasters (e.g. oil spills or extreme weather events), or following habitat loss (e.g. port expansion and dredging); thus, it is critical to understand the degree of fidelity and homing ability of turtles that have been displaced. For instance, if displaced turtles return to an oil spill area too soon after displacement, alternative conservation actions might be required. Additionally, if a turtle returns to its home habitat after displacement, as past studies have indicated, another question arises: how do the navigation mechanisms of sea turtles work? Current hypotheses theorise that sea turtles use geomagnetic cues for long-distance migration, but that they are likely to use non-geomagnetic cues during the last part of their migration. Details of their fine-scale navigation ability and potential cues are lacking. Fidelity to foraging habitat has been indicated by previous studies but key questions remain; (a) What is the extent of sea turtle foraging habitats? (b) How long do they remain in such habitat? and (c) What factors affect their faithfulness to specific foraging habitat? Answering these questions will improve credibility of home range analysis and its applications to conservation planning or evaluation such as determining the effectiveness of Go Slow Zones in Moreton Bay. Moreton Bay is a significant foraging area for green and loggerhead turtles in eastern Australia but the risk of boat strikes is high because of extensive recreational and commercial vessel traffic. Management systems, such as Go Slow Zones, are in place in some shallow areas but vessel collisions still occur more frequently in Moreton Bay than elsewhere along the Queensland coast. Therefore there is a need for assessing whether current regulation is providing adequate protection to sea turtles against vessel collisions in Moreton Bay. My thesis addresses these knowledge gaps and aims to advance our knowledge of ecology and conservation of sea turtles related to their spatial and temporal use of coastal foraging habitats, with particular focus on the Queensland region. Fastloc GPS (FGPS) is a powerful tool for investigation of fine-scale animal spatio-temporal ecology. Satellite-linked FGPS tags provide researchers with almost unlimited temporal and spatial range to monitor animal movements, and give more accurate and larger quantities of locations than earlier methods (e.g. platform transmitter terminals). These large and detailed locational data make understanding certain aspects of turtle ecology possible, whilst also enabling delineation of accurate areas for protection. However, it remains important to identify and remove locations with high error because some location fixes are much less accurate than others. I use FGPS tags as my primary tool to track turtle movements, and therefore need to handle FGPS estimates with high error prior to any ecological and conservation-based analyses. I began by investigating potential methods to screen FGPS data (Chapter 2). Increasing the number of source satellites required for a valid fix is a simple filter method but it comes at the cost of great data loss. Using data sets acquired from loggerhead turtles (Caretta caretta), I explored an alternative filtering approach, based on speed between successive locations, angles created by three consecutive locations, manufacturer's quality index, and number of satellites used for location calculation. The performance of the proposed filter method was evaluated by conducting terrestrial mobile tests. When my filter method was used, the linear error (mean ± SD) of Fastloc GPS data decreased from 2,645.5 ± 29,458.2 m (n = 1,328) to 47.1 ± 61.0 m (n = 1,246) while retaining more than 94% of data. My filter method also led to more accurate home range estimates than the simple filter method. This advance in processing satellite-derived data delivers an improved ability to analyse fine-scale animal movement. I went on to apply the filtering technique to my satellite telemetry data prior to subsequent analyses. In chapter 3, I investigated whether highly mobile sea turtles can be expected to remain at a new location after they were displaced. I addressed this question for sea turtles at foraging grounds along the coast of north-eastern Australia. I analysed 113 tracks comprising four species (Chelonia mydas, Caretta caretta, Lepidochelys olivacea, Eretmochelys imbricate) fitted with satellite-linked devices. Turtles released at their original "home" areas all remained there (n = 54). Among displaced turtles (released away from their original area, n = 59), the large majority travelled back to their respective home areas (n = 52) or near home (n = 4). Homing turtles travelled faster and adopted straighter routes in cooler water, and travelled faster by day than by night. My results showed that displacement up to 117.4 km and captivity up to 514 days did not disrupt homing ability nor did it diminish fidelity to the home area. However, for homing turtles I infer energetic costs and heightened risk in unfamiliar coastal waters. Confirmed homing suggests that moving individuals away from danger might offer short-term benefit (e.g. rescue from an oil spill) but moving turtles to a new foraging area is unlikely to succeed as a long-term conservation strategy. Priority must rather be placed on protecting their original habitat. As confirmed in Chapter 3 and previous studies, sea turtles have an exceptional ability to navigate accurately between known habitats as well as from unknown areas back to familiar habitat. In Chapter 4, I examined the turning and orientation behaviour of 29 displaced sea turtles of two species (Chelonia mydas, Caretta caretta), tracked en route back to their foraging habitats in eastern Australia. I found that sea turtles tended to alternate stationary and travelling phases during their trip home. Orientation corrections predominantly occurred immediately after a stationary phase and after sunrise. This is the first study to demonstrate time-restricted orientation by sea turtles and provided a new insight into their sophisticated navigational abilities. My remaining data chapters (5, 6) focused on turtles' behaviour in their foraging habitats, and resultant applications for conservation planning. In Chapter 5, I tracked adult green and loggerhead turtles foraging in the coastal waters of eastern Australia, objectively quantified home range size and site fidelity, and then examined how their spatial selections were affected by ecologically meaningful variables such as season, extreme weather events (tropical cyclones and extreme rainfall), habitat location and sex. Many individual turtles were observed multiple times over extended periods using satellite telemetry (PTT, FGPS or both) and mark-recapture methods. Evidence from these multiple observations inferred that many turtles maintained high fidelity to their coastal foraging habitats for long periods - up to 20 years. Within these long term foraging habitat areas defined by my analysis, turtles generally shifted their main foraging areas on a seasonal basis. These characteristics of sea turtles emphasise the importance of conserving areas according to their space use, with careful consideration given to identifying temporal trends in habitat selection. I also identified a geographical advantage of two sites in eastern Australia (eastern Moreton Bay and eastern Port Curtis) as foraging habitats for sea turtles with relation to extreme weather events such as tropical cyclones and extreme rainfall. These important foraging habitats would benefit from prioritised conservation planning and management actions. My findings have direct relevance to conservation managers for planning, or revision, of designated conservation habitat such as Marine Protected Areas or restricted area zones, to protect these threatened species from increasing human activities at their foraging habitats in Australia and other regions. Finally, I investigated whether existing Go Slow Zones are providing adequate protection against boat strike for sea turtles foraging in Moreton Bay. To do so, I examined space use of green and loggerhead turtles in relation to the Go Slow Zones and water depth (Chapter 6). I found that most of the habitats used by my tracked turtles were in shallow water, and up to 55% of their habitats were included within the Go Slow Zones in eastern Moreton Bay. However, turtles are not protected from vessel collisions in the deeper zones (water depth ≥ 5 m), which lie adjacent to the Go Slow Zones, or in other shallow water zones in Moreton Bay. In particular, little or no protection is given to sea turtles in southern, western and northern Moreton Bay. By designating all shallow areas in Moreton Bay as Go Slow Zones, approximately 50% or more of the Bay's turtle habitats would become protected from vessel collisions. Additionally, my data indicate that shallow zones plus a 1.2 km, 2.4 km, or 3.6 km buffer would protect ≥80%, ≥90% or ≥95% respectively, of habitats used by both species because they cover the deeper zones adjacent to the shallow zones. The results of this study will be highly informative for conservation managers when revising the current Go Slow Zones for improved management of these threatened sea turtle populations. The advanced technology and analytical tools I adopted in this thesis enabled me to overcome the difficulties associated with investigating sea turtle movements, and consequently to improve our understanding of their relationship with environmental variables. My approach has applications for investigating spatial ecology of other animals, including other populations of sea turtles. I concluded this study by discussing my key findings related to the behaviour of foraging sea turtles, highlighting conservation benefits that can be potentially derived from incorporating ecological knowledge into planning. I also suggest specific priorities for future research to enhance our knowledge of the spatial ecology of sea turtles, and consequently our ability to conserve these threatened marine reptiles which are necessary for healthy ecosystems