Thesis (M.S.)--University of North Carolina Wilmington, 2007. / Includes bibliographical references (leaves: 59-61)

Thesis (Ph. D.)--Massachusetts Institute of Technology and Woods Hole Oceanographic Institution, 1996. / Includes bibliographical references.

Thesis (Ph. D.)--Duke University, 2008.

Thesis (Ph. D.)--University of Hawaii at Manoa, 2003. / Includes bibliographical references (leaves 182-190).

No description available.

The bottlenose dolphin Tursiops truncatus is an extremely well-studied species. We have an extensive knowledge of certain aspects of their vocal behaviour, particularly from captive contexts. Bottlenose dolphins produce a rich tapestry of vocalisations, however, which have historically received minimal attention. Resident groups of bottlenose dolphins frequent the waterways of Fiordland in southwest New Zealand. These deep, sheltered fiords are ideally suited for acoustic studies. This thesis presents the first detailed study of bottlenose dolphin acoustics in New Zealand. Both narrowband and broadband systems were used to record the vocalisations of two resident groups. Effort was distributed evenly over three years for both Doubtful Sound and Milford Sound. From 875 recordings, I proposed a repertoire of 15 discrete calls. These categories were subsequently compared using parameters measured from almost 2000 individual vocalisations. Various multivariate techniques revealed some redundancy in the proposed repertoire, and it was subsequently reduced to 12 calls. The 12 call repertoire was compared between the potentially interbreeding populations of Doubtful Sound and Milford Sound. Fiord-specificity was revealed for many of the calls, particularly the sequenced calls and whistles. These differences suggest bottlenose dolphins use dialects, in keeping with studies of killer whales and sperm whales. As Fiordland dolphins are out of sight for 90% of the time, acoustic techniques allow inference in to subsurface behaviour. I investigated sequential relationships among sounds and between sounds and behaviours. Many calls were strongly implicated in social interactions. The vocalisations ratchet, orca and the sequenced calls were associated with periods of conflict. A number of the click-based calls were linked to diving and presumed foraging events. Inference on the functional significance of sounds allowed an interpretation of habitat use. This appears to be the first study relating the entire vocal repertoire of a cetacean population to a complete home range. Areas important for socialising, foraging and resting are proposed. Local management decisions may be well served by this information. This study uses benign techniques to build on previous research in Fiordland, and adds a new dimension to the study of these populations.

Interactions between humans and dusky dolphins in the coastal waters of New Zealand are increasing. My research focused on tourism interactions, with Kaikoura as the study site; and, on habitat use in an active aquaculture area, with Admiralty Bay as the study site. In Kaikoura, companies engaged in commercial cetacean tourism (For Hire Company) have permits issued by the New Zealand’s Department of Conservation, allowing them to take paying customers out to view and swim with wild dusky dolphins. During summer and fall of 2005, I assessed the effectiveness of a voluntary ‘rest period’ established to give time free of humans to the dolphins. I used a theodolite to track the movements of large groups of dusky dolphins and recorded the arrival, departure and behaviors of all vessels approaching within 400 m of the group. The ‘rest period’ resulted in a reduction of vessel visits compared to non-rest periods, yet one For Hire Company and private recreational vessels continued to visit dusky dolphin groups during this time. To increase compliance with the voluntary regulation, more education is needed targeting private recreational vessels. Weekend traffic was higher compared to weekday traffic, during both rest and non-rest periods; a large increase occurred in weekend non-commercial vessel traffic. Swimming with calves is prohibited by New Zealand’s Marine Mammal Protection Regulations of 1992, yet 71.4 percent of the swim attempts I observed on-board For Hire Company tours were conducted with groups containing calves. More should be done to reduce the number of swims conducted with groups of dusky dolphins containing calves. In winter of 2005, I used hourly theodolite scans to record the number of dusky dolphin groups using Admiralty Bay, a different near-shore environment with less tourism than off Kaikoura, and with near-shore mussel farms. Groups of dusky dolphins were observed in Admiralty Bay using the full extent of the bay. This re-enforces previous findings that Admiralty Bay is an important winter foraging ground for dusky dolphins, and further aquaculture development in the bay would remove available foraging habitat.

Abstract The aim of this study investigated the metal bioaccumulation in the cetaceans from the Taiwanese coastal waters. The differences of metal concentrations in the small dolphins between species, sex and sampling location were examined. The results were compared to the data all over the world to understand the pollution status in the western Pacific. Accordingly specimens-bank was established. Furthermore, the correlations between the size/age and metal concentrations were investigated in Stenella attenuata, S. longirostris and Neophocaena phocaenoides. This study cooperated with the Cetacean Research Laboratory leading by the Prof. L. S. Chou, Dept. of Zoology, National Taiwan University. The samples included 8 species of 83 individuals accidentally caught and stranded around Taiwan in 1994-1995. The 8 species included Delphinus delphin, Grampus grieus, Lagenodelphis hoseei, Stenella attenuata, S. coeruleoalba, S. longirostris, Steno bredanensis, Tursiops truncatus. In addition, one Kogia simus stranded in 1998 and two Neophocaena phocaenoides accidentaly caught in 2000 were used to examine the variation of metal bioaccumulation in the muscles at different parts of body. The tissues of muscle, liver and lung were used for analysis of Ag, As, Cd , Co, Cu, Fe, total Hg, organic Hg, Mn, Ni, Se and Zn. The flame atomic absorption spectrophotometry (AAS) was used in analysis of Cu, Fe, Mn, Ni and Zn. The cold vapor AAS and graphite furnace AAS were used for determination of Hg and Cd, respectively. ICP-MASS were used in measurements of Ag, As, Co and Se. The differences of metal bioaccumulation in muscles of different species were element ally depend. As, Fe, total Hg, organic Hg, Se and Zn presented significant species differences. Highest Fe (147 mg/Kg wet wt), total Hg (16.9), organic Hg (5.65), and Se (4.62) were found in Lagenodelpnis hosei, whereas highest As (2.92) and Zn (24.1) were found in Grampus griseus and Tursiops truncatus. However, lowest As (0.29), and Se (0.76) were found in Tursiops truncatus. Lowest Fe (90), organic Hg (0.93), total Hg (1.0) and Zn (10.16) were found in Delphinus delphis, Delphinus delphis, Steno bredanensis, and Stenella longirostris. No species differences were found in the muscle concentrations of Ag, Cu, Cd, Co, Mn and Ni. Except organic Hg found highest in muscle. The concentrations of rest elements were found highest in liver and lowest in muscle. There was no significant difference of metal bioaccumulation found in male and female Stenella attenuata. The correlations between the size/age and metal concentrations in muscle were found in Stenella attenuata, S. longirostris and Neophocaena phocaenoides. The relationships were significant in total Hg, oranic Hg and Se of Stenella attenuata, whereas in orangic Hg, Se and Zn of Stenella longirostris and Neophocaena phocaenoides. The metal concentrations found in this study were mostly equal to or lower than those records all over the world. Furthermore, by comparison the total Hg of Stenella attenuata, S. coeruleoalba and Tursiops truncatus with those data found in elsewhere. The total Hg levels in this study were slightly higher than that of eastern Pacific, but much lower than that of Mediterranean of Italy and France, Atlantic and Japan. These findings suggest that the metal concentrations found in this study representing the baseline metal concentration of western Pacific. Based on this, no specific sight of pollution was found in the coastal waters off Taiwan.

ABSTRACT The conservation of cetaceans in Pescadores was originated in 1990. However, the conflict between the fishery loss and cetacean protection keeps growing. The fishery loss caused by cetaceans are mainly: ¡§steal fish¡¨ and ¡¨frighten fish group¡¨, and the others like ¡¨damage fishing gear¡¨ and ¡¨interfere fishing operation¡¨, and etc.By the understanding of dolphins behavior, this study proposed acoustical deterrent methods; 1.harassment:look for echolocation system frequency range, and broadcast disarrange signals to produce an illusion, and prevent dolphins from locating the target. 2. threat: broadcast sounds of its predators which is killer whales to scare dolphins from approaching the fishing vessels. 3. warning: loud noise or alert sounds of dolphins. The circuits of generating above sounds are designed, test and modified after the field test . Underwater speaker was used to broadcast sounds of 10 kHz , 20 kHz , killer whales sound and distress call of dolphins. The test results showed these dolphins avoided the sound source, especially during killer whales sound and stress call of dolphins. They were effective to deter dolphins to reduce stealing fish on the sea. More effective deterrence device should be developed through this study to reduce the loss of fisherman, and achieve the cetacean conservation goal.

The study of dolphin cognition involves intensive research of animal vocal- izations recorded in the field. In this dissertation I address the automated analysis of audible dolphin communication. I propose a system called the signal imager that automatically discovers patterns in dolphin signals. These patterns are invariant to frequency shifts and time warping transformations. The discovery algorithm is based on feature learning and unsupervised time series segmentation using hidden Markov models. Researchers can inspect the patterns visually and interactively run com- parative statistics between the distribution of dolphin signals in different behavioral contexts. The required statistics for the comparison describe dolphin communication as a combination of the following models: a bag-of-words model, an n-gram model and an algorithm to learn a set of regular expressions. Furthermore, the system can use the patterns to automatically tag dolphin signals with behavior annotations. My results indicate that the signal imager provides meaningful patterns to the marine biologist and that the comparative statistics are aligned with the biologists’ domain knowledge.