Vessels cause considerable disturbance to cetaceans world-wide, with potential long-term impacts to population viability. Here we present a comprehensive review of vessel impacts to cetacean behavior in Australian waters (2003-2015), finding inadequate protections to be in place. The majority of these studies found trends of decreased animal travel and resting behavioral states as well as low compliance to regulations, and they recommended further regulatory action such as greater enforcement or monitoring, or passive management strategies. As a case study, we conducted the first field assessment of vessel compliance with the Wildlife (Marine Mammal) Regulations 2009 in Gippsland Lakes, Australia, and provide the first assessment of the endangered Gippsland Lakes Burrunan dolphin (Tursiops australis) population’s behavioral ecology. Dolphin behavior and vessel regulation compliance data were collected during boat-based surveys of Gippsland Lakes from July 2017 to January 2018, with a total of 22 dolphin group sightings resulting in 477 five-minute point samples. 77% of dolphin sightings involved vessel interactions (within 400 m), and 56 regulation breaches were observed. These breaches were most severe in summer (mean = 4.54 breaches/hour). Vessels were found to alter dolphin behavior before, during, and after interactions and regulation breaches, including increased mating (mate guarding) and milling behavioral states, and increased ‘fish catch’, ‘high leap’ and ‘tail slap’ behavioral events. These behavioral changes may indicate masking of the dolphins' acoustic communication, disturbance of prey, increased dolphin transition behaviors, and/or induced stress and changes to group structure (including increased mate guarding). While our results provide evidence of short-term altered behavior, the potential for long-term effects on population dynamics for this threatened species is high. In the context of reported inadequate cetacean protection Australia-wide, our management recommendations include greater monitoring and enforcement, and the utilisation of adaptive management.

Boat-based surveys were conducted across the Gippsland Lakes, Australia, in daylight hours during Austral Winter (July) and Spring (September) 2017, and Summer (December-January) 2017-2018. Survey periods were chosen to capture seasonal differences and peak (summer) and off-peak tourism periods. Surveys were only conducted on days of low wind (< 15 knots) and good visibility, enabling a 600 metre ‘survey-zone’. A 2C research vessel, 5.7m Ensign 570 powered by 90 hp Mercury engine, was driven along pre-determined line and zig-zag transects across the Gippsland Lakes system at depths ≥2 m, at ≥10 m from shore. Continuous horizon scans were conducted to sight Burrunan dolphins, with two to four researchers on board. The study utilised opportunistic observations of vessel-dolphin interaction rather than controlled experimental vessel approaches. Methodologies were approved by the Victorian State Government’s (Agriculture Victoria) Wildlife and Small Institutions Animal Ethics Committee (WSIAEC 33.14) and RMIT University Animal Ethics Committee, and conducted under Victoria State Government Wildlife Act 1975 Research Permit (Permit number 10008600).

Once dolphins were sighted, observers recorded dolphin group behavioral state and vessel interaction period through point sampling every five minutes. Dolphin behavioral events and vessel violation category were recorded through continuous sampling. Vessel observation included 11 vessel types to provide an assessment of vessel impacts across industries (fishing, tourism, and recreational). Focal study was conducted on whole dolphin groups rather than individual dolphins to maximise sample size. A group was defined as more than one dolphin within 800 m of each other engaging in the same behavioral state and, if travelling, travelling with the same heading. The primary behavioral state was defined as the behavior in which > 50% of the group was engaged. In the final survey period, vessel behavior was observed continuously to account for the high vessel traffic during this period. The minimum observation period for vessel violation category (before, during or after vessel interaction or violation) was five minutes.

Data underwent a series of processes to ensure accuracy but maintain optimal sample size. All behavioral data were collected by one trained observer to minimise observer error, and were validated by the senior observer. An adapted ‘before-during-after’ analytical design was used in this study. The data were divided into a single-exposure dataset (only including data from the first vessel interaction and first vessel violation observed in the sighting), and a repeated-exposure dataset (including data from the whole sighting). Mean vessel interactions (and violations) per hour of sighting were calculated by dividing the total recorded number of interactions (and violations) from unique vessels by the total dolphin sighting hours. The objective for this case study was to assess if the variation in Burrunan dolphin surface behavior in Gippsland Lakes is driven by both seasonal differences and/or vessel disturbance. To test this, multivariate analyses of variance (MANOVAs) were run in SPSS (IBM SPSS Statistics 25). The Tukey Honestly Significant Difference test (Tukey HSD) was applied as a post-hoc test to identify significant differences between means.

There are no missing values. Raw data contains five-minute observations with presence/absence and count data (where applicable).