Abstract
While white sharks (Carcharodon carcharias) are a large charismatic megafauna species that receives much public interest, little is known about their fine scale, minute-to-minute movement ecology or behaviors below the surface in coastal waters. This study used autonomous underwater vehicles (AUVs) to collect movement data (e.g., geo-position, depth) and record video to describe the fine scale habitat use and swimming behavior of white sharks in two distinct locations: shallow (<20m) waters off Cape Cod, Massachusetts (2012, n=4); and deep (<100m) waters off Guadalupe Island, Mexico (2013, n=4; 2015, n=3). The sharks were also tracked independently using an additional acoustic transmitter attached to the shark. Comparison of the tracks from the AUV, the vessel, and the shark indicated that the AUVs were able to follow the shark more effectively both horizontally and vertically than the traditional vessel-based tracking. In addition to providing more detailed tracking of the animal, the AUV provided more fine-scale, minute-to-minute acoustic data on the geo-position and depth of the shark than possible with conventional tagging methods (e.g., acoustic and satellite tags), along with environmental data (e.g., water temperature, bathymetry, current magnitude and direction) from around the shark and direct video observations of shark behavior and the surrounding environment. The AUV tracks provided a three-dimensional view of white shark movement through fine-scale, minute-to-minute recording of both geo-position and depth of the animal. Using these data, swimming behavior was quantified by calculating track linearity (index of tortuosity, IOT) and swimming speed taking into consideration water current direction/strength and changes in depth. The influence of environmental factors (e.g., time of day, tides, depth of thermocline, bathymetry, and water current magnitude and direction) on swimming behavior at these two sites were investigated. Median (Interquartile Range - IQR) shark speeds calculated across all the tracks in all three years based on vertical and horizontal distance moved over time (ROMD) across all the tracks in this study were 0.32 (0.26 to 0.37) TL/s (in meters per second - 1.52 (1.20 to 1.71) m/s), and accounting for water current magnitude and direction the median (IQR) shark speed (ROMC) across all the tracks in this study was 0.35 (0.10 to 0.74) TL/s (in meters per second - 1.62 (0.45 to 3.48) m/s). Tailbeats per second (TB/s) of the sharks counted from the video data remained consistent across tracks and swimming speeds, with a median (IQR) of 0.33 (0.31 to 0.35) TB/s (n = 69). Calculations of IOT per hour along the tracks as a measure of track linearity showed that the shark tracks taken off Cape Cod in 2012 were significantly more linear than tracks taken off Guadalupe Island in 2013, which could indicate different foraging strategies by the white sharks in these two areas. In both locations, however, white sharks were observed to swim close to the bottom at depths down to 20m in the Cape Cod tracks and 200m in the Guadalupe tracks. Comparisons of white shark swimming direction to water current direction found that the white sharks in the tracks taken off Cape Cod in 2012 (n=4) and off Guadalupe Island in 2013 (n=4) swam mainly against the current into higher water current magnitudes (median (IQR) = 2.11 (1.32 to 2.77) m/s and 2.27 (1.24 to 2.95) m/s,respectively), while the sharks in the tracks taken off Guadalupe Island in 2015 (n=3) swam mainly with the current at lower water current magnitudes (median (IQR) = 0.78 (0.45 to 1.52)m/s). This is the first report of white shark swimming direction compared to the water current. Further studies could continue to elucidate the role of water current on white shark behavior and habitat use. Video data from the AUVs provide novel views of white shark predatory and agonistic behaviors under the surface (see also Skomal et al. 2015) through attacks of white sharks on the vehicle, as well as a look at the surrounding environment and any visible other species around the white sharks. In addition, in one of the tracks taken off Guadalupe in 2015 in the evening following a female white shark, a novel behavior was observed characterized by a gaping mouth and lack of response to close approaches by the vehicle which has been suggested to potentially be the first observation of a white shark sleeping. The data collected by the AUVs in this study provide a unique, fine scale view of white shark swimming behavior. The results of this study show that AUVs can act as an effective and unique research platform that allow for the collection of fine-scale data on both swimming behavior and habitat use in large apex predators. These data can in turn be used by fishery managers and beach managers to determine essential fish habitat and may help in establishing marine protected areas for the conservation of this species as well as inform public safety efforts to avoid negative interactions between humans and this species in coastal areas.