Bottlenose dolphins produce a variety of sounds, most of which are either echolocation or whistles (Figure 1).  Echolocation is used mainly in navigation and foraging; a rapid series of short “pulsed” broadband sounds are produced by the dolphin, and the returning echo is used by the animal to determine information about its environment (for example, the presence, size, and movement direction of a fish).  Whistles are relatively long, lower frequency “tonal” sounds.   These sounds are thought to be mostly used in social contexts.  The Sarasota bottlenose dolphin community is one of only two populations of dolphins worldwide which have been shown to produce “signature whistles”, whistles which are unique to an individual.   The highly acoustic nature of dolphins makes the use of passive acoustics, or listening, a useful tool in detecting their presence.   The purpose of this project is to use underwater recordings to detect bottlenose dolphins in Sarasota Bay and to monitor noise levels which could potentially disturb these dolphins. 

Figure 1. Bottlenose dolphin whistle (left) and echolocation (right), recorded in New Pass, Sarasota Bay (48,828 Hz).

      Hydrophones were deployed in November 2005 in New Pass, Sarasota Bay (directly behind the Ann and Alfred Goldstein Marine Mammal Center at Mote Marine Laboratory; Figure 2).  Hydrophone data were digitized and saved onto an external hard drive, and all processing and data storage components were housed in a waterproof “dock box” located near the hydrophone.  Data from the external hard drives can be analyzed in two ways.  Manual inspection of the data files is practical for small data sets (by listening to them or visually identifying the recognizable waveforms of dolphin vocalizations in acoustics software).   However, for large data sets this method is too labor intensive.  The MATLAB program DPASS (Dolphin Passive Acoustics Surveillance System) uses a series of mathematical functions to identify dolphin whistles and echolocation pulses from acoustic files.  The New Pass recordings, which currently constitute several terabytes of data, can only be analyzed with such a technique.

Figure 2. Location of four hydrophone array (white circles) in New Pass, Sarasota Bay.

      One question that can be addressed with these data is the overlap between visual sightings conducted by SDRP and acoustic detection.  To investigate this question, 38 visual survey times were selected and the detection of dolphins by visual and acoustic methods were compared.  During a total of 10 visual sightings, four had corresponding dolphin vocalizations, and six did not.  In addition, there were five incidences of dolphins being heard, but not seen.  These results suggest that dolphins are frequently silent as they use the pass; however we are currently determining the precise acoustic range of the hydrophone.  Once this is determined, we can use the exact GPS position of the dolphin sightings to test for overlap in a more rigorous manner.  For this data set, whistles and echolocation clicks were identified manually; however DPASS was tested using a subset of the New Pass data.  50 files with dolphin whistles and 100 files without (and containing various levels of boat noise) were used to test DPASS.  The program was able to correctly identify 92% of the dolphin whistles correctly, with a 2% false detection rate. 
                New Pass is a high traffic area for recreational boats, being one of the only passes between Sarasota Bay and the Gulf of Mexico.   This results in a great deal of noise in the aquatic environment.  High levels of background noise have been definitively shown to affect cetacean behavior.  One of the goals in this project is to quantify the boat noise present in New Pass over time.  This has important conservation implications for the bottlenose dolphins of Sarasota Bay and the Gulf of Mexico.  In addition, the use of acoustics allows us to monitor these dolphins continuously, including at night and in poor weather conditions.  In the winter of 2008 we will be wiring the hydrophones into the Marine Mammal Building and developing a web page to broadcast recordings from the hydrophones.
       We wish to thank the Harbor Branch Oceanographic Institution’s Protect Wild Dolphins program for providing the funding making this research possible.