Dolphins can be detected by visual observations, and/or Passive Acoustic Monitoring (PAM; www.passiveacousticmonitoring.com) or Active Acoustic Monitoring (AAM). Dolphin detectors often refer to PAM devices known as the C-POD (www.cpodclickdetector.com) and its analogue predecessor, the T-POD (www.t-pod.co.uk). T/C-PODs both use proprietary signal processing algorithms to detect and classify dolphin and porpoise echolocation clicks (www.chelonia.co.uk). Other PAM systems also employ signal processing algorithms within software packages such as the open source PAMGuard (www.pamguard.org), that detects dolphin whistles and clicks. Both PAM techniques listen for marine mammal vocalisations and do not emit any noise. See www.dolphindetectors.com to learn how these two types of dolphin detectors work.
Dolphin detectors, such as T/C-PODs, are ideal for collecting long term data, as they can be left in the field for extended periods of time, collecting data continuously. This is useful especially in areas with low numbers of dolphins where other methods would not generate enough data for statistical analysis. Because these dolphin detectors operate 24 hours a day, researchers are also able to look at how dolphin activity may change throughout the day.
Dolphin detectors are used to better understand where and how dolphins are distributed. This is beneficial in minimising any potential effects caused by marine industries such as fishing, seismic, drilling and military sonar and renewables. For example, C-PODs were deployed off the Cornish coastline in the UK to enhance our understanding of dolphin distribution with the aim of minimising by catch (www.bbc.co.uk).
Records from a study in Doubtful Sound, New Zealand deployed nine T-PODs over a 12 month period. Results indicated that that bottlenose dolphin (Tursiops truncatus) distribution varied seasonally (www.tandfonline.com). Inner fiord sites were found to be used most often in summer (1 January – 31 March) and autumn (1 April – 30 June). Dolphins used outer fiord sites more during winter (1 July – 30 September) and spring (1 October – 31 December). When compared to Sea Surface Temperature (SST), a positive correlation to dolphin records was found (i.e. when SST went up so did dolphin detections).
Odontocetes (toothed whales, dolphins and porpoises) often make buzzing noises while foraging. This noise is created by echolocation clicks (www.en.wikipedia.org) occurring in quick succession. Animals tend to click more quickly when locking onto a target i.e. when catching prey. The New Zealand Fjordland study found that there was more echolocation activity and buzzes during dawn and/or dusk. Information from studies such as this is vital to the protection of the small resident population of bottlenose dolphins in Doubtful Sound (www.doc.govt.nz). With a better understanding of dolphin distribution, dolphin protection zones (www.doc.govt.nz) can be assigned appropriately.
The T-POD has also detected Hector’s dolphins (Cephalorhynchus hectori) effectively (www.dx.doi.org). T-PODs deployed at three sites around Banks Peninsula, New Zealand over winter (June – August) and summer (December – February) from 2004 – 2006, found that there were more dolphin detections over summer compared to winter, but that there was not a significant difference between day and night (www.int-res.com). The ability to detect Hector’s dolphins with PAM techniques is especially beneficial because their small size makes them difficult to observe visually. This study also highlights the importance of accounting for individual POD sensitivity; a significant difference was found in the sensitivity of the three T-PODs used. If POD sensitivity is not accounted for in studies that compare dolphin detections between sites, it is possible that the differences could be due to POD sensitivity and not actual differences in dolphin distribution.
T/C-PODs have been used on a variety of dolphin species throughout the world. A list of T/C-POD publications can be found at http://www.chelonia.co.uk, and a list of recent discoveries at www.chelonia.co.uk. T/C-PODs are also used to detect porpoises (www.porpoisedetectors.co.uk).
DOLPHIN WHISTLE CLASSIFICATION
Signal processing algorithms not only detect dolphins, but can be used to distinguish whistles from different species of dolphin. For example, using a simple semi-automated contour analysis and multivariate statistical techniques, researchers have been able to separate whistles of five delphinid species (long-finned pilot whale (Globicephala melas), Risso’s dolphin (Grampus griseus), striped dolphin (Stenella coeruleoalba), short-beaked common dolphin (Delphinus delphis), and bottlenose dolphin) found in the western Mediterranean basin (www.dx.doi.org). Using 14 variables from whistle contours, pilot whales (75.5%) and striped dolphins (79.8%) were classified correctly and most often, whereas Risso’s dolphins (37.3%) the least. Similar methods have been used for real time classification and detection (www.dx.doi.org). The ability to classify dolphin whistles and clicks to species level is beneficial, as their vocalisations can be similar and therefore difficult for a human observer to distinguish. These dolphin detectors improve the use of PAM for marine mammal distribution studies, particularly for species that are difficult to observe in the field. Dolphin detectors within software programmes such as PAMGuard allow PAM operators (www.pamoperator.co.uk) to better implement marine mammal mitigation measures (www.marinemammalmitigationplan.com), especially those in areas where dolphin identification is important (e.g. New Zealand, www.doc.govt.nz).