Chelonia Limited

  Cetacean Monitoring Systems

Monitoring methods

The following information provides a brief summary of the main features of different acoustic and visual methods of monitoring small cetaceans (excluding photo-identification) and how to select a monitoring device.

Small cetacean sounds - what can be monitored
 

Whistles

Clicks

Louder and fairly omni-directional from the animal, so that detection is easier. Clicks are highly directional with 3 dB beam widths of around 15 degrees.

Lower frequency so less absorbed by water.

Higher pitch, so larger data volumes if waveforms are stored.

Individuals may be identifiable.

Individuals are not identifiable at present.

Porpoises, beaked whales and others do not whistle.

All odontocetes click.

No species whistles continually.

Many species seem to click most of the time, particularly porpoises.

Comparing monitoring methods
 

Advantages

Disadvantages

PODs

 

Data volume less than 1 MB/day.

 

Long periods (weeks/months) of autonomous operation with no boat etc. present.

 

No false-positive rate.

 

Standardised sensitivity.

 

Fully automated detection process.

Unable to track individual animals.

Automated train recognition and export of train data allowing identification of behaviours.

Methodology for density assessment is under developed.

Can distinguish porpoises from dolphins.

Cannot distinguish between dolphin species (improvement is possible).

Can detect beaked whales, pilot whales, etc.

 

Static deployment down to 2000 metres.

 

Very low cost-per-detection.

 

Hydrophones

 

Some very successful tracking of vocalisations from baleen whales and whistles from dolphins using multiple hydrophones has been achieved.

System requires high speed digitised data storage either to tape or now more often, to disk or other medium and often requires a vessel or shore station to house the PC/ADC/power supply.

 

Data volumes for clicks can be very large - porpoise clicks at 130 kHz need sampling rates of 500 k/s i.e. around 0.1 MB/min or 1000 MB/week unless data selection is achieved.

 

Detection is done through a user-designed process generally involving spectral analysis, but lacks standardisation and often fails.

 

Click train recognition is usually by eye.

 

Cost per detection is usually high.

Towed acoustics*

 

 

The major issue for small cetaceans is responsive movement, with porpoises being averse to vessels to varying degrees in different places, while dolphins often show the opposite behaviour.

 

Cost per detection is relatively high.

Visual methods

 

The best established methods.

Cost per detection is very high.

Capable of producing absolute density estimates based on well developed theory.

Serious problems in the estimation of the detection rate on the trackline, which is essential to the density estimation.

Give by far the best species identification.

Small studies often fail because too few sightings are made to support statistical analysis unless animal densities are high.

* Various systems have been developed and are not described in detail here.

Choosing static acoustic monitors

Listed below is a brief check-list of factors to consider when making a choice between systems. These can be compared with the T-POD specification.

  • What is the duration of autonomous operation? What are the battery costs?
  • What is the false positive rate in adverse conditions e.g. shallow rough water over sand; long deployments in high productivity areas (hydrophone colonisation issues); exposure to varied boat sonars?
  • Standardisation data. Acoustic sensitivity affects detection rates.
  • How experienced are the users? Can datasets be compared?
  • Does the system have fully automated data processing? This ensures objectivity and reduces the time and cost of monitoring.
  • Does the system provide data about behaviour? Click rates are a valuable source of information on behaviour. Automated extraction has the same value as above.
  • Can you export data easily?
  • Does the system differentiate between species? Some dolphin clicks quite closely resemble porpoise clicks, although this is not much documented in published literature. To obtain an accurate assessment of porpoises in low density areas, where dolphins are also present, monitoring of both species is required.
  • What are the data access methods?
  • Is the housing buoyant? The benefit of buoyant monitors is that they are quite often recovered after being lost at sea.