Default Blog Post Feature Image - Baker Consultants Marine Logo

Acoustic porpoise detectors vs recorders

Harbour porpoises are the most abundant cetacean species in the North and Baltic Sea. They are very difficult to detect by visual means as they are small, shy and inconspicuous animals. Therefore more success has been achieved using passive acoustics methods. Porpoises use echolocation clicks to navigate their surroundings, find prey and communicate. These clicks can be detected using specialised equipment. This method has been successfully used for various research projects as well as commercial applications.

For commercial applications (e.g. offshore wind farm construction) it depends on the survey goal which passive acoustic monitoring equipment is used. For baseline surveys (to establish the status quo before any construction work is carried out), it is common to use towed hydrophone arrays in conjunction with visual surveys. Whereas during the construction phase it is common to use stationary recorders as a higher temporal resolution can be achieved.

Baker Consultants Marine has extensive experience using static equipment for porpoise monitoring. There are two main groups when it comes to stationary equipment. Baker Consultants Marine always favours equipment which is transparent in how it works therefore we recommend for example the SM3M (manufactured by Wildlife Acoustics) or the SoundTrap (manufactured by Oceans Instruments) which are RECORDERS. This means the as the name suggests the unit records a soundfile. The soundfile can then be accessed and analysed by bio-acousticians and can verify porpoise detections based on porpoise click characteristics. The other group are DETECTORS such as the C-POD (manufactured by Chelonia). A built-in algorithm analyses the data for potential porpoise detections and the output is a summary file of potential porpoise detections. You do not have access to the raw recordings data nor has the click classification algorithm been published. This equipment is what scientists call a ‘black box’ as although it has been proven to work nobody knows how it works. Baker Consultants Marine supports equipment which is transparent and the results can be justified.

Acoustic monitoring device being prepared for deployment.

Acoustic monitoring device being prepared for deployment.

As we carry out a lot of our work in Germany we encountered a discussion regarding porpoise acoustic equipment. The German guidelines specify the minimum requirements of the equipment but it is targeted towards one specific model- the C-POD. Nevertheless, the German government approved the Baker approach using recorders rather than detectors for various projects. Baker has successfully carried porpoise monitoring using recorders.

Whilst the detector has the advantage that the settings are kept constant for analysis it is not transparent. Furthermore, detectors work by identifying target signals above a threshold level against the background noise; therefore, their performance can be highly influenced by the environment (and mooring) that they are put into. The analysis from a broadband recording can be undertaken with different approaches but in line with common practice the method is stated in detail so that is replicable across studies and commercial monitoring projects and the raw data can be investigated visually and acoustically by experienced operators. Furthermore, the raw data is always accessible and a different analysis approach can be carried out if deemed necessary.

We have proven in a recent analysis with a direct comparison of SM3Ms and C-PODs that the recorders perform better and can be used reliably and comparably to C-PODs.

A different experiment recently carried out by Sarnocinska et al. (2016) compared the performance of the SoundTrap (Recorder) to the C-POD (Detector). The results of this study confirm our findings as well. Depending on settings C-PODs only detect 21-94% of the clicks which were detected using the recorder. The main reason for this discrepancy is likely due to that recorders can detect individual clicks whereas the C-POD algorithm only works on click trains (i.e. 5 or more consecutive clicks). Furthermore, the detection algorithm differs and it seems likely that C-PODs are more influenced by background noise. Whilst there was a correlation between these two methods, it is very variable. This may have implications on the overall results and especially in areas of low abundance (e.g. Baltic Sea); here, recorders should be favoured over detectors in order to obtain robust results.

Baker Consultants Marine believes in transparency and therefore recommends recorders to promote a robust data set for porpoise monitoring.