Abstract
This report provides an assessment of the feasibility of using a Passive Acoustic Monitoring (PAM) network on the Dutch Continental Shelf (DCS) to monitor harbour porpoises. It provides a description of what static PAM is currently able to do. It gives an overview of the technical and logistical requirements when using PAM at sea and the analytical limitations when using PAM to monitor harbour porpoises. Rijkswaterstaat was asking if PAM can provide:
• Absolute abundance estimates of harbour porpoises.
• Relative abundance estimates of harbour porpoises.
• Temporal and spatial changes of harbour porpoise presence in relation to abiotic and biotic factors.
• Data on diving behaviour of harbour porpoises.
Currently there are two types of autonomous static PAM devices to detect porpoise vocalizations: click detectors (such as CPOD’s) and wideband or full-spectrum sound recorders. The choice for a device depends on the research question.
The theory to estimate absolute porpoise density and subsequently abundance from echolocation clicks detected with PAM is solid. In practice, however, a number of the parameters needed to derive absolute densities from the acoustic detections are often not quantifiable. Where these parameters, such as click rates and detection probabilities, are not known, potential bias can be introduced. One option is to assume that these parameters, such as for example click rates and detection probabilities, are constant or random. In that case PAM can be used to derive minimum or relative density estimates of harbour porpoises. The great advantage of PAM is their continuous recording of sound, providing information throughout day and night and all seasons. This has made them a valuable tool to describe temporal and spatial patterns in harbour porpoise presence. Recently click patterns have also been analysed for behaviour providing more insights on porpoise habitat use.
To illustrate the steps needed to monitor harbour porpoises with PAM, we present two case studies for setting up a PAM network covering the Dutch Continental Shelf and a PAM network covering an operating offshore wind farm. These case studies include different options and associated costs.
Based on the fore mentioned points, the following conclusions are drawn:
-Deriving absolute abundance from passive acoustic monitoring is still in its infancy and while theoretically possible the needed information to do this is still not available.
-The primary advantages of PAM are the continuous monitoring, allowing insights in changes in behaviour and habitat use on short (hours) and long (years) temporal scales.
-PAM networks have shown to be a useful tool to provide a measure of relative abundance, longterm and of continuous habitat use, in particular in smaller areas or for populations that occur in numbers that are too low to be assessed by visual survey methods.
-For the aim of obtaining relative abundance estimates, the costs for PAM networks are relatively high compared to visual line transect surveys.
-Some options are available to derive some of the missing parameters needed so that PAM networks may provide absolute abundance estimates. For example, use of aerial surveys in combination with PAM stations as done in Jacobsen et al. (2017) could also be an option for Dutch waters to derive an estimate of EDA.
-Before a PAM network is to be established in Dutch waters it is important to clearly define the goals of such a network, and assess the feasibility of realistically achieving these.
-A step-wise approach could start with testing, developing and improving PAM network to assess harbour porpoise abundance in small-scale areas. Once the methodology is tested it can be further extended to a larger-scale (e.g. DCS-wide) monitoring.