The UK’s coastline, and increasingly its inland areas, are home to globally important populations of a host of seabird species. However, our seabird populations are increasingly being negatively impacted by a suite of anthropogenic threats including climate change, pollution, and bycatch, hence it is ever more important that we obtain robust data on the health of their populations.
Yet, the often-remote nature of seabird breeding colonies can prove difficult to monitor resulting in sometimes sparse collection of data on their population status. Our ability to understand how these seabird populations are responding to and being affected by threats is then limited. Procellariiformes, the order of the ‘tubenoses’ containing storm petrels and shearwaters, are acutely impacted by many of the anthropogenic threats facing seabirds and so it is vital that we obtain comprehensive data on how their populations are faring. However, UK-breeding Procellariiformes species are generally only active at their colonies at night and may breed in completely inaccessible or hard to reach colonies making them particularly difficult to monitor using observer-based monitoring methods.
Fortunately, however Procellariiformes are famously very vocal species, as those who have been lucky enough to spend an evening in one of their breeding colonies will know. As a result of their noisy behaviour RSPB scientists are investigating whether measures of the calls of these species could be used as an alternative monitoring method.
To investigate this a team from the RSPB had the privilege and joy of working on RSPB Mousa this summer to test whether the population size and breeding success of European Storm Petrels (hereafter ‘petrels’) may be monitored through measuring their calls alone. To this end the team installed a number of acoustic recorders, specifically ‘AudioMoths’ see below photo, across the island to record the calls of the petrels across their breeding season from June through to August.
These recorders were attached to homemade ‘parabolic receivers’ (read, IKEA salad bowls spray painted black) that were then fixed in position facing stone walls where the petrels are known to breed.
For those of you who have been out to Mousa over the summer you may have noticed some of these alien-like structures appearing in the centre of island:
The sound system worked incredibly well and the team were able to obtain over 900 hours of usable sound data from the petrels. Through the use of an automated classifier made by collaborators at Google these sound recordings were then translated into simple outputs that quantified how many seconds the storm petrel adults and chicks were calling, i.e., their call rate. Following the results of previous studies the team used the adult call rates as the proxy of population size, and the chick call rates as the proxy of breeding success.
Across the season the team also undertook the traditional observer-based monitoring techniques to which to acoustic measures could be compared. The team used a technique called playback to monitor population size in June, which involves playing storm petrel calls at set distances throughout suitable habitat, modelled by the excellent research assistants Lucy Williamson and Miguel Hernandez Gonzalez:
The number of birds that call back to the recording are then recorded and after applying some statistics an estimate of population size may be obtained.
The team then used endoscopes to monitor the breeding success of nests in August that had been identified in the playback census. For this technique the team used small endoscope cameras to search for chicks at each nest, the presence of which would indicate that the nest had been successful. While it was very difficult to obtain good views of chicks the team persisted to capture some lovely images of adults brooding chicks (left) and the chicks themselves (right):
Currently, all the numbers from the summer are being crunched and while full analyses are still ongoing it is looking very likely that bioacoustics will be usable as an alternative method of monitoring petrel population size and breeding success. Further, the team were able to detect both adult and chick calls from at least 50m away from nests. Therefore, bioacoustics may be used to monitor currently hard to reach colonies such as those on steep-sided cliffs and sea stacks, for which we currently have no population data. If bioacoustics do prove to be a valid method, then this method may be used to monitor currently overlooked and hugely important seabird species like storm petrels across the UK, and consequently better understand how their populations are faring across Scotland.
The team would like to extend a huge thank you to the Mousa Boat team and Kevin Kelly and Helen Moncrieff from the RSPB Shetland team for all of their hard work in facilitating the work on Mousa this year.
All monitoring was undertaken with the required licences and permits granted by NatureScot and the RSPB.
Dr Sophie Bennett, RSPB Centre for Conservation Science
sophie.bennett@rspb.org.uk