Image courtesy of Sorcha Lewis
The Cambrians are justly acknowledged for their exceptional tranquillity. The absence of traffic noise and human development means that in most of the uplands you can really hear the sounds of nature – wind, water and birdsong – in a way that can be hard to find in much of Britain. But at the same time, being largely remote from towns and roads, it can be challenging for bird watchers to observe what birds are present and absent across the region on any systematic and regular basis. As a result, though we know we have an exceptional and biodiverse bird population, we found that we do not have a good picture of how those populations are distributed and how they are changing over time.
An AudioMoth in the field
To address this, the Society has set up a Soundscape project to try to capture a better understanding. Partly from Society funds and partly funded by some very generous donations from Members, we have bought a batch of AudioMoth automatic recorders. These are tiny devices developed by the University of Southampton which can be set up to record at pre-determined intervals all day and all night for up to three months at a time. Once the memory card is full, we retrieve them and download the data to be analysed by BirdNet software (developed by Cornell University in the USA and familiar to many as the Merlin bird identification app). Subject to a few caveats (extraneous noises do confuse it into giving some quite improbable results!) this can identify the bird species which come within range of the microphone (about 100m) in a location hour by hour and month by month.
We’ve deployed these in a wide range of habitats, from heathlands to the oak woods, and from north to south, east to west across the uplands, since October 2024. Our plan is to keep them in the field, and even expand the array to fill in some gaps, until we have a full 12 months of recordings. This is equivalent to having a team of expert ornithologists sitting out night and day, rain or shine, for a year – something which neither is nor ever will be practicable.
Leach’s Storm Petrel (image courtesy of Newfoundland birder Bruce Mactavish)
Of course, we’re going to be taking great care in analysing the data to make sure that we’re only reporting genuine birds. One of our first ‘false positives’ was an identification of a Leach’s Storm Petrel – not impossible, as apparently they are “infamous for being driven inland in the fall”. But on listening to the recording which the software believed to include this bird call we heard what was definitely a pig rootling in the undergrowth. On another recording, hum from a nearby power line somehow reminded the software of a Woodcock. So the software is certainly fallible and human ears and judgment still have an important role to play. We’re drawing on the County Bird Reports and NBN Atlas data as a first line of sanity checking. Nevertheless, it’s worth the time and effort: after all, it is also infamous that ‘out of sight is out of mind’. If we collectively don’t know what birds we have, what chance is there of making the right choices to conserve them to inspire future generations?