PippyG - A new way to record Bats?

 One of the challenges I have set myself is to create a complete species list for my patch. One taxon that is currently under recorded is Bats. I have made a few visits to the site in the summer with a handheld detector; I am currently using an Echo Meter Touch 2 from Wildlife Acoustics. This device fits on to an iPhone and is remarkably easy to use. However, what I needed for a more systematic survey of bat abundance was something I could deploy in the field and leave, a kind of trail camera for sound.

This is not as strange as it sounds and there are devices out there that make devices for recording sound. Wildlife Acoustics are obviously the leader in this technology and produce a number of models for this purpose with their song meter range. This range allows for a range of species identification and their MiniBat records ultrasonic sounds. These devices, however, range from £160, up to £1300, a little pricey for me but very tantalizing.

The more I researched the more interested I became the more I found, the Titley Chorus is about £462 and AudioMoth offers a low cost build your own version although assembled devices cost about £69.63.

During my searches I came across the BTO acoustic pipeline. This section of the British Trust for Ornithology’s website enables you to upload recordings from audio equipment that it will analyse and present the data. There is a cost to this, but for a non-commercial operation like mine its pretty cheap with many facets of its access free or purchasable via credits that can be topped up as needed by payment.

This solved one of my problems, I am in no way an audio engineer nor am I skilled at interpreting Bat sonograms, I am even shaky on bird song to be honest. The question of how I could analyse the information is what was holding me back the most. There are bat call analysis software but this represented an additional cost.

As I investigated the pipeline I looked at what hardware was being used by the website and stumbled across PippyG. PippyG is one of a range of inexpensive bat detectors developed by Phil Atkin, which like AudioMoth can be assembled from scratch. Phil is a hardware designer and software developer, and his models are based around the Raspberry Pi the small, affordable single board minicomputer that can be used for a range of micro electronic devices. The benefit is that the addition of a microphone to such a device makes a very cost-effective detector. PippyG and the whole range of the pipistrelle family of products is now being sold by the Dutch company Apodemus and in the UK by NHBS.

Now I do warn you, the device is very barebones. Do not expect a fancy case or easy to flick switches or an extensive manual. The PippyG is a device in its rawest form and right at the edge of my technological understanding. It arrives in an unassuming box, and I struggled for a while on how to ‘reflash’ the device to install software updates.  It is not intuitive, and you do need the accompanying free app to make it work but slowly I started to understand it. Phil Atkin has several YouTube clips available and Apodemus is starting to develop its support resources. I was at one point very stuck and frustrated and contacted Apodemus by email, they were incredibly responsive and helpful even replying in the evening and weekend to ensure that I was able to get the device to work. If you’re not familiar with reflashing then it’s all very confusing.

With the software updated and an understanding of how to use that app I was ready to field test in it the garden. The device lacks any waterproofing and despite the hot weather I was concerned about damage and so I set the device out in my garden under a plastic strawberry Punnett. To set the device you use the Pipistrelle app, this codes the location of the device via GPS, identify the device and sets the wake and sleep times. You can then select the trigger sensitivity and the frequency you wish to record. The device offers Bats – 384kHz and Birds 48kHz. You then select the length of recordings, I would recommend the 5 second option as it is much easier for analysis and then you press configure. Like a bat itself the information is transmitted to the device via a chirp.

With the device primed you switch it on and walk away. The files are recorded on the microSD card which can later be uploaded on to your computer. I would recommend you purchase the model of SD card that Apodemus suggest, I bought a different one first and had difficulty.

Using the BTO Acoustic Pipeline you can upload your data and the software on the website will process the data giving you a list of species identified with a probability of certainty percentage.

It took a couple of attempts to test it, I obviously needed bats to be present to get a positive result and on the second night of trying I obtained some results. On Test 4 76 detections were made resulting in 5 species being recorded. Unidentified bird, Soprano Pipistrelle, Common Pipistrelle, Noctule and Pygmy Shrew. All had a maximum probability of correct identification of 91 or greater except the shrew which was 46%. The discussed level of acceptance is 50% however given my lack of understanding of the software, hardware and general innate cynicism I will only be accepting 90% probability going forward, at least for awhile.

The shrew record is very interesting, whilst I doubt it was a shrew it does indicate how useful the device could be in identifying other ultrasounds and could be useful in future mostelas studies I carry out when differentiating between common and pygmy shrews is difficult.

Clicking on any species reveals a time plot revealing when this species was recorded.

 Overall I am impressed by this piece of kit and the BTO software. It is opening up another avenue of research for me and I look forward to deploying it on my patch once I have decided how to protect it from the elements.

This seems at present to be a very cost effective way to record species and I look forward to experimenting over the next few weeks in its use.