Sphinx

Yes, it’s been a quiet month on here. Too much heat for my liking; a lot of time spent watering the chillis and aubergines in the greenhouse as a result; and [whispering very quietly so the gods don’t hear] I have actually been feeling slightly more myself at last, so have actually made some progress with writing. But I’ve also been doing some other stuff, in particular organising the Castle Cary Big Bat Count that took place on Saturday.

One of my lockdown investments – probably the second-best one after the new pond (which is looking rather sorry for itself at the moment – but water is still clear, what’s left of it, and there are still lots of mating dragonflies) – was a bat detector. We get lots of bats flying over the garden, but had no idea what they were – basically, due to combination of smallness, speed and darkness it’s pretty well impossible to identify bats visually beyond ‘some sort of pipistrelle’ (very small, fluttery) and ‘maybe a serotine’ (bigger, more swooping). The answer is some sort of ultrasound detector, as all the different species have distinctive echolocation and social calls; there are gadgets that allow you to listen (is it a rapid peep-peep-peep or a slower pwoop pwoop?), and fancy expensive gadgets that plug into a tablet to give you a visual representation of frequency range and ‘shape’, and then the software offers its best guess.

Having now had this thing for two years, I’ve got reasonably good not just at spotting and identifying the common signals before the software does, but even occasionally spotting things that the software doesn’t. It records and analyses a segment, and bases its suggestion on frequency of identifiable calls within that segment, so does not cope well when several bats are squeaking at once; the image above shows the utterly distinctive call of the Greater Horseshoe Bat, which came in at the tail end of a vocal Common Pipistrelle so the latter is what the software identified. And a fair number of bats have similar calls, and can be mistaken for one another especially if they’re further away, so I now analyse the more unusual ones with the help of a detailed guide, rather than just accepting the offered identification. We regularly get seven or eight species, and over two years I’m reasonably confident that we’ve had at least ten or twelve and maybe as many as fourteen of the UK’s eighteen breeding species – I’m realistic enough to mistrust some of the software’s more esoteric suggestions without a really clear recording.

I have in various occasions taken the detector round to friends’ houses, though the results have been a little humdrum – either our garden is a particular hotspot, or, more likely, you need to spend quite a lot of time detecting, which isn’t entirely compatible with being sociable. The idea of the Big Bat Count is, in a way, to test these two hypotheses by getting a decent picture of bat activity in the area. It’s part of a Somerset-wide endeavour to record wildlife; the Somerset Bat Group is organising some big public events, e.g. in Taunton, where they bring along a load of detectors and organise volunteers into groups to survey the town, but they’re also happy to loan the equipment and support the endeavour if other people want to organise such an event. Which is what I did, rounding up assorted friends and neighbours who had at one time or another unwisely expressed interest. We gathered at eight pm for briefing, then headed out after sunset to detect bats for an hour and a half before reconvening at the pub.

It was, overall, a great success; we doubled the number of Somerset bat observations (these now all need to be verified by experts), detecting eleven species on the night, and we have at least a partial picture of the different areas of concentrated activity. It’s not a complete picture owing to the phenomenon of “thank you Neville for your carefully designed division of survey areas and instructions, but we think there should be a load of bats up at the old Priory so we’re going there” – from someone I didn’t actually know, so couldn’t tell them not to…

I guess this must be a relatively common phenomenon in ‘citizen science’ – the tension between different motives, between enthusiasm and rigour. For a proper survey, you’re actually interested in identifying areas of low activity – evidence of relative absence in some parts is potentially significant for getting a sense of where bats may be roosting, where their feeding grounds and flight routes are etc. But the people who’ve volunteered to take part are there to detect bats – hence the drive to go to places where they expect to encounter them. It is perhaps a bit like the impulse behind certain kinds of metal detecting – the primary desire is to find rare and interesting things, rather than to survey an area as an end in itself, even if that would be more significant from a scientific perspective.

This was the first time the Bat Group has supported one of these events, so I think lessons have been learned, including about how to communicate the aims of the activity effectively – certainly I’d do some things differently next time. In the meantime, I am fighting the urge to become obsessively rigorous in my own record-keeping – at the moment, I record what species I detected on a given evening, rather than the number of observations. I feel compelled to start creating spreadsheets…

Update: as I noted above, all ‘casual’ observations need to be verified by actual experts in order to be counted as ‘research grade’ on the iNaturalist app. For the observations made on Saturday night, that’s a huge job for the Somerset Bat Group, who have to work through recordings to verify them and then match these to the observations made by different people so they can upload the supporting data – and if they’ve drawn one lesson from the event, it’s the need to make a record of who has which detector… But I’ve also started uploading my own recordings, and these are now starting to be verified – not just the common and obvious ones that turn up reliably night after night and I just had to pick the best recordings, but also some of the more unusual ones where I really wasn’t certain – including one that’s so rare that the precise location of the sighting is automatically obscured in the official map except for verified researchers…

For the sake of completeness: bat observations now officially verified as ‘research grade’: common pipistrelle, common serotine, common noctule, Natterer’s Bat, Western Barbastelle; bat observations where I am 95% confident they will be verified and/or I just need to get a good recording: soprano pipistrelle, Brown Long-Eared, Greater Horseshoe, Leisler’s Bat (aka the Lesser Noctule, Irish Bat or Hairy-Armed Bat); bat observations where software offers identification but experts uncertain: Daubenton’s Bat, Whiskered Myotis; Bat observations made in the past before I had much experience or could upload data, so a matter of waiting for future observations: Nathusius’ pipistrelle, Bechstein’s myotis. Bat the software detected but not actually found in Somerset: Grey Long-Eared.