Wednesday July 26 - time to wean myself off identification apps?

 I'm going right off tech. Around this time last year I was introduced to an identification app called Obsidentify. It came with plenty of warnings about over-reliance on it, but once you start using you think you can handle it.

Come the moth season I found myself checking everything using the app, even when I knew perfectly well what I was looking at. It was always gratifying to have that 100% certain (or anything over 90%) backing up your identification. 

The trouble is, it makes you lazy. Why trawl through hundreds of images looking for a match when app tells you straight away? This week, some sobering app misidentifications have shaken my faith.

Tuesday night was too cold for mothing really, but I needed a moth fix and I ended up with an OK figure of 74 moths of 26 species. It was mostly species I'd seen already seen this year but I still found myself using the tech on obvious things like Canary-shouldered Thorn.

Canary-shouldered Thorn

Now to be fair to the app, the first properly small micro I looked at would have taken me a very long time to work out. Also, I should have been mindful of its 75% certainty. Nevertheless it really did look like a Plum Fruit Moth when I checked the books.

Plum Fruit Moth Grapholita funebrana...or is it?

It was only later that I realised there was a lookalike from which it can only be distinguished under a microscope. Never mind, it probably is one.

But things took a turn for the worse when I potted a rather worn moth which didn't ring any bells at all. Lazily I Obsidentified it, and was told it was definitely Crambus perlella. This is a fairly common grass moth, but I had only ever caught one before. I let it go, and moved on. Later, I checked the photograph and wondered how come it was that. It didn't even look like a grass moth. I sent the image to an on-line expert and he confirmed my worse fears. It was not Crambus anything, but was a worn Phycitodes species which could not be identified from the crap photo I'd sent him. He didn't actually say that, but I'm sure he was thinking it.

Not a clue.

On the bright side I did see three Black Arches and no technology was needed.

Black Arches

But it gets worse.

Today, after a rather uninspiring walk around Morton Bagot (Kingfisher new for the year, hardly any Green Sandpipers, an adult Peregrine and not a lot else) I joined Leigh K to help her with any tricky identifications on a butterfly transect at College Wood.

The adult Peregrine

Leigh was particularly concerned about distinguishing Essex and Small Skippers. I was pretty confident we would find some, having seen at least one at Morton Bagot earlier. The task was to prove harder than expected. A couple of good candidates was seen, and I made the call. Then I took another photograph, and backed down. They are not easy, but I was quite sure one at least was a Small Skipper.

Orange at the tip of otherwise black antennae tips, Small Skipper surely.

When I got back to the car I decided to run them past Obsidentify. The photo shown above was described as 100% Essex Skipper. What? So back home I checked on-line to see if I could find anything to support the app. I couldn't.

So apart from dodgy Skippers we saw relatively few butterflies as the cloud cover increased, but did at least see a very smart Small Copper.

Small Copper

No need for any technical support this time, or perhaps ever again!

Full list: Small Skipper 4, Small White 12, Small Copper 1, Red Admiral 5, Small Tortoiseshell 1, Peacock 8, Comma 1, Speckled Wood 11, Gatekeeper 13, Meadow Brown 15, Ringlet 1.