Tag: Plant databases

Google Photos as a natural history resource.

Most of these photographs were taken on what turned out to be a life changing trip to Dartmoor in March 2016. The Potwell Inn didn’t exist; we’d been through retirement and a major family crisis when our second grandchild was born with a still undiagnosed genetic disorder; we’d moved from a large vicarage to a small flat in Bath and the allotment was yet to come. Christmas had been a write-off and we had taken ourselves off to St Ives in the campervan in search of some respite – in the dog days of the old year just as a major Atlantic storm reached the south west.

January 1st 2016 – a stupidly hubristic resolution

The key photograph here is the one of the (at the time) unidentified little seagull. We were leaning on the railings above the harbour just down from St Ives lifeboat station and watching the gulls when I realized that this gull was very different from some of the other gulls and I had no idea what species it was. I remember feeling faintly annoyed with myself and later that day we went into the bookshop and bought the first bird guide we’d ever owned. It was about then that quite the stupidest resolution I’ve ever made began to form in my head. I would not – I resolved – pass by any plant or animal, without naming it. The picture I took that day was completely hopeless in terms of identification, even with an 80mm lens there was no detail that had any significance so far as I could see. Just for the record I now know perfectly well that the dark smudge behind the bird’s ear is all that remains of the Black Headed Gulls’ distinctive summer plumage on December 29th.

Before we retired (i.e when we had a bit more money), I bought an iPhone and a Macbook. Madame had already settled into the Apple ecosystem because she was teaching photography at the time and it made sense for us to share resources. For the first time I owned a mobile that would add location data to the EXIF file. At roughly the same time I started keeping a private journal using Day One software and this integrated seamlessly between the iPhone, the Macbook and a big desktop. But it wasn’t so long before we came up against Apple’s policy of abandoning their products by making it impossible to upgrade the software. Day One suddenly stopped talking to my Macbook and when I complained I was told that it was my fault for not upgrading. We struggled on for several years, but applications like Photoshop and Lightroom were less and less easy to run on our aging machines and in any case their memory hogging features were far beyond my own needs. I was using Microsoft Windows for work and so we landed up in the worst of all worlds, moving photographs between incompatible software. Heaven knows how much data was lost; but in the end I began a long process of ditching Apple and moved as much data as I could find into Dropbox as a security for the future.

It was our oldest son who broke the impasse by giving me an old Pixel 3 phone after an upgrade and I started using Google Photos. To be honest there was nothing about the iPhone that I particularly missed, and when – a couple of years later – he gave me a basic HP Chromebook that he’d acquired during another upgrade, and after he helped me with a fast and pretty seamless setup and data transfer, I put my agonisingly slow Macbook on a shelf and started enjoying an entirely new and fast cloud-based ecosystem.

Anyway, to get back to the photographs at the top of this post, they represent something of a sea change. My ludicrous resolution to try to name things brought about a move towards much more purposeful photography. As my interest in plants moved up a step, I discovered that well composed photographs of plants I didn’t recognise, meant I didn’t have to dig them up and take them home which is at best inconsiderate and at worst illegal. It’s taken ever since to refine the technique so that I remember to photograph the sometimes apparently insignificant details that make the difference between a correct and a wrong identification.

None of this particularly mattered, of course, as long as I was the only person ever to see the record. But then, as I wrote in a post last week, I began sending them for verification by the local BSBI East Cornwall recorder – which is when it dawned on me that my photographs were more akin to archaeological compost than a filing system. Looking for old photos meant trawling through thousands of them in the hope of alighting upon the right one.

By now I’d got myself a powerful little Pixelbook and I was beginning to understand that the originally utilitarian Google Photos has a few new tricks up its sleeve. One of Google’s most useful – although much feared – attributes is its formidable artificial intelligence software. If you’ve ever used Google Lens on the phone you’ll know that. It can be a curse as well as a blessing – this morning I had an email from someone asking what was the launching fee at Percuil. We were on holiday for the last two weeks and I mentioned Percuil several times in the blog and the odd request must have flowed from a bit of AI that was too clever for its own good.

Google Photos uses AI to do a lot of basic sorting and indexing of photos behind the scenes; so I can search on months and years, faces and places. But last week I wanted to start indexing keywords like genera, species, English and Latin names and various other bits of searchable data. I’d contemplated using some third party software but after hours of searching I couldn’t find anything I liked better. Any records I’d submitted or had accepted would be somewhere in one of a number of vast external databases, but I wouldn’t be able to access them in a simple personal list.

So here’s the exciting thing. After a bit of experimentation it seems that contrary to what I read on the internet, I can do keyword searches on terms that I type into the “details” column at the head of all the EXIF data. The only downside is that the AI facilities remain in play, so a keyword search on primrose might yield a few lookalikes as well – not the end of the world in my view. My guess is that some of the AI used in Google Lens is already being built into Google Photos. The software connects seamlessly to WordPress, so I can use my photographs in this blog; and I can easily forward diagnostic photographs to the referees for acceptance.

When we go out plant hunting I take the phone with a macro lens so I can take the necessary photos. I’ve got stacks of useful apps on the phone as well, including OS mapping, accurate National Grid references, altitude, and shortcuts to the BSBI plant distribution database. I’ve also got iRecord, I Naturalist and a link to the local list of plants which I download as a PDF before leaving home. Google Lens is surprisingly helpful in steering me in the right direction but it’s not infallible. Of course I will take a book along with me but I always take photographs because I can use them with my library of plant books at leisure when I get home. It doesn’t always work, and I often have to go back, look more closely and take more photos in order to do a full ID.

The upshot is that the more of this kind of intense work I do, the more my plant ID skills improve. Typing details into the software drills the English and Latin names into my memory very effectively too. I think the take-home point is that while for some people, staggering across a bog with a large camera and tripod, an expensive GPS unit and a copy of Stace 4 weighing in at well over a kilogramme is the only way forward. But for my purposes I can get it all into a pocket without my trousers falling down. I take a 6″ ruler, I use my walking pole as a 1.5 metre measuring stick. I take a plastic zip bag, just in case; a Rite in the Rain waterproof notebook and a pen that writes on it underwater and upside down – who knows when that might be needed! a scalpel and a couple of hand lenses. Total weight around 500g.

Maybe I’ll be forced to eat my words one day, but this seems to me to be the simplest and cheapest way – it’s free – of creating a searchable, automatically backed up photographic database for plant recording. No good at all, of course for insects, mammals and birds because they all race around too far away for a phone camera. But my subjects stand still – unless it’s very windy – and best of all the photos escape the dangers of being lost forever in my fallible memory. What’s not to like?

Computer says no!

Having gone on – perhaps a little too lyrically – about the joys of big data in yesterday’s post; I came up against one of its limitations today. It was nothing; or it was something …… who knows, but it was a wonderful example of the way that – re-using the phrase I used yesterday – absence of evidence is not evidence of absence. Or to put it in more immediate terms – there is little to no evidence that the plant in the photo has ever been found in, or anywhere near, the Lost Gardens of Heligan – or even in East Cornwall. This may of course be because the gardens – being lost – cannot be invaded by this plant because it can’t find them. On the other hand, there it was in front of me on Wednesday, and all I knew for sure was that I’d never seen one before …… but it was almost certainly a form of comfrey.

So I spent quite a few painful hours checking and then double checking but, as an amateur, you have to hang your hat somewhere and so I decided it was probably but not certainly Symphytum bulbosa. The haunting possibility is that it’s something entirely different; so much so that any halfwit would have known its real name. Common as muck! Where have you been living, you botanical imposter. And there’s the crux of the problem. There are two possible outcomes to the conundrum and one of them is to submit the record to the scrutiny of someone who actually knows what they’re talking about, but who may dismiss me scornfully they may also be looking at the same maps, records and data as me and go on to say that tuberous comfrey simply doesn’t grow where I saw it and so, QED I was wrong.

Now this isn’t that big a deal – it would hardly be a hardly a paradigm shift if I turned out to be right; but the sheer inertia of the system resists change and for a learner like me, taking on a system is a really big deal. Nobody dies either way, whether I’m right or wrong; but if I don’t submit the record because I don’t want to rock the boat, everybody loses. I miss an important lesson, and the maps and data are just that little bit less accurate than they could be.

So I bit the bullet and filled in the online form; emphasising my lack of experience and humility (I wish!) in the company of experts and all that blah blah, and then spent twenty minutes trying to change a GPS reading from my phone to British National grid without realising that the programme accepted it either way. The only thing I forgot to fill in was the species name which by this time I had forgotten. So once I’d found the name again I filled in the form and immediately a red triangular warning sign appeared against the entry; I’d been flushed out before I’d even pressed the send button!

So there we are. My contribution to citizen science is out there, somewhere in cyberspace, awaiting final judgement by a recorder who may well be knackered and daunted by a stack of 10,000 records that they wished they’d never signed up for.

Today the sun shone. There was a stiff northwesterly wind and so we found a sheltered spot in full sun on the beach and I recorded some celandines and sea beet. At least I know what they are.

%d bloggers like this: