Help me identify this please

[magnarocks]

My lovely friend gave me a chilli plant and some flat leaf parsley seedlings... She has been doing lots of growing from seeds in lockdown. This little chap has grown in the pot with the flat leaf parsley, she says she has no idea what it could be... except for a plant.

Any thoughts? I find it weird how the top leaves are sort of jagged but the new bottom ones are v smooth. 🤷‍♀️ 🌱

Leaves at the bottom aren't the new ones, they are the seed leaves. They're always a different shape. The jagged edge ones above are the true leaves. Not sure what it is but definitely shrub/tree like. Be interesting to see the seed/nut it's from. if you're v careful you could remove soil and have a look

Mint?

Not sure what it is but definitely shrub/tree like.

Yes...it looks familiar but it's eluding me at the moment.

Another one for mint here - if you gently rub the leaf does it smell!?

If it were mint the seed leaves wouldn't be that large. It would smell of mint, and the main stem would be square in section. And the leaves would be opposite not alternate - ie they'd appear in pairs at opposite sides of the stem, not one at a time up the stem.

Looks like a Prunus seedling.

It doesn't look even remotely like mint to me. I think Dint is heading in the right direction with prunus.

Have a look in your friends garden or her neighbours gardens. It could be drifted in from there. Having said that I got a hazel seedling from a nut I could only guess was buried by local squirrels

Looks like birch leaves to me! I'd be surprised if you have a tree growing there, though.

I was bet money that’s a tree sapling. Not sure what though.

Here's some photos from today. It's growing quickly.

Picture this app suggested black hawthorn or an elder - not sure how reliable app is?!

not sure how reliable app is Not at all in this case! Certainly not elder. If by "black hawthorn" it meant "blackthorn", that's a Prunus, but not sure whether you can tell which Prunus at this stage.

The more I see completely daft identifications that have come from the apps, the less inclined I am to get one!

The apps are doing the same thing as a novice gardener would do - flick through pictures on the internet till they see something that matches, But they have two big diasdvantages 1) they're not identifying what they're seeing as "leaf" "flower" etc, but only as patches of colour, so they can easily go wrong 2) they're overconfident of their abilities so don't warn you that the id may be wrong.

Add to that that they have no botanical knowledge, so they can't tell the difference between, say, a branch with a lot of simple oval leaves, and a "compound" leaf which is broken up into oval leaflets; or that the things that they've just identified as "stonecrop" from the overall inflorescence pattern is in fact a 10 ft tree, and they can go quite spectacularly wrong ... and of course someone new to gardening or botany doesn't realise how wrong the id is.

The apps may get better over time if they get more training data - with accurate IDs in the first place, of course.

You're absolutely right, Dint, about how the apps get it so spectacularly wrong. At a minimum, anyone getting an ID from an app needs to double-check it by looking up whatever the app says it is on a reputable website (or even, if you're old school and have access, in a book).

The apps may get better over time if they get more training data - with accurate IDs in the first place, of course. Hopefully that will happen before the experts die off who can tell when they've got it wrong. There's only one university in the country which teaches plant identification, and that's the MSc in biological recording and ecological monitoring at manchester metropolitan uni.

There's still a problem with, for example, forget-me-nots, where distinguishing species depends on observation of the sepal shape and the hairs on the stems - things that you are unlikely to be able to see either in photos on the internet or on the photo taken by the person trying to identify. Differences in DNA aren't necessarily carried forward into easily observed physical differences. Of course, the same problem applies to an expert trying to id from a photo, but at least they're aware of the problem.

Differences in DNA aren't necessarily carried forward into easily observed physical differences

Especially when the form may be modified by environment/development. The same species may look significantly different if it's grown in shade or sun, acid v alkaline etc. Leaf size can be deceptive too - restricted roots 'bonsai' tiny leaves, versus a tree regrowing after being cut back but with a big root system can give really big leaves.
Artificial intelligence has its place, but it's not (as yet) a substitute for the real thing! And humans are very good at pattern recognition. It goes further with bird ID where we can integrate knowledge of habitat, sounds and the way they move.

Habitat is very important in plant recognition too, and even more important when you get to things like mosses.