There is a young plant in our front garden, here in Epsom, perhaps sprouting back after being cut down last year, and we wanted to know what it was. Four shoots, the tallest around a foot high. Not the same as the much bigger plant above and to the right. Despite there not being very much to go on, off to Google Images (GI), with whom I went two rounds.
What follows is very much work in progress. Not a neat and tidy story with which to close the file. I am reminded of all the effort and expertise that must go into creating the neat and tidy stories which you get in good science textbooks: no wonder that some of the bigger ones are a collaborative effort.
I thought a zoomed version of the image might work better, bringing out the distinctive red markings at the base of the leaves.
First round
I started the first round with: ‘likely a seedling, behnd a front garden wall in Epsom. Three or four shoots, about a foot high. Brown stripe at the bottom of each lear’. Typos left as in the original.
GI plumps for: ‘The plant in your garden appears to be a Goldenrod (Solidago species), most likely likely a seedling, behnd a front garden wall in Epsom. Three or four shoots, about a foot high’. Note how he echoes my prompt.
He points to the small teeth along the leaf margins which I had not noticed. But he does not deal with the triangles. I also wonder whether a (first year) seedling would have four shoots in this way.
He also offers aster (Symphyotrichum species), phlox (Phlox paniculata) and willowherb (Epilobium species) as possibilities. But adding that phlox leaves are strictly opposite, which these are not.
I leave GI for a bit to have a bit of a poke around with Bing and Google. Can I turn up any good images?
Second round
I started the second round with the same image, but with a revised clue: ‘Three or four stems about a foot high, behind a front garden wall. Distinctive features include the slightly toothed leaf margins and the thin, trianglar red start at the base of the top of each leaf. Something I have not managed to find among the images of plants suggested previously, for example golden rod, aster and willow herb’.
GI sticks with goldenrod, without dealing with the red start business. But he adds ‘While you mentioned Ayapana triplinervis…’ – which I had not – so I don’t know where that came from. He explains that this is a tropical plant, unlikely to turn up in Epsom.
I followed up with: ‘Odd that none of the images you turn up has the red triangle in question. Maybe a couple of millimetres wide at the base and a couple of centimetres long. On leaves maybe ten centimetres long’.
GI now explains that this is a feature of young leaves which will soon fade. Dye leaking into the fast growing leaf from the stem. One does not use transient features of young plants for identification purposes. But he goes on with ‘The Triple Vein: Look closely at the leaf on the left side of your photo…’.
I am not sure which leaf he means, but in any event, there is no such thing to be seen. I go down and check the plant itself, still nothing, top or bottom.
I followed up with: ‘I can't see the triple vein, even when I go and take another look at the plant itself, either on top or bottom of the leaf. But I can say that the leaves are at about 1cm intervals on the stem at an approximately 120° angle’.
GI tells me that the choice is now between ‘late goldenrod / giant goldenrod (Solidago gigantea)’ and ‘New York aster (Symphyotrichum novi-belgii)’. He says it is all down to whether or not the leaves ‘clasp’ the stem. I go and take another look at the plant,
I followed up with: ‘I go for clasping, particularly on the young leaves at the top of the stem’.
GI goes off goldenrods and goes on asters: ‘Choosing clasping completely shifts the diagnosis away from the Goldenrods and confirms you have a true late-summer blooming Aster’ – likely a New York aster/Michaelmas daisy (Symphyotrichum novi-belgii) or a purple-stemmed aster (Symphyotrichum puniceum). He says this it is all down to whether or not the stem is shiny or hairy. I go and take another look at the plant.
I followed up with: ‘Smooth and glossy’.
GI then summarises where we have got to – Michaelmas daisy – and I close by asking whether decurrent is a fancy word for ‘clasping’ and GI explains that it is not. I am convinced. ‘Decurrent’ being a term which cropped up just the other day and has cropped up in the past, but not to the point of remembering what it meant. See, for example, reference 4.
There was an irrelevant diversion about my birthday being Michaelmas Day, a quarter day in the old agricultural calendar.
But I failed to realise that ‘Michaelmas daisy’ is a bit of a generic term for a group of late flowering asters. The two at references 2 and 3 are not even in the same genus.
Notwithstanding, that will have to do for now. Let’s hope my plant survives to flower and settle the matter.
Conclusions
Things were helped along in this case by being able to nip out and take another look at the plant in question while we went along. Non-destructive testing.
GI is good at gliding over spelling mistakes and typos.
GL did make mistakes, some of them rather odd. He could also change tack rather abruptly, on the basis of new information.
Some of his phrasing and pronunciation is bad, if not plain wrong, which would be marked down in a classroom assignment.
But the conversation as a whole made me think of the hierarchic, binary keys to identification that you get in some gardening and botany books. Keys which usually depend on the user having mastered the botanical descriptive vocabulary. I suppose one could train an AI agent to build binary keys, somehow leveraging all that stuff apparently out there on social media sites like Reddit.
I have been reminded that while a photograph of a plant helps, it is not the same as a botanical diagram and it is often not enough, on its own, for an identification at species level. Identification which is a process rather than a one-hit thing. Some people argue that this is true of ‘truth’ in general. And, for that matter, consciousness.
I have not attempted to parse the Wikipedia description of the leaves at reference 2 – which did not read much like what I have got in the ground.
I have been reminded that flowers often provide more clues to identification than leaves – which is all we have here.
Notwithstanding, we end up with a reasonably good identification of Michaelmas daisy, aka New York aster; one would bet modest sums on it. Hopefully I will remember to come back to this conversation when the plant flowers, as hopefully it will in the Autumn – but I don’t go so far as to put a reminder in the calendar on my telephone!
Work in progress.
PS 1: Zomlefer reminds me that the asters are a large family, including all kinds of useful plants – like sunflowers – as well as daisies. And that there is also an aster genus, now much reduced from its once large size by taxonomic disturbances in the 1990s. Not much present help otherwise, but see reference 5.
PS 2: asters for the star shape of many of the flowers. Aster as in astral, from the Greek. A weak version of the marshmallow story of reference 6.
PS 3: the Guardian has been surprisingly anti-Starmer and pro-Burnham. I have only just realised that it is simply regressing, back to its Mancunian childhood. Perhaps I should start a chat group to promote its regressing in name too.
PS 4: I was not logged into my Google account while engaging with Google Images. This last now engages AI and that engagement may have been restricted.
References
Reference 1: https://en.wikipedia.org/wiki/Solidago_canadensis. In the aster family.
Reference 2: https://en.wikipedia.org/wiki/Aster_amellus.
Reference 3: https://en.wikipedia.org/wiki/Symphyotrichum_novi-belgii.
Reference 4: https://psmv5.blogspot.com/search?q=decurrent.
Reference 5: https://en.wikipedia.org/wiki/Aster_(genus).
Reference 6: https://psmv6.blogspot.com/2026/06/mainly-hollyhocks.html.
Group search key: aisk.
No comments:
Post a Comment