In message , Nick Maclaren
writes

In article ,
Janet Tweedy writes:
|
| I am now totally confused as I have seen pictures of the darker flowers
| on the web and in books and also pictures of the baby/sky blue colour I
| saw last year. Could there be different strains of this then?
| The Ipomea in Swallowfield had set what looked like early seed pods and
| I planted them but nothing germinated.

Given that David has confirmed my memory, if it has seed pods, it is
NOT I. indica. No ifs or buts. I. indica does not set seed.

I found a claim in Am. J. Bot. That I. indica is self-incompatible. If
this is correct then it is possible that it does set seed in the wild,
but that the cultivated stock is a single clone. (For comparison,
Lavatera 'Bicolor' is resolutely sterile in normal cultivation, but does
produce seed when grown alongside Lavatera acerifolia. I want to obtain
wild-collected seed of Lavatera maritima to test the nature of Lavatera
'Bicolor' further.)

There are something like 500 species, of which at least half a dozen
are in cultivation and will grow outside in the UK during the summer;
quite possibly, several of the others will, too. The cultivars of
the commonly grown species (see below) are very variable in size and
colour, so odds on it is one of them.

There has been a fair amount of botanical renaming, too, and I have
never worked out what the situation is between I. nil, I. purpurea,
I. triloba (a synonym of I. nil?) and I. tricolor. I believe that
there are also hybrids. I can state definitely that there is a HELL
of a lot of confusion on the Web and in books over this, and even
quite respectable books conflict badly with each other, so I am in
good company.

I can't really help with this, but after a little digging. I find that =

The true Ipomoea triloba L. is more closely related to the sweet potato
(subg. Batatas) than to the other species. Ipomoea triloba Thunb. Is
given by IPNI (http://www.ipni.org) as a synonym of Ipomoea hederacea.
(The existence of an Ipomoea purpurea var triloba may add to the
confusion, regardless whether that plant is correctly placed in I.
purpurea.)

I found a citation, requiring JSTOR access, to Austin, Nomenclature of
the Ipomoea nil Complex (Convolvulaceae), Taxon 35(2): 355-358 (1986),
which might clarify things further.

No doubt you'll be skeptical, but a molecular paper on the subject is
Miller et al, Systematics of Ipomoea subgenus Quamoclit (Convolvulaceae)
based on ITS sequence data and a Bayesian phylogenetic analysis, Am. J.
Bot. 91(8): 1208-1218 (2004). See http://wwww.amjbot.org

From then above cited paper, I. hederacea, I. nil and I. indica are a
group of closely related species. (Reading between the lines lumpers may
be tempted to place them in a single species.) Again from the cited
paper, I. hederacea selfs, I. nil is self-compatible and I. indica
self-incompatible.

The resolution of the study isn't great, but the group of species
consisting of I. purpurea, I. pubescens and I. lindheimeri appears to be
separated from the preceding. I. tricolor is more distant. There's no
indication in IPNI of any confusion of the application of the names I.
purpurea and I. tricolor, though the usual lumper/splitter issues may
well apply.

Regards,
Nick Maclaren.



--
Stewart Robert Hinsley