> 1.) First of all, I was wondering why you placed your subgen.
> Thelocalyx first, and not subgen.Regiae? Wouldn't you say the latter is
> probably most primitive? If I remember well, D.burmanni even has the same
> basic number of x=10 as subgen.Drosera. I don't remember now (and
> unfortunately I'm not in my lab and don't have my literature nearby) but
> aren't palinological and anatomical (wood) characters in D.regia more
> primitive than in other Drosera?
The primitive corner in _Drosera_ is not well resolved. There are
many character bridges and no clear picture as of the "most
primitive" species. Much of this distant history was severely obscured
by specific habitat adaptations, migrations, and climatic changes.
The vegetative makeup of _D.regia_ is rather primitive, indeed, but
the flower (with the exception of polllen) shows little deviation from
apparently higher members of _Drosera_. If we assume (I do so!) that
the shift from pentamerous to trimerous ovaries has occurred but once
in the genus, and if we consider the wide present distribution of
Thelocalyx, it appears to be a good candidate for a starting point
(i.e. in fact only a close derivative of the starting point) for the
whole genus. However, the key does not necessarily represent a true
phylogenetic tree (although I am rather convinced that it is closer
to such a thing than all precursors).
> 2.) Why was subgen.Lasiocephala placed before subgen.Drosera?
> Wouldn't you say the former would be better placed between the latter and
> subgen.Bryastrum, which also has subpeltate lamina (plus peltate?) and
> according to rbc-L sequence is closer?
Lasiocephala shares distinctive features with subgen. Drosera (basally
bifid styles, a very important synapomorphy in the latter!!),
Bryastrum (subpeltate laminae, but styles **NEVER** divided to any
degree), and Ergaleium (peltate laminae and divided styles but no
stipules). It seems to be close to the origin of all these three very
distinct subgenera and should therefore not be united with any of
them.
> 3.) Don't you think maybe subgen.Meristocaulis is closer to
> subgen.Drosera than subgen.Lasiocephala? After all, D.meristocaulis
> does have petiolate lamina
So have many other subgenera.
> and the single flowers on short scapes are found in
> several species in subgen.Drosera, such as D.felix
... which has ***divided*** styles.
> and D.uniflora.
It has a short but distinct scape and *repeatedly divided* styles.
> Also,
> the 3 undivided styles in D.meristocaulis seem closer to subgen.Drosera
NEVER.
> than the multifid ones present in subgen.Lasiocephala.
They are not multifid from the base (like in Phycopsis and many
representatives of Ergaleium) but repeatedly dichotomously divided
(like some representatives of subgen. Drosera, e.g. _D.aliciae_,
_D.oblanceolata_ etc.).
> 4.) Why create a whole subgen. for D.glanduligera? Wouldn't it be
> better placed as a section under subgen.Thelocalyx?
You want to do so? Well, this may be feasible technically but
Thelocalyx has (as mentioned above and in the key!) a pentamerous
gynoeceum. There are important taxonomists who would have lynched me
if I united Coelophylla (trimerous ovary!!!) and Thelocalyx and
separated the rest of the genus (trimerous ovary in almost all
species, some deviations in Bryastrum: you do not want to unite this
with Thelocalyx, do you?) from that assemblage.
> I don't remember
> floral morphology, but don't the annual habit, unusual yellowish color,
> similar leaf shape, and similarly-shaped outer tentacles with rectangular
> tips and wide flat stalks (is there a name for the 'stalks' of the
> glandular hairs?)
Yes, they are called "stalks", but they are *not* hairs but
emergences.
> suggest some relationship?
Yes but the pollen is different, and I think pollen morphology is
important in infrageneric classification in _Drosera_.
> Another thing about the
> outer tentacles is that in D.glanduligera they may bend in less than a
> second, faster than any other known Drosera, while in D.burmanni and
> D.sessilifolia they may bend in a few seconds only, also faster than the
> rest of the Drosera. Coincidence?
I don't know.
> 7.) The difference in leaf length you point out between
> D.sessilifolia and D.burmanni I believe is nonexistent. Apparently
the
> main differences are the leaf shape (which you mentioned), flower
color
> (pink and white, respectively)
This is a weak character not reliably constant. I have tried to avoid
it where possible in the key.
> 5.) As to the inclusion of D.brevifolia, D.uniflora, + D.trinervia
> in sect.Ptycnostygma, you'd already mentioned this to me in the beginning
> of the year and I still find it a bit hard to swallow.
I know.
> D.uniflora and
> D.trinervia both look a bit odd in this group, but I don't know enough on
> them to argue with you. D.brevifolia is a delicate annual which I find
> hard to be closely related to something as large and magnificent as
> D.cistiflora and D.pauciflora. And as I've mentioned before, it is VERY
> similar to type D.montana in the juvenile stage.
Maybe. But the stipules are very distinctive even in very young
plants.
> The fact that type
> D.montana and several other Brazilian Drosera have been observed to go
> dormant as roots during the dry season may show a link between sections
> Oosperma and Ptycnostygma, maybe indicating that they could be united into
> a single section. A bit confusing at the moment.
Yes, indeed. This is the reason why I grouped them in sections, not
subgenera.
> 18.) Why did you decide to create sect.Osperma based only on
> seed morphology, especially if you believe both groups are not
> monophyletic. In D.villosa the seeds are somewhat fusiform, but in
> the D.villosa var. graomogolensis which has not been published yet
> they are ovoid. Wouldn't this show that seed shape can not be used
> for such a deep separation?
In fact, Oosperma is rather certainly inhomogeneous,
and rearrangements will probably be necessary (like I have stated
already in the article!). But I had to find a formal solution and had
the least (NB: *not* none!) stomach ache with the version presented.
Enough work for future researchers left.
> 6.) What do you mean by calling D.capillaris a "collective
> species"? Is it as complicated a taxa as D.spatulata, D.montana, and
> D.leucoblasta?
Yes. Some taxa in America are obviously misclassified, and there is
no unanimous conclusion where _D.intermedia_ ends and where
_D.capillaris_ begins.
> Where does D.tenella fit in this? It's in your article but
> I could not find it as valid on your CP list on the web.
It is not entirely clear if this taxon can reliably be distinguished
from _D.capillaris_ (with which it is united so far in the list). I
have tentatively cited the alleged differences from the type in the
key.
> 8.) What did you mean by putting D.cistiflora twice, once followed
> by "s.l." and once by "s.str."?
s.l.=sensu lato (broad sense)
s.str.=sensu stricto (srict sense)
> The same was done for D.capillaris, D.montana, and maybe others
e.g. _D.leucoblasta_.
> 10.) As to D.chrysolepis leaf shape, I'd say it's more lanceolate
> than spathulate.
The key says "oblong spathulate". The important point in the key is
that it is decidedly "rounder" than "linear".
> 11.) As to D.villosa scapes, only those of the ex-D.ascendens
> are glabrous (and even then only in the lowermost parts). Type
> D.villosa have glandular + simple hairs thickly covering the scapes
> and D.ascendens can be found with many variations of hair
> distribution on the scapes.
I would propose that you publish your stuff (or send me the
*complete* manuscript with all systematic conclusions) on Brazilian
species before we start discussing details. It seems that there will
be severe rearrangements especially in this corner (you seem to want
to reestablish all described microspecies and add some new ones). I
can only use the data available to me. I can and will not comment on
taxa like "ex-D.ascendens" & al.
> 12.) I liked very much your use of both subsp. + var., for
> allopatric and sympatric species. I was just wondering how I should refer
> to a few taxa now. For example the ex-D.microscapa. Can we just call it
> D.occidentalis var.microscapa
Yes.
> or does it have to be D.occ.subsp.occid. var.microscapa?
No, not necessarily.
> 13.) As to the true subsp.omissa, couldn't the 'circular stigma'
> be a misidentification at the time,
No.
> analysing herbarium specimens
Which our friend from Australia has expressedly *NOT* done in this
case.
> later where the stigma were maybe not so visible?
It is very well visible, and not shaped like in what was called
"D.nitidula omissa" by Allen Lowrie (cf. an article by Dr. Martin
Cheek in CPN specifically regarding this problem).
> Maybe they are like in Lowrie's book.
No. They are *not* like in his book!
> What about the rest of the morphology, isn't it closer
> between these two plants than it is between the other varieties
> (var.nitidula, var.allantostigma, and var.leucostigma).
Not at all.
> And is the
> subsp.omissa you mention allopatric, to be placed as a subsp.?
Yes. BTW "D.ericksoniae" is a synonym.
> 14.) What do you mean by synanthropous?
Dispersed together with man.
> 15.) What do you mean by sheating leaf bases?
In the exstipulate Arcturia, the leaf bases fulfil the job of the
stipules to protect the bud. They are broader as in the species with
stipules and sheat the centre of the rosette.
> 16.) Isn't auriculata sympatric with peltata, and thus shouldn't
> it be a variety?
It is perhaps partly, but AFAIK, _subsp.auriculata_ was never found in
W AU. It is usually regarded to be the E AU part of the taxon. It
should better not be a variety.
Kind regards
Jan