|
[Page
16]
David Hunt
The Manse, Chapel Lane, Milborne Port, Sherborne, DT9 5DL, U.K
Following the
preliminary discussion on this topic (08 11: 23-29) and the six hours
of discussions at the recent ICSG meeting (see this issue pp. 4-8) it
seemed to me we had made sufficient progress to attempt a tentative
re-working of Ritter‘s classification of the genus (Ritter, Kakteen in
Südamerika 3: 1048. 1980). First, I made a free translation of his text,
using the abbreviations devised by Urs Eggli some years ago for the
projected 108 Lexicon of Succulent Plants and adapted by me for the
Cactaceae volume:
Subg. Pilocopiapoa
(Ritter) Ritter
|
[Sect.] 1 |
Bo
hard; r non-tuberous; ri high, etuberculate; ar
>1 cm ø pc regular, with very small elongate scales; scax
with dense trichomes surrounding the whole pericarpel and perianth;
tep <1/4 flower-length; sta relatively short; hilum basal,
impressed [solaris]. |
Subg.
Copiapoa
Bo
soft to hard; r tuberous or non-tuberous; ri less high,
seldom etuberculate, in old plants occasionally fully resolved into
tubercles; ar < 1 cm Ø, pc almost always with broad scales
at the apex and sometimes a few smaller scales below, occasionally
naked; scax naked or occasionally with very small tufts of hairs;
tep >1/2 flower-length, naked; sta relatively long; hilum
subbasal or basal.
|
Sect. 2. |
Bo relatively large, hard, grass-green or
slightly grey-green, sometimes brown, never pruinose; ri few;
sp subulate. |
| |
Series a. |
Southern group. R non-tuberous or only very slightly tuberous
(cuprea, dura [= echinoides], marginata, bridgesii] |
| |
Series b |
Northern group. R tuberous, hard (hornilloensis,
desertorum, rupestris, rubriflora) |
|
Sect. 3. |
Bo relatively small, soft- or not very
hard-fleshed grass-green to grey-green or black, not or only very
slightly pruinose; r strongly tuberous with narrow neck;
ri not numerous, always notched (crenate); sp acicular,
straight or slightly curved (longispina, chaniaralensis,
esmeraldana, taltalensis, humilis, paposoensis, vari(i)spinata,
tenuissima, tocopillana). |
|
Sect. 4. |
Bo small to fairly large, in some species
relatively very long, somewhat soft- or not very hard-fleshed, green
to grass-green, sometimes white-pruinose; r strongly tuberous
with narrow neck; ri numerous or fairly numerous, strongly
crenate or not crenate, sometimes with a “chin” beneath the areole;
sp stout-acicular or thin-subulate, occasionally lacking (pendulina,
coquimbana, pseudocoquimbana, vallenarensis, fiedleriana,
alticostata, echinata, megarhiza, calderana, cinerascens, hypogaea,
mollicula, grandiflora, montana, olivana, rarissima, boliviana [sensu
Ritter; = atacamensis]). |
|
Sect. 5. |
Bo large, hard, grey-green, usually white-pruinose;
r non-tuberous; ri numerous to very numerous, slightly
or not crenate. |
| |
Series a. |
Southern group. Ri moderately numerous, the furrows very
tortuous; ar oval, fairly widely spaced; seed
tuberculate (carrizalensis, dealbata). |
| |
Series b. |
[no
name]. Ri moderately to very numerous, furrows straight or
less tortuous than in series a. (except C.
serpentisulcata);
ar round, more close-set; seed
almost smooth (serpentisulcata, columna-alba, longistaminea,
melanohystrix, cinerea, tenebrosa, gigantea, krainziana, eremophila). |
[Page 16]
Currently, it seems accepted that
Copiapoa solaris
deserves a section or subgenus to itself. For the rest, I have the
impression that the texture of the plant body (whether “hard“ or “soft“) is
a more significant feature than its colour and degree of pruinosity, which
seems to have dictated the linear order of Ritter‘s sections. This
impression arises from Nigel Taylor‘s observation that mucilage is present
in the stems of the so-called soft-bodied species but few of the hard-,
doubtless explaining, at least partly, the difference in texture. There is
also a correlation to be observed between body-texture and the nature of the
roots, tuberous roots being more characteristic of the soft-bodied /
mucilaginous species. For this reason, my first move would be to place the
hard-stemmed groups adjacent to one another, i.e. to place Section 5 next to
Section 2. Not to get confused, we now need names rather than numbers for
all the Sections, and the group names adopting during the “Copiapoathon“
will do for the time being, i.e.:
(2). MARGINATA group (Bo hard, never pruinose;
r non-tuberous or tuberous)
(5). CINEREA group (Bo hard, usually pruinose;
r non-tuberous)
(3). HUMILIS group (Bo soft to hard, not or
slightly pruinose; r tuberous)
(4). [THE REST] (Bo somewhat soft to hard,
pruinose or not; r tuberous)
From our discussions, it
seems the MARGINATA group contains four or five species. Of
these,
- C. marginata
(incl. C. bridgesii) and C. sp. nov. “Botija“,
- C. desertorum
and
- C. rupestris
(incl. C. hornilloensis, C. rubriflora) all have tuberous
roots.
- C. echinoides
(Lem. ex Salm-Dyck) B. & R. was not recognized by Ritter but is now
regarded as the prior name for his C. cuprea and C. dura,
both placed in this Group. lt differs from the others in having
fibrous roots and also has quite distinctive seeds with high-convex
testa-relief. For these reasons Fred Kattermann feels it merits a
separate group.
The CINEREA group
is now regarded as containing five species:
- C. dealbata
(incl. C. carrizalensis, ?C. malletiana),
- C.
serpentisulcata,
- C. cinerea
(incl. C. melanohystrix, C.
tenebrosa, C. eremophila),
- C. haseltoniana
(incl. C. gigantea),
- C. krainziana.
Excluded: C. longistaminea (allied
to C. calderana?).
The HUMILIS group includes
- the variable C. humilis (incl.
C. longispina, C. chan(i)aralensis, C. tocopillana,
C. paposoensis, C. taltalensis)
- C. varispinata.
Excluded: C.
esmeraldana, C. tenuissima nom. inval.
During our discussions
“THE REST” were divided into four
subgroups:
-
C. calderana
(including C. atacamensis as a
potential subspecies) and C. longistaminea.
-
C. cinerascens
and C. grandiflora.
-
C. megarhiza, C.
echinata, C. fiedleriana, C. coquimbana
(incl. C. pendulina, C.
pseudocoquimbana, C. vallenarensis, C. alticostata, ?C. mollicula, C.
esmeraldana)
-
C. hypogaea
(incl. C. tenuissima, C. rarissima), C.
laui, C. montana (incl. C. olivana).
[page 17]
In his account of the genus
for the European Garden Flora (Taylor 1989: 253), Nigel Taylor
treated most of these species, excluding C. hypogaea but including
C. montana,
under C. cinerascens. C. calderana and C.
grandiflora were regarded as “probably only varieties” of C.
cinerascens itself. In that the three subgroups he treated under
C. cinerascens, and the species included in them, are allopatric
this grouping is a logical one; it meets the “rule of thumb” that taxa
belonging to a single group are not sympatric, i.e. they do not normally
grow together. The distribution of the C. hypogaea subgroup
(including C. montana)
does, however, overlap with that of what I will now
call the CINERASCENS group, and in that the relationships of C. hypogaea
and its allies with other species are debatable, I believe it would
preferable, for the time being, to treat them as a separate group, i.e. the
HYPOGAEA group.
In the CINERASCENS group
(as delimited here), the status and relationships of both C.
atacamensis and C. longistaminea remain under discussion.
When Fred Kattermann tested cultivated plants of C. calderana FK
32 it did not have mucilage, but a seedling of FK 46 (var.
longispina) did. He therefore considers the presence/absence of
mucilage unreliable as a means of separating C. atacamensis and
C. calderana from C. marginata.
C.
longistaminea
has usually been regarded as belonging to
the C. cinerea Group, but grows sympatrically with C. cinerea
var. columna-alba and has mucilage. For this reason a
relationship with C. calderana was suggested during the
Copiapoathon, but it (C. longistaminea) does not have tuberous
roots.
I
conclude that subg.
Copiapoa can
be divided into five groups, as indeed was implied by Nigel
Taylor in his identification key to the genus (Taylor 1989, l.c.). The
key is reproduced here, with group names replacing individual species:
|
