ECHINOPSIS REVISITED, April 28, 2004.
By Bob Schick
I. INTRODUCTION
For the past two years, I've been carrying out a study on the
genus Echinopsis, s.l. (sensu latu, in a broad sense) with the
aim of defining its component groups as full genera or subgenera.
I had intended to maintain a Web site to show my progress but,
for various reasons, I have not done that. Thus, this post to
the Cactus etc. Web site to briefly describe the current state
of my study (and just in case something happens to me). The
study is based solely on the collection of Echinopsis, s.l., at
the Huntington Botanical Gardens in San Marino, California.
Though the collection is extensive, there are many forms not
represented. Because of the only partial representation and
because my understanding of some floral structures is still in
the learning process, this report should be regarded as a work in
progress.
Several major groups of species are represented in Echinopsis,
s.l. My original intent was to treat these groups as separate
genera, because Echinopsis, s.l., in my view, was a too massive
and overly diverse genus. Moreover, I was able to define the
major groups by using the floral and seed characters not usually
utilized for the genus. Then I realized that dividing
Echinopsis, s.l., into the several genera would mask their
presumed common evolutionary origin, so I decided instead to
treat all the major groups (except for a new genus containing
leucantha) as subgenera of Echinopsis, s.l. A further advantage
of recognizing only a single genus was that it gave me more
freedom to subdivide Echinopsis, s.l., into major subgeneric
categories. But now, in a compromise, I've decided to recognize
three genera in Echinopsis, s.l., instead: Echinopsis,
Trichocereus and the new genus represented by leucantha. This
divides Echinopsis, s.l., into its two major groups--Echinopsis
and Trichocereus--while still emphasizing the distinctness of
leucantha and continuing to show, at least to a considerable
extent, the evolutionary cohesiveness of Echinopsis, s.l., and
permitting me to recognize additional subgenera with a clear
conscience.
The several groups, with some of their component species as
examples, are listed below:
ECHINOPSIS, S.L.
A. genus Echinopsis
1. subg. Echinopsis (e.g., eyriesii, oxygona; Mamillosa group)
2. subg. Pseudolobivia (e.g., Obrepanda group, ancistrophora)
3. subg. (new) for subdenudata
4. subg. Lobivia (e.g., arachnacantha, aurea, cinnabarina, haematantha)
5. subg. (new) for Pseudolobivia x Lobivia hybrids (e.g., calorubra,
rojasii)
6. subg. (new) for Maximiliana group
7. subg. Chamaecereus (chamaecereus, saltensis)
B. genus Trichocereus
8. subg. Helianthocereus (e.g., bruchii, formosa, huascha)
9. subg. Trichocereus (e.g., candicans, spachianus, thelegonus)
10. subg. (new) for Trichocereus x Helianthocereus hybrids
(pseudocandicans?)
C. new genus (e.g., leucantha, rhodotricha)
To simplify presentation of data, I will indicate generic or
subgeneric status only for Echinopsis and Trichocereus, e.g.,
genus Echinopsis and subg. Echinopsis. Lobivia is cited as an
ancestral group to the other subgenera because of its great
diversity and the apparent primitiveness of some of its features.
Regarding terminology, I will use the terminology in Edward
Anderson's, "The Cactus Family", whenever practical, and if there
is a choice between technical and less technical variants of
terms in the book, I will choose the less technical. In some
cases, though, I will have to employ new terms.
I will now very briefly discuss some characters I've used to
define the genera and subgenera which might not be familiar to
readers of Cacti etc. Unfortunately, photos or illustrations are
necessary to adequately visualize them, but this is not possible
on a mailing list. Instead, as a poor substitute, I will refer
to the photographs in five popular books to illustrate at least
some of the characters. The books are, (a) "The Cactus Family,"
by Anderson, (b) "Cacti," by Innes and Glass, (c) "Cacti, the
Illustrated Dictionary," by the Preston-Mafhams, (d) "The
Encyclopedia of Cacti", by Cullmann, Goetz & Groener, and (e)
"Kakteen," by Goetz and Groener.
II. STAMENS
1. Rim and throat stamens.
One of the hallmarks of Echinopsis, s.l., is the presence of two
sets of stamens, one an upper series and the other a lower
series. There has been no consistent terminology for either set
of stamens and terms that are used may not be entirely clear
without knowledge of their definitions. To make the terms more
accessible and also proper, I will call the upper series, *rim
stamens*, since they originate on the rim of the throat, and the
lower series, *throat stamens*--perhaps not an original
term--since they originate within the throat itself. The rim
stamens are arranged in a circle around the throat, their
orientation varying from an erect position to forming a rather
flat horizontal spray against the petals, while the throat
stamens take on different configurations in the different groups.
2. Radial symmetry of flowers.
The stamen arrangement of lobivias is probably the most primitive
of Echinopsis, s.l. In this type of symmetry, the rim stamens
are all similarly curved so that when arranged as a horizontal
spray the petals, form a more or less perfect circle, while the
throat stamens form a broad swathe completely circling the
throat. For photos showing the rim and throat stamens, see
Anderson, Echinopsis coronata, p. 264 and Echinopsis pentlandii,
p. 276; Innes and Glass, Lobivia jajoiana, p. 146; Preston-Mafhams,
L. cinnabarina, p. 76 and L. jajoiana, p 79; Cullmann et al.,
Lobivia spp., p.197-200; and Goetz & Groener, Lobivia spp., p.
189, 192, 194-196.
Notice that the rim and throat stamens are arranged so that they
extend out from or around the flowers' central axis in a radial
pattern. This is obvious for the rim stamens when they form a
more or less flat spray over the petals and not so obvious for
the throat stamens since they are arranged cylindrically up and
down the throat. Nevertheless, the symmetry for both is
*radial*, i.e., any plane, whether vertical, horizontal or
slanted, that passes through a flower's central axis will divide
the flowers into more less symmetrical halves. Unfortunately,
many of the above photographs do not clearly show the radial
symmetry, especially for the throat stamens, because the photos
were shot to one side of the flowers and not directly above.
3. Bilateral symmetry of flowers.
The throat stamens undergo a remarkable transformation in subg.
Echinopsis, Pseudolobivia and subg. Trichocereus, the rim
stamens also becoming altered, though not as markedly as the
throat stamens. Regarding the throat stamens of these subgenera,
those in the ventral half (see below for explanation) of the
throat have grown in length to form a prominent exserted
(projecting) cluster, the *throat stamen cluster*, while the
throat stamens in the dorsal half (see below) of the throat tend
to decrease in numbers, usually greatly so (Pseudolobivia), or
disappear entirely (subg. Echinopsis and subg. Trichocereus).
The stamen symmetry in these groups is *bilateral*, i.e., a
flower divided down its vertical midline results in a right and
left half, which are more or less mirror images of each other.
Some structures of a flower whose stamens are bilaterally
symmetrical can be described as having a top-side and a bottom
side or, to be more concise, a *dorsal* side and a *ventral*
side.
I must note I'm using dorsal and ventral here more like a
zoologist, whose interest is in fish or four-legged animals,
than a botanist, although Barthlott and Hunt, in their book on
cactus seeds, do refer to a dorsal and ventral side of seeds. As
a zoological example, the fin of a fish located on the top side
of its body is the dorsal fin and the fin located on the fish's
bottom side is the ventral fin. The dorsal fin was made famous
in the movie, "Jaws", as it projected menacingly above the
surface of the water.
In the Trichocereus photographs of the following references,
notice that the throat stamen cluster is quite prominent and
restricted to the ventral (bottom) half of the throat and there
are no stamens in the dorsal (top) half: Anderson, "Echinopsis"
macrogona, p. 273 and "E." spachiana, p. 280; Innes & Glass,
Trichocereus spp., p. 297; Cullmann et al., Trichocereus
'theleflora', p. 305, Trichocereus hybrid, p. 306 and T.
candicans, p. 307; and Goetz & Groener, Trichocereus spp., p.
301, and Trichocereus-Hybriden, p.302. The Preston-Mafhams'
photos do not show the throat stamen cluster clearly. But what
about the photo of "Echinopsis" arboricola on page 257 of the
Anderson book? The throat stamen cluster appear to be located in
the dorsal half of the flowers! However, Myron Kimnach informed
me that the photo is actually upside down.
An interesting modification of the filaments of the throat stamen
cluster occurs in subg. Echinopsis and Pseudolobivia. An
extensive basal portion of the filaments in these subgenera have
an adhesive surface that serves to hold the stamens together as a
compact group in the throat. Many throat stamens in subg.
Trichocereus are also held together by the stickiness of their
filaments to form a similar adherent cluster, but other stamens
run courses through the throat independent of the adherent
cluster so that, unlike subg. Echinopsis and Pseudolobivia, the
cluster is not compact in the throat, but appears as a broad,
loose assemblage of stamens.
The rim stamens in the three subgenera also show bilateral
symmetry: Looking at a flower from a top view, the rim stamens on
the ventral half of the flower tend to be hook-shaped while those
on the dorsal half are merely weakly curved. As a result, when
the rim stamens are arranged in a horizontal spray over the
petals, they do not form a perfect circle, as in radially
symmetrical flowers, but a circle that is somewhat flattened on
its ventral side. The flower tube of bilaterally symmetrical
flowers is sometimes also modified: Its ventral side is slightly
longer than the dorsal side, so that the surface of the flower is
slightly inclined.
Bilateral symmetry is obviously a very important feature in
Echinopsis, s.l. But why develop such a radical modification of
the stamens in the first place? An obvious answer seems to be
that bilateral symmetry functions as part of the pollination
syndrome (strategies for attracting particular kinds of
pollinators) and serves to attract a different class of
pollinators than those of radially symmetrical flowers.
III. THE THROAT OF ECHINOPSIS, S.L., FLOWERS
To see the throat divisions clearly, a flower must be cut
longitudinally in two. The throat in all Echinopsis, s.l., is
transversely partitioned into two or three horizontal sections.
In those with three sections, the divisions, from the base
upward, are: the nectar chamber, the zone of throat stamens and
the distal throat division. (a) The *nectar chamber* is
typically a distinctly defined, hollow compartment at the base of
the throat. It is a very long, narrow cavity in all subg.
Echinopsis and Pseudolobivia; very short to only moderately long
or completely absent in lobivias; very short and broadening
toward its tip in Chamaecereus; very short to absent in
Helianthocereus; and in subg. Trichocereus, very short to
moderately long, usually the former. The absence of a nectar
chamber doesn't necessarily mean that nectar is not secreted by a
flower, for at least in some cases, nectar is still produced,
though perhaps not in the volume of those flowers in which the
chamber is developed. (b) The *zone of throat stamens*, between
the nectar chamber and distal throat division, is the region
where the throat stamens originate as free structures. The
division is present in all Echinopsis, s.l., of course, since all
have throat stamens. (c) The *distal throat division*, the upper
section of the throat, is a hollow chamber whose dividing line
with the zone of throat stamens is not always distinct. It is
greatly reduced or essentially absent in some species of Lobivia
and in Helianthocereus and Chamaecereus.
In many Echinopsis, s.l., a terminal portion of the distal throat
division appears to be modified to serve as a nectar guide.
There are many different types of guides, but I'll mention only
one of the common types. It is the *throat-circle*, a term
coined by Buxbaum, found in lobivias and helianthocereus. The
throat-circle appears as a narrow colored ring that circles the
inner wall of the distal throat division along the division's top
edge. Its color differs from and contrasts to the portion of the
distal throat division below it and to the color of the flower
petals situated adjacent to the circle. The throat-circle
apparently serves as an attractant for day-flying pollinators to
an opened flower and/or to orient a pollinator, already on a
blossom, to the throat itself. The throat-circle is often set
upon a slight bulge of the throat wall that circles the tip of
the distal throat division. I term this bulge, the *annulus*.
The annulus is variably developed in Echinopsis, s.l., but is
most prominent in lobivias. It apparently functions to make the
throat-circle more evident to a pollinator. In some cases,
though, an annulus is not necessary at all, for a throat-circle
may be present when an annulus is greatly reduced or apparently
absent. The flared throats of some members of some lobivias,
such as kuehnrichii, which have a throat-circle, would appear to
have no need for one since much of the white of the throat is
exposed due to the flaring. (The flaring could act as a nectar
guide in itself, but possibly just functions as a platform for
insects, such as butterflies, to rest upon while feeding on
nectar.)
Examples of a throat-circle are shown in the following
references: Anderson, Echinopsis backebergii, p. 258 (the partial
yellow circle); Innes & Glass, Lobivia jajoiana and L. marsoneri,
p. 146 (the partial or nearly complete yellow or bright purple
circles); Preston-Mafhams, L. acanthoplegma and L. acanthoplegma
v. patula, p. 74 (the white circles); L. amblayensis, p. 78 (the
partial yellow circle), and L. jajoiana, p. 79 (the purple
circles); Cullmann et al., L. pentlandii, p. 191 (the partial
white circles), L. jajoiana, p. 197 (the partial purplish
circles, some appearing white due to glare) and L. kuehnrichii
group, p. 198 (the partial yellow circles); Goetz and Groener, L.
pentlandii, p. 189 (the partial white circles), L. jajoiana (the
partial purplish circles, some appearing white due to glare) and
L. kuehnrichii (Formenkreis), p. 195 (the yellow circles).
If you look closely at the throat-circles in the above
photographs, you will see they actually consist of a circular
line of tightly packed, small squares or rectangles, these giving
the throat-circles their color. I call the squares and
rectangles *trace terminals*. They are the expanded tips of rim
stamen traces that are covered by a thin, transparent layer of
throat tissue. *Stamen traces* are vascular bundles located
within the tissue of the floral tube itself. The stamen traces
extend as narrow filaments from the base of the throat upward to
various places in the tube, where they exit as free stamens. The
traces of throat stamens exit in the zone of throat stamens and
the traces of the rim stamens exit from the end of the throat or
from the throat-circle, when present.
The throat-circle is of special interest in the Marsoneri group
of Lobivia, which includes jajoiana. In many forms in the group,
the throat, including the throat-circle, is covered by a thin,
stiff, transparent layer which is highly reflective. An insect
looking down at a flower would see a series gorgeous, iridescent
purple trace terminals that flashed on and off as the flower
fluttered in the breezes. This apparently is a great signal to
attract six-legged pollinators. Buxbaum used the term,
*hymen*, to describe the reflective throat-circle of jajoiana.
Actually, the hymen is not membranous at all, despite the name.
Moreover, it is just a special case of the throat-circle that is
developed only in some forms of the Marsoneri group. Various
authors (including myself) have applied this term to the
throat-circle of other Echinopsis, s.l., but strictly speaking,
hymen should be applied only to the condition in the Marsoneri
group.
I'd like to inject a personal anecdote about traces here. Many
years ago, when I drove through the southeastern United States, I
came across national monument signs indicating historic traces.
I was intrigued by the term "traces"--romantic and all that.
These traces, I later learned, were narrow trails made by the
passage of people or animals through the early 19th century
wilderness of the old southeastern United States (considered the
southwest, then). They were main routes before roads and rail
lines were built. The one I still remember seeing is the Natchez
Trace which, according to the Britannica, was an Indian trail
that extended from Alabama to Tennessee and was the most
important highway in the Old Southwest. Well, a couple of years
back, when I started dissecting echinopsis flowers, I noticed the
internal vascular trails of stamens in floral tubes. I'm not a
botanist, my background being in Entomology and Arachnology, so
in my ignorance, I thought I needed a term for these vascular
trails. The Natchez Trace immediately came to mind and, in
reports on my research, trace was the word I used. No one
objected to the term. Then I found out why when I read "The
Cactus Primer" by Gibson and Nobel. Trace was actually the
accepted botanical term for those vascular bundles! I still
smile when I think of the coincidence. Apparently, botanists
were also fascinated by the Natchez trace.
IV. THE FRUITS OF ECHINOPSIS, S.L.
Cactus fruit seeds are connected to the inner walls of the ovary
by cord-like bundles called *funiculi*. In Echinopsis, s.l., at
least, the funiculi comprise *closely pressed mucous-filled
globules with a single vascular trace running through each
funiculus. Two major types of fruits may be recognized in
Echinopsis, s.l., based upon the fragility of the membrane
enclosing the globules.
1. Fragile funicular membranes and wet fruits.
A fragile-type of membrane is found in genus Trichocereus. At a
fruit's dehiscence (splitting open) in this genus, the membranes
of the globules break down within only a day or so after exposure
to air, releasing their mucousy contents into the interior of the
fruit. The membranes also show their fragility when subjected to
a mechanical stimulus. I can manually rupture the membranes of a
newly dehisced fruit by an even very gentle prod with a sharpened
probe. (The membranes of the outermost funiculi, on the other
hand, persist to form a thin but rather tough outer layer that
prevents the mucousy contents of the fruit from escaping.) This
type of mucous-filled fruit is described as *wet*.
2. Resistant funicular membranes and dry and semi-dry fruits.
In the genus Echinopsis, the membranes of the funiculi are tough
and relatively resistant to exposure to air and even to my
pricking with a fine needle. Two types of funiculi are found in
this genus: (a) In subg. Echinopsis, Pseudolobivia and most
lobivias, the funicular membranes do not break down after a
fruit's dehiscence. Rather, they dry out within a few days, the
funiculi shriveling to form irregular dry cords. The mucousy
contents of the funicular globules apparently have simply
evaporated. This type of fruit is described as *dry*. (b) In
some lobivias, the mucousy contents of the funicular globules are
apparently of a different type and the globules remain turgid
with mucous for a long period after dehiscence. This type of
fruit is apparently referred to as *semi-dry* by Anderson. I
still have to make observations on semi-dry fruits to compare
them with the dry type.
V. THE SEEDS OF ECHINOPSIS, S.L.
I used the book, "Seed-diversity in the Cactaceae, subfam.
Cactoideae," by Barthlott and Hunt as a reference to the
structure of seeds. It is a must for anyone interested in the
use of seeds in taxonomy. I do deviate, though, from some of the
terminology in the book by changing the name of the "border"
(Saum in German) to *collar* and by considering a seed to be
divided into a usually narrow collar and broader *body*. The
collar is the usually narrow portion of a seed bordering the
hilum-micropylar region (HMR), this region a part of a seed upon
which the funiculus is adpressed, though the seed's actual
connection to a funiculus is though the hilum (a simple cavity in
the HMR). To study the seeds, I used my old stereoscopic
microscope with magnifications of up to 150 diameters. Some
characters I found useful for taxonomy were not covered in the
Barthlott and Hunt book. They are explained as follows.
1. Cuticular striations (microrelief) of the body cells
Three types of cuticular striations were apparent as seen under
high power with my dissecting microscope. (a) In one type, found
in subg. Echinopsis, the body cells are flattish and, owing to
reflections from my microscope lamp, dotted with sparse white,
and often blue or green pinpoints of light amidst barely
perceptible cuticular folds. I found myself staring at these
cells in fascination and could not help but to imagine another
world down there, a quaint city at nighttime with homes lit by
dim lights. How to name this fantasy land? *"Micro-city"*, I
decided. At low power, the luster of the cells showing
"Micro-city" cuticular striations is semi-matte or matte. (b) In
Pseudolobivia and Pseudolobivia x Lobivia hybrids, the body cells
are low-domed and awash in barely discernable reflected faint
specks of white light like distant stars on a clear night and, as
if in the background, the surface of the cells are dotted with
nano-sized punctations. I called this type of striation,
*"Starry Night"*. Groups of "Starry Night" cells, under low
power, appear mirror-like as they brightly reflect light of the
microscope lamp. (c) In the Subdenudata subgenus, the testa
(hardened seed coat) is uniquely glass-like in appearance and
lacks striations. The luster of these cells under low power is
glossy. I must mention that the appearance of an unprepared seed
under a dissecting microscope, especially the cuticular
striations and color of the HMR and other structures, is quite
different from that of a seed seen with a scanning electron
microscope and might be useful as an adjunct to an SEM.
2. "Magmatized" multicellular sculptures.
Multicellular sculptures are depressions of the testa involving
several or many adjacent body cells. Two types of sculptures may
be recognized in Echinopsis, s.l.: rounded craters and elongate
rills (like the rills of the moon). In some subg. Echinopsis
with multicellular sculpturing and in subdenudata, the interior
of the seeds seems to be a source of an orangish, magma-colored
light. In subg. Echinopsis, the floor of the craters and rills
are handsomely lit with this magma color. The cause of the
color, unfortunately is not very romantic. It is apparently a
case of neoteny, i.e., the seeds retaining their immature color.
Regretfully, the magma color eventually fades. A bright
magma-colored light also appears to emanate from inside of the
translucent seeds of subdenudata. Gorgeously jewel-like, this
light seems to illuminate the entire glass-like seed's surface
from below, the spaces between the conical portions of the cells
lit more brightly. This magma color, much to my regret, also
fades after a few months.
3. "Golden halo".
The HMR in Echinopsis, s.l., is a flat, mostly soft-tissued
structure, usually whitish in color and either recessed below the
rim of the surrounding testa or situated about level with the
testa's rim. In some subg. Echinopsis (Sucrensis group), the
HMR is produced to form a broad, yellowish donut-shaped
structure that just barely projects beyond the testa of the
collar. I can find no reference to this structure in the few
sources I've read, so I angelically refer to it as the "golden
halo". A somewhat similar, but more striking modification of the
HMR occurs in the Mamillosa group of subg. Echinopsis. The HMR
here is also yellowish, but nearly the entire HMR, not just the
periphery, is puffed out to extend well beyond the collar.
Although not particularly halo-like, I'll also refer this
structure as a "golden halo" so as to not coin still another
word. The lateral (side) edge of the halo of mamillosa is shown
in the Barthlott & Hunt book, p. 91, fig. 36.5, as a broad pale
band.
VI. TAXONOMY SECTION
The classification of Echinopsis, s.l., is discussed in the
following sections.
A. GENUS ECHINOPSIS
The only reliable character distinguishing this genus from genus
Trichocereus is the type of fruit. It is dry or semi-dry in
Echinopsis vs. wet in Trichocereus. That said, I certainly do
not like to rely on a single character to distinguish genera, or
any taxon. In this case, though, the type of fruit seems to
represent a fundamental difference between subg. Echinopsis and
subg. Trichocereus and supporting data relating to their
pollination syndromes (strategies for attracting particular kinds
of pollinators) appear to support this interpretation, as further
suggested in the following discussion.
Subg. Echinopsis and subg. Trichocereus appear to be adaptive
twins as far as their pollination syndromes are concerned:
flowers open at night, stamens bilaterally symmetrical, petals
white and floral tubes and throats long. Lobivia and
Helianthocereus also are adaptive twins: flowers of most or all
open diurnally, stamens radially symmetrical, petals generally
are colored, and floral tubes and throats short. All these are
adaptive features that relate to their pollination syndromes. If
lobivias represent an ancestral stock of Echinopsis, s.l., and
subg. Echinopsis and subg. Trichocereus are derived from
lobivias, the similar pollination syndromes of the two subgenera
can be explained as a result of convergence. This would explain
why the two have been difficult to distinguish as separate taxa.
The similarity of the pollination syndromes of Helianthocereus to
Lobivia can be accounted for by Helianthocereus simply
representing a line derived from lobivias. But this is all
speculation. The more important question, I think, involves
subg. Echinopsis and subg. Trichocereus: Is there more
substantial evidence for evolutionary convergence of these two
subgenera? I'll address that question, as follows, by pointing
out the differences between them that suggest separate
evolutionary origins.
Though the flowers of both subg. Echinopsis and subg.
Trichocereus are bilaterally symmetrical, their throat stamen
clusters differ in a basic way. Those of all subg. Echinopsis
I've examined are compact within the throat, this the result of
an adhesive coating of the filaments, while the throat stamen
cluster of all the subg. Trichocereus I've seen, form a more
diffuse assemblage in the throat, due to a different arrangement
of the stamens within the throat, though a perhaps weaker
adhesiveness is still a factor in holding many of the stamens
together. On one hand, the two forms of the throat stamen
cluster appear to represent a fundamental difference between the
two subgenera, which can be interpreted as resulting from
convergence. On the other hand, the presence of adhesive
filaments seems like such a unique character that it might
indicate a closer relationship between the genera. But before I
discuss this possibility, I should point out that there seems to
be a general tendency in cacti to form bilaterally symmetrical
flowers, specifically by the development of throat stamen
clusters. This is shown in many diverse genera: epicacti and,
based upon photographs, also in Aporocactus, Brasilicereus,
Epiphyllum, Harrisia, Heliocereus and Selenicereus, as examples,
each presumably developing its own version of the bilateral
condition but none, as far as I can tell, with the aid of
adhesive stamen filaments. A possible explanation for this is
that the Echinopsis, s.l. condition uniquely leads to the
adhesive filament condition. Perhaps a way to investigate this
is to determine if the filaments of some lobivias whose throat
stamens in the ventral half of the throat appeared to have
lengthened--and could represent a first step toward the
definitive bilateral condition--are indeed adhesive. If I have
time, I will investigate this. But there are other differences
between subg. Echinopsis and subg. Trichocereus that suggest
they represent separate phyletic (evolutionary) lines.
One of the differences involves the scales of the floral tubes.
In subg. Echinopsis, the scales tend to be more elongate (8 to
29 times longer than broad, though only 5 times longer than broad
in the Mamillosa group), and in subg. Trichocereus only 3 to 10
times longer than broad. This suggests a different evolutionary
process for the lengthening of the floral tubes in the two
subgenera. Perhaps associated with this evolutionary process,
the nectar chambers tend to be much longer in subg. Echinopsis
than in subg. Trichocereus: 12 to 40 mm vs. 3 to 13 mm. Of
interest, the rim stamens are differently colored in the two
subgenera, always white in subg. Echinopsis and the inner side
of the rim stamens being green in subg. Trichocereus. This
difference is possibly a factor in attracting different
pollinators or just a different way of attracting the same
pollinators. There is another perhaps very importance difference
between the two subgenera involving the traces of rim stamens in
the throat, but I still have to study this.
1. SUBGENUS ECHINOPSIS.
Diagnosis: Stem with relatively few ribs, 8-12, except in
Mamillosa group, 22-26, central spines straight. Flowers
typically white, fragrant and "early nocturnal-type" (fully
opened in the evening before 9PM, standard time), symmetry
bilateral, throat stamen cluster forming compact bundle in the
throat owing to adhesiveness of filaments, throat stamens absent
in dorsal half of throat. Rim stamens white. Floral tube with
scales 8-29 times longer than broad, except in Mamillosa group,
only 5 times longer than broad. Throat long, length 103-188 mm;
nectar chamber long, length 12-40 mm. Fruit dry-type. Seeds
with body cells semi-matte to matte black, cuticular striations
of body cells "Micro-city".
Subgenus Echinopsis comprises several species groups. Briefly,
the major groups are, (a) the Eyriesii group, defined by features
of the stem and seeds, and comprising, e.g., eyriesii, oxygona,
?tubiflora and ?tapecuana (labeled obrepanda subsp. tapecuana in
Anderson, p.275), (b) the Sucrensis group, defined by floral tube
and fruit scale characters, type of fruit dehiscence and HMR of
seeds with "golden halo", (c) a group comprising two unidentified
minutely spined species, and (d) the Mamillosa group,
characterized by large globular stem with many ribs that are
divided into low rounded tubercles, and relatively short floral
tube scales and unusual seeds whose HMR is produced into a puffy,
strongly exserted "golden halo". Also, very unusual for subg.
Echinopsis, the petals of two members of the Mamillosa group are
colored, kermesina in red (from lit.) and a plant from the
Huntington collection, dark pink tones. The latter might have
resulted from hybridization.
There are interesting similarities between the Mamillosa and
Sucrensis groups, conceivably indicating a common phyletic line:
The floral tube of the Sucrensis group and at least some of the
Mamillosa group have deeply incised fleshy scales, distal scales
with prominently hooked bracteoles and an HMR that is produced
into a "golden halo", though the halo is much more prominent in
the Mamillosa group. Because of this possible relationship I've
taken a conservative approach to the Mamillosa group and am
treating it as a member of subg. Echinopsis. Besides, I've seen
only three members of the group and can't generalize on their
condition.
Also at the Huntington is E. cochabambensis whose narrow, curved,
columnar stem is peculiar for subg. Echinopsis, and E. huotii.
Huotii shows some similarity to the Sucrensis group in its hooked
bracteoles and presence of a "golden halo".
The Echinopsis subgenus is contrasted to other subgenera as follows:
Subg. Echinopsis vs. Pseudolobivia.
(a) Throat stamens: absent in much of the dorsal half of the
throat in subg. Echinopsis vs. typically present in the dorsal
half of Pseudolobivia, (b) central spines: straight in subg.
Echinopsis vs. gently curved to hook-like in Pseudolobivia, (c)
body cells of seeds: matte to semi-matte in subg. Echinopsis vs.
glossy in Pseudolobivia and (d) cuticular striations of body
cells: "Micro-city" in subg. Echinopsis vs. "Starry Skies" in
Pseudolobivia. Less reliable because of overlapping, (e) flowers
open nocturnally in subg. Echinopsis vs. at least sometimes
diurnally in Pseudolobivia, and (f) number of ribs: tending to be
fewer in subg. Echinopsis, 8-14 (but up to 26 in Mamillosa
group) vs. 12-28 in Pseudolobivia.
Subg. Echinopsis vs. Lobivia and Helianthocereus.
(a) Stamen symmetry: bilateral in subg. Echinopsis vs. radial
in Lobivia and Helianthocereus and (b) throat length, as measure
of floral tube length: longer in subg. Echinopsis, 103-188 mm
vs. 15-59 mm in Lobivia and Helianthocereus. Less certain due
data lacking in some species, (c) flowers opening nocturnally in
subg. Echinopsis vs. opening diurnally in Lobivia and
Helianthocereus. And less reliable because of overlapping, (d)
flower color: white in great majority of subg. Echinopsis vs.
usually colored in Lobivia and Helianthocereus, (e) relative
length of floral tube scales: 8-29 times longer than broad in
subg. Echinopsis (only 5 times longer than broad in Mamillosa
group) vs. 1-15 times longer than broad in Lobivia and
Helianthocereus, and (f) length of nectar chamber: longer in
subg. Echinopsis, 12-33 mm vs. 0-13 mm in Lobivia and
Helianthocereus.
Subg. Echinopsis vs. subg. Trichocereus.
(a) Fruit: dry-type in subg. Echinopsis vs. wet-type in subg.
Trichocereus, (b) throat stamen cluster: a compact bundle in
throat of subg. Echinopsis vs. a diffuse bundle in throat of
subg. Trichocereus, (c) rim stamen filaments: white in subg.
Echinopsis vs. inner surface green in subg. Trichocereus, and
(d) floral tube base: narrower in subg. Echinopsis, diameter
near base, 8-11 mm (13 mm in Mamillosa group) vs. 15-23 mm in
subg. Trichocereus. Less reliable because of overlapping, (e)
floral tube scales: tending to be more elongate in subg.
Echinopsis, 8-29 times longer than broad (but only 5 times longer
than broad in Mamillosa group) vs. 1-10 times longer than broad
in subg. Trichocereus, and (f) nectar chamber: tending to be
longer in subg. Echinopsis, 12-33 mm vs. 0-13 mm in subg.
Trichocereus.
2. SUBGENUS PSEUDOLOBIVIA
Diagnosis: Stem with relatively moderate number of ribs to many,
12-26, spines at least moderately long, central spines curved or
hooked. Flowers white, variable in opening times, from "late
nocturnal-type" (opening fully after 9 PM, standard time) to
diurnal, symmetry bilateral, throat stamen cluster forming
compact bundle in throat, throat stamens typically present in
dorsal half of throat (absent in "coronata"), rim stamens
filaments white or green on inner surface. Floral tube with
scales 8-15 times longer than broad. Throat long, length 64-168
mm; nectar chamber long, length 18-48 mm. Fruit dry-type. Seeds
with body cells glossy brown or black, low-domed in most species,
conical in subdenudata, cuticular striations "Starry Skies".
Pseudolobivia is defined here in a much narrower sense than its
author, Curt Backeberg, who interpreted Pseudolobivia as standing
"halfway between Echinopsis and Lobivia". In my treatment, I've
included only those species which have bilaterally symmetrical
white flowers, throat stamens in the dorsal half of the throat
and a long floral tube. Species represented are ancistrophora,
obrepanda and similar forms, and polyancistra.
I've seen only a few species of Pseudolobivia, but an Obrepanda
group appears to be recognizable. Their rib tubercles and
areoles are more or less offset to one side and the radial spines
tend to be characteristically long, stout, recumbent and
pectinate.
Three specimens from the same collection, labeled coronata, pose
a problem. The description and photo of coronata in Anderson
does not match this specimen, so I've indicated the three plants
as "coronata", indicating their questionable identity. More
disturbing from a taxonomic view, the plants lack throat stamens
in the dorsal half of the throat. This is unusual for
Pseudolobivia, but the plants do show the other pseudolobivia
characters in the curved central spine and seeds whose body cells
are mirror-like glossy with "Starry Night" cuticular striations.
Subg. Pseudolobivia is contrasted to other subgenera as follows:
Pseudolobivia vs. Lobivia and Helianthocereus.
(a) Stamen symmetry: bilateral in Pseudolobivia vs. radial in
Lobivia and Helianthocereus, (b) throat length, as measure of
floral tube length: longer in Pseudolobivia, 64-168 mm vs. 15-59
mm in Lobivia and Helianthocereus, and (c) nectar chamber: longer
in Pseudolobivia, 18-48 mm vs. 0-13 mm in Lobivia and
Helianthocereus. Less reliable due to overlap are (d) flower
color: white in Pseudolobivia vs. usually colored in Lobivia and
Helianthocereus, and (e) floral tube scales: 8-15 times longer
than broad in Pseudolobivia vs. 1-15 times longer than broad in
Lobivia and Helianthocereus.
Subg. Pseudolobivia vs. subg. Trichocereus.
(a) Fruit: dry-type in subg. Pseudolobivia vs. wet-type in
subg. Trichocereus, (b) throat stamens: forming compact bundle
in throat of Pseudolobivia vs. diffuse bundle in throat in subg.
Trichocereus, and (c) floral tube base: narrower in
Pseudolobivia, diameter near base 6-9 mm vs. 15-23 mm in subg.
Trichocereus. Less reliable because of overlapping, (d) relative
length of floral tube scales: tending to be more elongate in
Pseudolobivia, 8-15 times longer than broad vs. 1-10 times
longer than broad in subg. Trichocereus, and (e) length of
nectar chamber: tending to be longer in Pseudolobivia, 18-48 mm
vs. 0-13 mm in subg. Trichocereus.
Differentiation from Lobivia, Helianthocereus and subg.
Trichocereus is covered above under subg. Echinopsis.
The throat stamens in the dorsal half of the throat of
Pseudolobivia show a great deal of variation in number even from
flower to flower of a given plant, ranging from none or a few to
many. (I will refer to these dorsal stamens as *radial throat
stamens* for conciseness.) The mere presence and, importantly,
the variation of the throat stamens in the dorsal half of the
throat raises an interesting question. Is their presence merely
a retention from some ancestral line or is it the result of
hybridization? This really involves the origin of Pseudolobivia.
But why consider hybridization at all? The reason is that radial
throat stamens also occur in my subg. Echinopsis x Lobivia
hybrids where they show the same variation as in Pseudolobivia,
and a similar situation occurs also in Pseudolobivia x Lobivia
hybrids, discussed below. It appears, then, that the variation
is the result of hybridization, which implies that Pseudolobivia
is a hybrid genus, presumably derived from a subg. Echinopsis x
Lobivia cross or crosses. Is there any hint of what the lobivian
parent could be? Subg. Echinopsis and Pseudolobivia both have
seeds of similar form, a type almost round in cross-section, but
the luster of their body cells are different (matte or semi-matte
in subg. Echinopsis vs. glossy in Pseudolobivia). The only
present-day lobivias with a rounded type of seed, as far as I
know, are arachnacantha and mizquensis, so if seeds are an
indicator of parentage, an ancestor from the arachnacantha or
mizquensis lines might be the lobivia parent. Just for the fun
of speculating which of these two lines would more probably
represent the lobivian parent, Mizquensis would be the answer,
for the ribs of this species are divided into well-developed
acute tubercles, as in the Obrepanda group.
This isn't the end of the subject of possible hybridization
concerning Pseudolobivia, though. There are several collections
from Entre Rios in Bolivia (Lau 400 collection) of
ancistrophora-like plants at the Huntington. All these plants
differ from typical ancistrophora, as described in Backeberg's
Lexicon, in the fewer number of ribs, 12-13 rather than 15-16.
Each of the plants shows an extreme range of variation from
flower to flower in the number of the radial throat stamens,
varying from none to many. There is also an oddity about the
ovaries. They generally lack ovules or have only a few!
Obviously enough are present, though, for the Entre Rios
population to persevere. There is one other oddity about the Lau
400 plants. They appear form fruits without a pollinate partner.
A very interesting population!
3. NEW SUBGENUS FOR SUBDENUDATA CARDENAS
Diagnosis: Stem with 13 acute ribs, divided into low tubercles,
spines minute. Flowers white, "early nocturnal-type" (fully
opened in the evening before 9PM, standard time, as in subg.
Echinopsis), symmetry bilateral, throat stamen cluster forming
compact bundle in throat, throat stamens absent in dorsal half of
throat, or several present more usually, rim stamens filaments
white. Floral tube with scales 13 times longer than broad.
Throat long, length 131 mm; nectar chamber long, length 33 mm.
Fruit dry-type. Seeds with translucent, glass-like testa and
seeming internal source of magma light when seeds fresh, body
cells low-conical, "lit" by magma glow, cuticular striations
lacking at 150 power under dissecting microscope.
Subdenudata can be distinguished from all other Echinopsis, s.l.,
by its seeds, as described above, and also from Pseudolobivia by
its minute spines. Though the presence of throat stamens in the
dorsal half of the throat suggests a relationship with
Pseudolobivia, the minute stem spines is a character shared with
some species of subg. Echinopsis as is the early nocturnal-type
flowers, which are typical of subg. Echinopsis.
4. SUBGENUS LOBIVIA
Diagnosis: Stem with ribs undivided or divided into tubercles,
tubercles when present, either in straight or oblique rows.
Pollination syndrome very variable, involving petals, stamens, or
throat structures. Floral symmetry typically radial with
suggestion of incipient development of bilateral symmetry in some
cinnabarina. Rim stamens usually colored, at least distally, but
not green. Floral tube with scales usually short, 2-15 times
longer than broad. Throat short, length 15-58 mm; nectar chamber
often absent, when developed, tending to be short, length 0-13
mm. Fruit dry- or "semi-dry" types. Seeds extremely variable.
Lobivia is a large and remarkably diverse subgenus, so diverse
that it would appear to be polymorphic (represented by more than
one evolutionary line). This diversity points to Lobivia as the
oldest major group of Echinopsis, s.l. Many species groups are
represented and it would be a major project to characterize them
all, which I shall not attempt. Instead I will only discuss two
species groups which differ so significantly from the "usual"
type of lobivia that, in my estimation, both deserve the rank of
subgenus separate from lobivias. And they are so treated here, a
new subgenus for the Maximiliana group and the subg.
Chamaecereus for E. chamaecereus, as follows:
5. MAXIMILIANA NEW SUBGENUS
There are many different forms, listed as varieties of species
maximiliana, that were ascribed to this group by Walter Rausch in
his "Lobivia '85", the last complete revision of genus Lobivia.
Obviously, maximiliana is quite variable then. The variation I
have personally seen involves a great difference in the length of
the floral tube (the throat, as a measure of the length,
substantially varying from 8 to 32 mm long), the relative length
of the scales, the breadth of the nectar chamber and the
thickness of the superior (upper) ovary wall. But what is really
peculiar about this subgenus is a major modification of the
floral tube and of the rim stamens found in no other Echinopsis,
s.l. The distal end of the tube is uniquely produced to form a
very thin, broadly truncated-conical tube-like structure,
obliterating the distal throat division in the process, and the
rim stamens originate at the tip of the tube-like structure as an
adpressed conical cluster that completely encloses the stigma
when the flower first opens. The group of rim stamens opens
later the same day or the next day (I'm not certain which) to
form a tiny aperture over the stigma, thus permitting
pollination. Also very unusual, for at least some members of the
subgenus, the throat stamens in the base of the zone of throat
stamens tend to be rather widely spaced, adding extra room to the
nectar chamber for nectar production and/or storage.
6. SUBGENUS CHAMAECEREUS
Diagnosis: Throat essentially a reddish urn-like chamber with a
very short, broadening pale nectar chamber and without a distinct
distal throat division, the throat thus almost entirely
comprising only a zone of throat stamens; cells of tissue lining
zone of throat stamens elongate-low-tuberculate, glistening in
the cut-opened throat. Throat stamens red, short, hook-shaped
and sparsely distributed over the entire zone of throat stamens.
The throat and throat stamens of Chamaecereus, as described
above, distinguishes this subgenus from all other Echinopsis,
s.l. Because of the almost identical throat and throat stamens
of E. saltensis and E. chamaecereus, which seem very unlikely to
have resulted from convergence, I'm placing both in the same
subgenus to emphasize their relationship, despite the differences
in their stems.
E. saltensis differs from E. chamaecereus not only in its
short-cylindrical stem, as opposed to the uniquely
thin-cylindrical and soft trailing stem of chamaecereus, but also
in the orientation of the rim stamens which, in saltensis,
converge toward the center of the throat when the flower first
opens, leaving a somewhat narrow aperture over the stigma (until
it opens more fully later on, probably on the second day of
flowering, as in the Aurea group of Lobivia), while in E.
chamaecereus, the rim stamens open fully right from the start,
though the circle formed by the rim stamens is not obvious from a
top view of the flower.
7. PSEUDOLOBIVIA x LOBIVIA HYBRIDS.
In the Preston-Mafhams' book, published in 1991, there is the
following passage in Appendix 1 concerning Echinopsis obrepanda:
"Its wide distribution through the highlands of central and south
Bolivia has caused this species to show a perplexingly broad
range of variation. As a result, many individual populations
showing their own distinctive characteristics of body morphology,
spination and flower colour have been named as discrete species
or varieties. A number of these names were based mainly on
flower colour, but as populations are now known where a full
range of colours-- from white through pink and magenta to red and
scarlet--can be found on adjacent plants, this form of splitting
now seem untenable... In some of these populations only a single
flower colour is found, e.g. white or magenta, while in others
the full range described above occurs."
Is obrepanda just a very variable species, or is there another
explanation to explain the broad range of variation in its
flowers and stems? E. obrepanda is interpreted here in a very
restricted sense, as a group of perhaps varieties of the
Obrepanda group (subg. Pseudolobivia) whose flowers are white
and whose radial spines are more or less characteristically long,
recumbent and pectinate. But what about those plants mentioned
by the Preston-Mafhams with colored flowers? There are many of
these in the Huntington collection. They have been interpreted
as varieties of E. obrepanda or full species (as mentioned by the
Preston-Mafhams). I have studied them carefully, and though
there is still further work to be done, one thing seems evident.
These plants are hybrids derived from crosses between a member or
members of the Obrepanda group and one or more unknown lobivias.
Which lobivias they could be, I don't know, but two in the
Huntington collection, Echinopsis (subg. Lobivia) mizquensis and
E. (subg. Lobivia) pojoensis Rausch, or species like these, are
possible candidates. Both are red-flowered forms regarded as
varieties of Lobivia calorubra by Rausch in his "Lobivia '85"
book, but listed as subspecies of Echinopsis obrepanda by
Anderson.
To investigate the possibility that the variant plants can really
be interpreted as Obrepanda group x Lobivia hybrids, I made a
table of five characters which show quite a range of variation in
the putative hybrids. The five are, (a) throat length (as a
measure of floral tube length): short in the lobivias and long in
the Obrepanda group, (b) flower color: red in the lobivias, white
in the Obrepanda group, (c) throat color: red in the lobivias,
light greenish in the Obrepanda group, (d) rim stamen color: red
in the lobivias and white in the Obrepanda group, and (e) throat
stamen color: red in the lobivias, distally white and light green
basally in the Obrepanda group. I refer to these characters in
the table as lobivia-type, obrepanda-type and intermediate-type
(intermediate between the two, possibly resulting from
hybridization). I hope to add more characters this coming
flowering season.
I then listed the three groups of plants--suspected ancestral
lobivia types, putative hybrids and obrepanda--in a table
arranging them in order of their throat lengths, from shortest to
longest. The suspected lobivia parental types (pojoensis Rausch
and mizquensis) came first since their throats were the shortest,
while the representative of the Obrepanda group, an obrepanda
specimen from Cordoba in Argentina, was last since its throat was
the longest. The hybrids fell in between, from those with the
shortest throat (megalocephala Rausch, a taxon missed by
Anderson) to those with the longest throat (the "Calorubra
assemblage").
The throat lengths in the table revealed four groupings, the
first two overlapping some: (a) throat lengths 33 to 43 mm (the
suspected ancestral-type lobivian forms), (b) lengths 36 to 60 mm
(five collections of hybrids I refer to as megalocephala), (c)
lengths 77 to 129 mm (15 miscellaneous collections of hybrids,
the "Calorubra assemblage") and (d) lengths 150 to 166 mm
(obrepanda).
How did the five characters mentioned above sort out in these four groupings?
(a) In the suspected ancestral lobivian forms (33 to 43 mm
throats), all five characters were all lobivia-type, of
course,
(b) in the hybrids I refer to as megalocephala (36-60 mm
throats), the five characters in some
plants were all lobivia-type, but in others was a mixture of
lobivia- and intermediate-types,
(c) in the hybrids of the "Calorubra assemblage", (77-129 mm
throats), most plants showed a mixture of intermediate and
obrepanda-type characters and others showed a mixture of
lobivia, intermediate and obrepanda-type characters in
different combinations,
(d) in obrepanda (150-168 mm throats, a range in this plant),
all characters were obrepanda-type, of course.
From the above data, the putative lobivia and obrepanda parental
forms showed no variation in the five characters; megalocephala
showed relatively little, only lobivia- and intermediate-type
being present; while in the "Calorubra assemblage" there was
marked variation, every individual showing combinations of either
two or three of the lobivia-, intermediate- and obrepanda-type
characters. The occurrence of intermediate-type characters and
mixtures of the three different types in virtually all of the
presumed hybrids, I think is a good indication that, indeed, they
do have a hybrid origin.
I should note that I had not considered E. (subg. Lobivia)
cinnabarina as a possible parental type at first. I've changed
my mind since then. This species, which is a complex of
different forms, shows the same lobivia-type characters as
pojoensis Rausch and mizquensis. But it doesn’t really matter if
cinnabarina was excluded from the table. The results would have
remained the same. However, it seems more likely that one or
more cinnabarina forms would represent a lobivian parent or
parents since cinnabarina is more widely distributed than the
other two species.
There are distinct geographically isolated populations of the
obrepanda hybrids, centered from the mid Cochabamba department to
the northern Chuquisaca department in Bolivia. The differences
in the data of megalocephala vs. the "Calorubra assemblage" in
the table does indicate such a population diversity. The
localities are known for many of the hybrid specimens listed in
the table. There are five collections at the Huntington from
Pojo in which four hybrids have red flowers (a lobivia-type
character) and the fifth, an orange flower, which I interpret as
an intermediate-type. Regarding the four other characters, more
than half are lobivia-type and the rest, intermediate-type. The
question now arises, should the hybrid Pojo population be
considered as a distinct taxon and, if so, what should it be
called? I'm inclined to recognize this population as distinct,
indeed a species, mainly because the short floral tube sets it
apart from the other populations and there is a preponderance of
lobivia-type characters in the four other categories, as shown in
the table. Since Pojo happens to be the type locality of
megalocephala, calling this form E. megalocephala seems justified.
But what about the other populations in the table? There is
collecting data only for two others, one from Aiquile and the
second from Sucre. The specimens from both places showed a
mixture of lobivia-, intermediate and obrepanda-type characters
(one of the situations mentioned by the Preston-Mafhams). Should
these two populations each be regarded as distinct taxa?
Probably not. They apparently have no feature or features to
distinguish them from each other and probably not from other
populations, unless adaptation to a specific habitat be
considered. But they cannot be considered as varieties of a
single species, either, since the origins of the different
populations have occurred in different places and probably
involved different parental varieties or species, for this would
make the single broadly distributed species polyphyletic
(comprising forms of different evolutionary lines). (This is why
I favor calling megalocephala a full species.) Some of the above
putative hybrids in the Huntington collection have been
identified as calorubra or as varieties of calorubra by Friedrich
Ritter, so I'm just informally calling the entire group, the
"Calorubra assemblage" and let someone else worry about it.
Now that that problem's solved, I'll briefly mention that the
seeds of the hybrids I've examined are more or less similar
enough to consider all the hybrids as belonging to the same
group. Most of the seeds are beautiful, glossy gems (not as
great as subdenudata, though), but I've not been able to connect
the seeds to any extant lobivian species for possible parentage.
Species or variety names that are probably Obrepanda group (or
perhaps other Pseudolobivia) x Lobivia hybrids are:
calliantholilacina, calorubra, frankii, megalocephala, purpurea
and rojasii. There are undoubtedly others.
The Preston-Mafhams mentioned there are localities where only a
single flower color is found. The white-flowered populations
would probably be a form of the Obrepanda group, as defined here.
Pseudolobivia x Lobivia hybrids (those known to me at least) can
be distinguished from both subg. Echinopsis and Pseudolobivia by
their colored flowers and often from subg. Echinopsis by their
rib tubercles, most of which are hatchet-shaped at least in the
apical portions of the stems.
B. GENUS TRICHOCEREUS
As previously mentioned, the only significant character
distinguishing Trichocereus as a genus from Echinopsis is the
fruit: wet type in Trichocereus vs. the dry or semi-dry type in
Echinopsis. There may also be significant genetic barriers to
successful hybridization between the two genera, especially when
columnar Trichocereus are involved, based upon some crosses I
carried out, but I don't have enough data to substantiate this.
Two subgenera may be recognized in Trichocereus, Helianthocereus
and subg. Trichocereus. These can be distinguished by (a)
stamen symmetry: radial in Helianthocereus vs. bilateral in
subg. Trichocereus and, with possible exceptions, (b) flower
pigmentation: colored in Helianthocereus vs. white in
Trichocereus, and (c) opening of flowers: diurnal and remaining
open more than a single day in Helianthocereus vs. "early
nocturnal-type" (fully opening before 9PM, standard time) and
remaining open only one day (or less) in subg. Trichocereus.
Most or perhaps all of the above Helianthocereus characters,
except fruit type, are shown by Lobivia, the presumably ancestral
group and Helianthocereus may be assumed to be derived from that
subgenus. Helianthocereus and subg. Trichocereus are possibly
monophyletic (representing a single evolutionary line). If this
is the case, subg. Trichocereus may branched off from a
Helianthocereus ancestor as it underwent convergence with subg.
Echinopsis in its pollination syndrome (white, scented,
bilaterally symmetrical nocturnal flowers).
8. SUBGENUS HELIANTHOCEREUS
The species I've seen belonging to this subgenus are bruchii,
crassicaulis (and an allied species), grandis (a large globose
form sometimes incorrectly referred to as formosa, which is a
tall cylindrical species), huascha and rubriflorus. Except for
crassicaulis and its allied species, the stems and floral tubes
of the other species are more massive than those of Lobivia.
Also, the nectar chamber of Helianthocereus is less well
developed than in Lobivia, from being absent to very short, 0-2
mm in length vs. 0-13 mm in Lobivia.
Trichocereus pasacana, currently synonymized with atacamensis,
might belong to this subgenus, but I have not seen its flower
clearly enough to determine that. Pasacana conceivably could be
a Helianthocereus x subg. Trichocereus hybrid.
9. SUBGENUS TRICHOCEREUS
This is a very large group and I've seen only a few of its
species. Its pollination syndrome, similar to that of subg.
Echinopsis, subg. Trichocereus is distinguished from the other
by its diffuse cluster of throat stamens in the throat, rim
stamens which are green along their inner surface, a stouter
floral tube and less elongate scales. These differences were
detailed under subg. Echinopsis.
I've had experience with but a few species: candicans,
'imperialis', spachianus, thelegonus and werdermannianus
(=terscheckii). These represent different growth forms: shrubby,
trailing and treelike. Nevertheless, I can offer little insight
into this subgenus except for contrasting it to Echinopsis.
10. HELIANTHOCEREUS x SUBGENUS TRICHOCEREUS HYBRIDS.
Helianthocereus x subg. Trichocereus hybrids evidently occur in
nature. The hybrids can be recognized by the presence of a
throat stamen cluster and throat stamens in the dorsal half of
the throat (radial throat stamens). The clearest indication of
hybridization in the wild, to my knowledge, was shown by Andreas
Wessner in the Hybriden-Journal (v. 2001, no. 2 and 3), a
publication published by the German "Arbeitsgruppe
Echinopsis-Hybriden". Wessner, traveling in northwestern
Argentina, observed presumed hybrids wherever T. (subg.
Helianthocereus) huascha and T. (subg. Trichocereus) occurred
together (syntopy), but not where only one was present. The
flowers in these populations were variously colored and some
individuals in Wessner's photographs evidenced the hybrid type of
throat stamen development. Wessner suggests that
pseudocandicans, a rejected name according to David Hunt's "CITES
Cactaceae checklist", may also be a hybrid derived from these
inter-subgeneric crosses. There are also specimens at the
Huntington that appear to be Helianthocereus x Trichocereus
hybrids. One of them, a shrubby plant in the Desert Garden,
becomes rather spectacular when covered by its dark red flowers.
C. NEW GENUS FOR LEUCANTHA AND RELATIVES
I originally had considered the Leucantha genus as outside the
realm of Echinopsis, s.l., but am taking a more conservative
approach now. Fundamental differences between the two are, an
absence of a distinct circle of rim stamens, the development of
radial symmetrical, extraordinarily numerous throat stamens and a
type of seed I haven't seen in Echinopsis, s.l. There is little
to connect the genus with s.l., only its elongate floral tube
with some minute tufts of hairs in the areoles and, of uncertain
significance, a wet-type fruit, similar (this is only from
memory) to that of genus Trichocereus. If this new genus is
indeed a member of s.l., its closest ally, would seem to be subg.
Trichocereus. Characters indicating a relationship--a very
distant one to be sure--are white, nocturnal flowers and wet-type
of fruit. On the other hand, the narrow stalk of its floral tube
is like that of subg. Echinopsis. However, the radial symmetry
of the Leucantha flower sets it far apart from both genera.
I have seen two species which belong to this genus, the globose
to short cylindrical leucantha and the taller cylindrical
rhodotricha. Stephen Jankalski considers adolphofriedrichii as a
third member of the group.
VII. POLLINATION SYNDROMES AND EVOLUTION
The flowers of many or perhaps most Pseudolobivia and, more
certainly of most or all Lobivia, open during the daylight hours.
But, regarding Pseudolobivia, why should a plant with white
flowers, an adaptation for nocturnal pollinators, be diurnal
rather than nocturnal? Four suppositions are necessary explain
this. The first is that Pseudolobivia originated from one or
several subg. Echinopsis x Lobivia crosses, as discussed under
Subgenus Pseudolobivia. The second supposition is that Lobivia
flowers are adorned with UV patterns on their petals (bees can
see UV), the presence of UV on petals known for other
day-flowering plants. Third, as a result of such subg.
Echinopsis x Lobivia crosses, the lobivia UV patterns was passed
on to the hybrids (Pseudolobivia). But what advantage would
Pseudolobivia gain from its hybrid origin? That leads to the
fourth supposition, that there was a climate change which
resulted in the reduction in numbers or disappearance of night
flying insect pollinators of subg. Echinopsis, these insects
attracted to the flowers of this subgenus by their perfume and/or
white color. Plants of subg. Echinopsis would thus be put at a
disadvantage and hybridization with a lobivian parent would have
been selected for.
Another question is, what advantage would Helianthocereus have
gained in differentiating from its lobivian ancestors, since both
subgenera typically have colored flowers? The main advantage
actually might not relate to flower color, but to fruit type,
more specifically to method of seed distribution.
Helianthocereus has a wet-type of fruit, presumably derived from
the dry or semi-dry type of one of more ancestral lobivias.
Perhaps a change of climate took place here, too, resulting in a
selective advantage for Helianthocereus to evolve a different
type of fruit for attracting different insects or birds to
distribute their seeds.
Any advantage to the hybridization between Trichocereus huascha
and T. candicans, as mentioned under Helianthocereus x
Trichocereus hybrids is less clear since they both have a
wet-type of fruit. I can speculate on that, but without knowing
if the petals of Helianthocereus has UV markings, it would just
be guessing.
VIII. A LAST NOTE
Although my suppositions on the phyletic relationships within
Echinopsis, s.l., may be in question,
or indeed wrong, I am sure the stamens provide very important,
even crucial, information in the study of the taxonomy of
Echinopsis, s.l., and recommend strongly that anyone working on
the group include an analysis of these structures in their work.
IX. ACKNOWLEDGEMENTS
I would like to thank Frank Goetz for supplying me with a copy of
Walter Rausch' Lobivia '85; Dr. David Hunt for his
recommendations of literature to read and for pointing out the
possibility that hybridization had played a part in the evolution
of Pseudolobivia, which led me to consider that possibility;
Stephen Jankalski for supplying me with a much appreciated
literature and a document he compiled showing the subgeneric,
sectional and species composition of Echinopsis, s.l., and
related genera; Brian Kemble for his photographs of subg.
Helianthocereus and subg. Helianthocereus; Myron Kimnach for his
frequent advice on generic nomenclature and terminology of
flowers, including comments on the hymen; Dr. Martin Lowry for
notes on Echinopsis ancistrophora and for providing me important
information about collections; Gary Lyons, Curator of the
Huntington Desert Gardens, for providing me with cuttings of
Echinopsis and Trichocereus; Roy Mottram for his very
considerable help in general and particularly his input on
pollination syndromes, generic interpretations and identification
of species; Bob Ressler for sending me trichocereus seeds, which
showed that subg. Trichocereus, like other Echinopsis, s.l., is
also a complex group; and John Trager, curator of the Huntington
Desert Collections, for his generous loan of specimens for me to
study (however, I must say, John, that I found a weed in one of
the pots!). And last but not least, to my wife, Aiko, who
assiduously read and corrected typos in the presentation.