ArticlePDF Available

A synopsis of the genus Cleistocactus Lemaire (Cactaceae)

  • International Organisation for Succulent Plant Study

Abstract and Figures

The history of the genus Cleistocactus Lem. (Cactaceae) is reviewed and significant eras outlined. A conservative species concept has been employed to group the over 300 published names into 24 species and 3 heterotypic subspecies. A complete synonymy is given for each of the accepted species along with an expanded description, distribution maps, ecological region associations and a brief commentary. The descriptions are expanded from the respective protologues to include additional data observed both in the field and in ex-situ culture.
Content may be subject to copyright.
148 Bradleya 34/2016
Summary: The history of the genus Cleistocactus
Lem. (Cactaceae) is reviewed and significant eras
outlined. A conservative species concept has been
employed to group the over 300 published names
into 24 species and 3 heterotypic subspecies. A
complete synonymy is given for each of the
accepted species along with an expanded
description, distribution maps, ecological region
associations and a brief commentary. The
descriptions are expanded from the respective
protologues to include additional data observed
both in the field and in ex-situ culture.
Zusammenfassung: Die Geschichte der Gattung
Cleistocactus Lem. (Cactaceae) wird betrachtet
und bedeutende Perioden skizziert. Ein
konservatives Artkonzept wurde angewendet, um
die über 300 publizierten Namen in 24 Arten und
drei heterotypische Unterarten zu gruppieren.
Eine komplette Synonymie wird für jede der
anerkannten Arten gegeben, zusammen mit einer
erweiterten Beschreibung, Verbreitungskarten,
Zuordnung zu ökologischen Regionen und einem
kurzen Kommentar. Die Beschreibungen werden
ausgehend von den entsprechenden Protologen
um zusätzliche Beobachtungsdaten ergänzt, die
sowohl in der Natur als auch in ex-situ-Kulturen
erhoben wurden.
Introduction and History
Slow Beginnings
Little did Lemaire suspect in 1844 when he
described a “pretty Cactus recently blossomed in
the stove-house of Courant at Le Havre” that it
was one of the most widespread species of what
would eventually become a very popular genus in
cactus horticulture. He named it Cereus
baumannii after Mr Napoleon Baumann of
Bolwilliers from whom he had received a plant.
Napoleon and his brother Augustus owned a
nursery and they had recently imported the plant
from the Cordilleras. It seems their marketing
strategy was excellent as the plant rapidly spread
to major collections within France so that
Monville was able to observe the flowers soon
thereafter. Fortunately, he was able to
communicate this to Lemaire prior to the
publication of volume 5 of L’Horticulteur
Universel allowing Lemaire to make an addition
to his preliminary description. This presumably
accounts for the curious separation of the first
description over two widely-separated pages of the
volume. Monville described the flowers as being
bright orange and very different in shape from
other cerei;more like those of Epiphyllum
truncatum (now Schlumbergera truncata, the
Christmas cactus). It appears Lemaire did not see
the flowers himself and continued to call the plant
Cereus baumannii in subsequent publications
over several years. Only when eventually
considering the production of a monograph did he
accept that the genus Cereus needed dividing.
Thus, in 1860, he created the genus Aporocactus
and transferred his Cereus baumannii there along
with C. flagelliformis. This unfortunate mistake
was rapidly pointed out to him by a “distinguished
botanist” and by an amateur cactophile, neither
of whom he names. The amateur botanist was
very likely Schlumberger of Rouen who had
amassed a large collection often used by Lemaire.
Lemaire subsequently made a trip to the botanical
garden at Caen to see the flowers of Cereus
colubrinus Hort. which had been described as
identical to those of his C. baumannii.
Recognising his error he published another new
generic name (the 27th in the family and for which
he feared he might be called a splitter!) for his
plant the following year. The new name was
Cleistocactus, named for the feature that the
stamens almost close the flower.
Although we now unequivocally accept the
genus it was not so for the first 40 years or so,
even though other species with partially closed
flowers were discovered and described. In 1897,
when Schumann published the first edition of his
Gesamtbeschreibung der Kakteen, he listed 5
names that we now include in Cleistocactus
(baumannii, hyalacanthus, laniceps,parviflorus
and smaragdiflorus) but continued to call them
Cereus. It was not until 1904 that Roland-
Gosselin, in an obscure French journal,
transferred these species to Cleistocactus, based
on a manuscript left by the late Frédéric Weber.
Perhaps because of the obscure nature of the
publication and its limited distribution the name
still did not come into common usage and it was
Bradleya 34/2016
pages 148–186
A synopsis of the genus Cleistocactus Lemaire (Cactaceae)
Martin Lowry
Northgate Avenue, Macclesfield, UK (email:
Bradleya 34/2016
only with the publication of Volume 2 of The
Cactaceae by Britton & Rose in 1920 that the new
name became more widely known, although they
still only included the same names as had
The Numbers Game
It was not until the exploratory activities of
Backeberg, Cárdenas and Ritter in the mid 20th
century that the genus began to grow. Cárdenas
was primarily interested in expanding the cactus
flora of Bolivia thus he contributed 27 new names,
all of which derive from single populations, but
failed to provide a systematic treatment of the
genus. On the contrary, Backeberg’s forte was
systematic treatment and, although he only
described 15 new taxa he did recognise that 4
older names also belonged to the genus and made
the appropriate combinations. In his final
treatment of the genus, published shortly before
his death in 1966, he accepted 47 taxa in
Cleistocactus and a further 7 in 3 genera now
considered to be synonymous. Ritter was the last
of the three to publish a treatment, in 1980, and
was extremely critical of Backeberg’s work
considering his descriptions riddled with mistakes
and inaccuracies and his genera over-inflated.
Even so he too proposed a large genus of 44 taxa
along with 2 monotypic genera for taxa now
included in Cleistocactus. Among those were 26 of
his own descriptions, only 10 of which were known
by name to Backeberg. In the background,
Buxbaum was giving serious consideration to the
use of morphology in cactus taxonomy. Having
discovered what he perceived as significant
differences in the shape of the flowers of C.
baumannii compared to those of the remainder of
the species he proposed to divide the genus into
two subgenera and created the subgenus
Annemarnieria for those species differing from the
type (Buxbaum, 1956).
The Modern Era
Since 2000, three major treatments of the
Cactaceae have appeared (Anderson, 2001; Hunt
et al., 2006; Lodé, 2015) each with a different
treatment of Cleistocactus. The first (Anderson,
2001), and its subsequent edition in German
(Eggli, 2005), was quite uncritical and accepted
many of the species erected by Cárdenas,
Backeberg and Ritter with little or no re-
evaluation of their status. Although managing to
reduce the taxon count to 38, there were still
many names based on single populations. The
basis for this work appears to have been the
CITES Cactaceae Checklist (Hunt, 1999).
Unfortunately 15 of the names listed in that work
as provisionally accepted were treated as fully
accepted without evaluating their potential as
synonyms. This, however, would have been a
difficult task at the time since little or no
information was available on the distribution or
variability of the species in habitat. That
information became more readily available in
subsequent years and was able to influence the
treatment in The New Cactus Lexicon (Hunt et al.,
2006). In that work only 24 taxa (17 species + 7
heterotypic subspecies) were fully recognised with
a further 5 species considered as provisional, all
placed in Cleistocactus.
The most recent treatment of the family (Lodé,
2015) claimed to be “The new classification of
Cacti mainly based on molecular data ...”.
However, the treatment of Cleistocactus lies
somewhere between those of Anderson (2001) and
Hunt et al. (2006) in accepting 27 species. It
differs markedly from both, however, in accepting
two of the species under Borzicactus Riccob. For
one of these (C. samaipatanus (Cárd.) Hunt) the
reason appears to be based on a misreading of
taxon names in the cladogram of Hernández-
Hernández et al. (2011) since the authors did not
include the species in their study.
Limited Specific Characters
The clear message that can be extracted from
this history is that, although it is relatively easy
to identify plants as belonging to the genus, it can
be very difficult to define the specific limits, as
perhaps is the case for all major genera of the
Cactaceae. In the treatment that follows I cannot
claim to be very different since I have encountered
this situation on several occasions when observing
the plants in habitat. Unfortunately, there are few
characters that can be used to separate the
species since all have a similar growth form,
essentially the same flower shape (with some
obvious exceptions) and more-or-less identical
fruits and seed. Those characters that I have
found useful include branching mode (basitonic
only vs. basi- and mesotonic), flower colour and
flower inclination (erect, angled or horizontal).
These characters become especially useful when
encountering sympatric species. This is a
frustratingly frequent occurrence, particularly in
the core of the range where it is common to find
two species, and sometimes three, at a single
Systematic Position and Circumscription
Buxbaum (1958) placed Cleistocactus as a core
member of his subtribe Borzicactinae within the
tribe Trichocereae when creating a formal division
of Schumann’s ‘Cereoideae’. He placed it alongside
several other genera including Arequipa,
Bolivicereus, Borzicactus, Denmoza, Loxantho-
150 Bradleya 34/2016
cereus, Matucana, Morawetzia, Oreocereus, Oroya
and Seticereus based on their possession of
common floral traits. It is testament to Buxbaum’s
careful and detailed analysis that this grouping
has remained together throughout all subsequent
modern classifications with no dispute. Although
some of the genera have been absorbed, for
example Arequipa into Oreocereus, Cleistocactus
has remained intact albeit at times taken to
include species from the other genera. Perhaps
the earliest examples are the inclusion of
Echinocactus rhodacanthus by Lemaire in 1861
and subsequently the inclusion of Oreocereus
celsianus by Weber in 1904. More recently,
following the deliberations of the IOS Working
Party (1986), the scope of the genus was expanded
to include all those cereoid species from Andean
South America having ornithophilous flowers and
solid fruits. Thus many species of Borzicactus
(including Akersia, Borzicactella, Clistantho-
cereus, Loxanthocereus, Maritimocereus and
Seticereus) subsequently gained combinations in
Cleistocactus through the publications of several
authors (Hunt & Taylor, 1987; Hunt, 2003;
Ostolaza, 1996, 1998; Ostolaza et al. 2003;
Leuenberger, 2002). This was the definition of
Cleistocactus used, with some reservation, in The
New Cactus Lexicon (Hunt et al., 2006).
At that time few studies using molecular
techniques to investigate phylogeny had been
published and those available had not
concentrated on Cleistocactus and related genera.
It was not until 2012 that a study including a
significant number of species from the
Borzicactinae was published (Schlumpberger &
Renner, 2012). The results clearly demonstrated
that Cleistocactus as circumscribed by Hunt et al.
(2006) was polyphyletic. The genus appeared in a
moderately well supported clade sister to a clade
containing only some members of Echinopsis s.l.
as well as Harrisia Britton, and distant from that
containing Borzicactus sensu Kimnach. The clade
containing Cleistocactus s.s. was poorly resolved
and contained several other genera, specifically,
Cephalocleistocactus Ritter, Vatricania Backeb.,
Samaipaticereus Cárd., Weberbauerocereus
Backeb. and Yungasocereus Ritter. Thus the
situation is very complex and currently only
partly resolved. A first step was taken by Hunt
(2013) in an update to The New Cactus Lexicon
where Borzicatus was reinstated. A broader
alternative view was taken by Anceschi & Magli
(2013) who, since the study also demonstrated
Echinopsis s.l. to be polyphyletic, transferred
several species of Cleistocactus, including the
type, into Echinopsis thus making the former a
synonym. There are several other, less
consequential, options that could be considered
but, given the limited taxon sampling from within
Cleistocactus, the low number of markers
employed and the lack of nuclear sequences I feel
it prudent to wait for further molecular data
before making further adjustments to the
circumscription of the genus.
Study Methods
Names. An initial list of names was obtained
by conducting a search of The International Plant
Names Index (IPNI; 2012) for records from the
following genera: Cleistocactus, Bolivicereus,
Hildewintera, Winteria, Winterocereus,
Borzicactus and Loxanthocereus, and
subsequently for the species epithets: strausii and
baumannii. Over 300 records were identified as
belonging to members of the genus Cleistocactus.
These records were downloaded in extended
format and loaded into a searchable, indexed
spreadsheet (Libre Office Calc) for improved
access. This file is available from the author on
Occurrence data. The primary dataset of
occurrence records was obtained several years
previously through a search of the Global
Biodiversity Information Facility (;
last accessed 8th April 2009) for all records of the
family Cactaceae. This resulted in 21,538 records
which were incorporated into an SQL-based
database using custom written software. The
database was augmented with further records
donated by many friends and travelling
companions and is continually updated with
occurrence information from my own field trips.
The number of records now stands at 36,768 of
which 420 relate to Cleistocactus.
Mapping. The 420 records of Cleistocactus
were exported from the database and converted
into ESRI shapefiles using the free software
DIVA-GIS ( The package was
used to generate point distributions for each
species using base maps built from country level
spatial data available at the same website. Shape-
files depicting the ecoregions of the world (Olsen
et al, 2001) were obtained from the World Wildlife
Fund (
restrial-ecoregions-of-the-world; last accessed 9th
August 2016). These were overlaid on the base
map and allowed identification of the major
biome(s) and ecoregion(s) within which each
species can be found.
The following list describes all those taxa that
I consider sufficiently distinct to require formal
recognition. The descriptions are based on those
given in The New Cactus Lexicon (Hunt et al.,
2006) with additional information included from
Bradleya 34/2016
the protologues of the basionym and subsumed
synonyms where relevant. They are also expanded
with information from my own observations, both
in the field and in culture, when details, for
example, of flowers and fruits, were not given in
the first description. All references to the
International Code of Nomenclature refer to the
Melbourne Code (McNeill et al., 2012).
Cleistocactus Lem., Ill. Hort. 8: Misc. 35. (1861).
Type: Cereus baumannii Lem.
Bolivicereus rd., Cact. Succ. J. (US) 23: 91 (May
1951). Type: Bolivicereus samaipatanus Cárd.
Borzicactus subg. Bolivicereus (Cárd.) Buxbaum,
Kakteen (H. Krainz) 58: CVc (1974). Basionym:
Bolivicereus Cárd.
Cephalocleistocactus Ritter, Succulenta (NL)
1959: 108 (1959). Type: Cephalocleistocactus
chrysocephalus Ritter.
Hildewintera Ritter, Kakt. and. Sukk. 17: 11
(1966), nom. inval. [Art. 33.4]. Type: Winteria
aureispina Ritter.
Hildewintera Ritter ex G.D. Rowley, Regnum.
Veg. 54: 15 (1968), nom. illeg. [Art. 52.1]. Type:
Winteria aureispina Ritter.
Loxanthocereus subg. Hildewintera Buxbaum,
Kakteen (H. Krainz) 58: CVb (1974).
Seticleistocactus Backeb., Descr. Cact. Nov. 3: 13
(1963). Type: Cleistocactus piraymirensis
Winteria Ritter, Kakt. and. Sukk. 13: 4–8, figs.
(1962), nom. illeg. [Art. 53.3]. Type: Winteria
aureispina Ritter.
Winterocereus Backeb., Kakteenlexikon p.455
(1966). Type: Winteria aureispina Ritter.
Shrubs, generally with several slender,
cylindrical stems which can be erect, ascending,
decumbent, procumbent or pendent. Stems with
5–30 ribs and closely spaced areoles bearing few
to many acicular spines. Flowers produced along
the upper section of each stem below the apex
normally from undifferentiated areoles but
occasionally along with extra bristles and spines.
The flowers are tubular, regular to zygomorphic,
diurnal, and vary in colour from red to green. The
hypanthium, occasionally markedly kinked above
the ovary, is covered with many small scales.
Tepals scarcely larger than the scales and in most
species barely spreading, though if spreading then
often obliquely. Stamens inserted in two series,
included or exerted; anthers and pollen often
purplish. Nectar chamber partially closed by
lowermost filaments or by a diaphragm. Style
mostly exerted, generally white or yellow. The
fruits are small and globose retaining the floral
remains. They are solid with fleshy pulp enclosing
small pyriform shiny black seeds.
Etymology: From Greek kleistos, closed, and
Cactus, referring to the way the throat of the
flower is almost closed.
Habitat & Distribution: Arid and humid zones of
the chaco, inter-Andean valleys, Yungas and pre-
puna generally amongst low shrubs and bushes
but occasionally in deciduous forest. Generally
terrestrial but frequently lithophytic or even
epiphytic, 100–3700m.
Ecological regions: The species are found in 10
different ecoregions with major concentrations in
the Bolivian Montane Dry Forest and adjacent
Southern Andean Yungas. Other frequently
occupied regions include the Central Andean
Puna and Bolivian Yungas. One species occurs
extensively throughout the Dry & Humid Chaco
whilst two others are found in the Peruvian
Occurrence: AR (CC, CD, CT, FO, JY, LR, SA, SE,
TM), BO (CB, CH, LP, PO, SC, TR), BR (MS), PE
(AP, AY, CZ, HC, JU), PY (AG, BQ, CE, CN, CR,
GU, PG, PH), UY.
1. Cleistocactus baumannii (Lem.) Lem., Ill.
Hort. 8: Misc. 35 (1861). Basionym: Cereus
baumannii Lem.
Aporocactus baumannii (Lem.) Lem. Ill. Hort. 7:
Misc. 66–68 (1860). Basionym: Cereus
baumannii Lem.
Cereus anguinus Gürke, Monats. f. Kaktk. 17: 166
(1907). Type: Paraguay, (not preserved?).
C. baumannii Lem., Hort. Univ. 5: 126, 315
(1844). Type: cult. hort. Paris from hort.
Baumann, not preserved. Neotype selected by
Hunt & Taylor, Cact. Syst. Init. 21: 6 (2006):
Jardin Fleuriste 1: t.48 (1851).
C. baumannii var. colubrinus K. Schum. (non C.
colubrinus Otto), Gesamtb. Kakt. p.133 (1898).
C. baumannii var. flavispinus K. Schum.,
Gesamtb. Kakt. p.133 (1898).
C. colubrinus Otto, in Förster’s Handb. Cactkde.
p.409 (1846).
C. colubrinus var. flavispinus Salm-Dyck, Cact.
Hort. Dyck. 1844: 32 (1845).
C. grossei Weingart, nom. nud., Monats. f. Kaktk.
18: 8 (1908).
C. tweediei Hooker, Bot. Mag. 76: t.4498
(1850).Cleistocactus anguinus (Gürke) Britt. &
Rose, The Cactaceae 2p.175 (1920). Basionym:
Cereus anguinus Gürke.
C. aureispinus Frič, nom. nud., ller’s Deutsche
Gärtn.-Zeitung xliii: 267 (1928).
C. baumannii ssp. anguinus (Gürke) Braun &
Esteves, Succulenta (NL) 74: 84 (1995).
Basionym: Cereus anguinus Gürke.
C. baumannii ssp. chacoanus (Ritter) Braun &
Esteves, Succulenta (NL) 74: 84 (1995).
152 Bradleya 34/2016
Basionym: Cleistocactus chacoanus Ritter.
C. baumannii ssp. croceiflorus (Ritter) Braun &
Esteves, Succulenta (NL) 74: 84 (1995).
Basionym: Cleistocactus croceiflorus Ritter.
C. baumannii ssp. paraguariensis (Ritter) Braun
& Esteves, Succulenta (NL) 74: 84 (1995).
Basionym: Cleistocactus paraguariensis
C. baumannii ssp. santacruzensis (Backeb.)
Mottram, Cact. Succ. J. (US) 61: 156 (1989).
Basionym: Cleistocactus santacruzensis
C. baumannii var. colubrinus (Otto) Riccobono,
Boll. Reale Orto Bot. Palermo 8: 266 (1909).
Basionym: Cereus colubrinus Otto.
C. baumannii var. flavispinus (Salm-Dyck)
Riccobono, Boll. Reale Orto Bot. Palermo 8:
266 (1909). Basionym: C. colubrinus var.
flavispinus Salm-Dyck.
C. baumannii var. grossei (Weingart) Marshall,
nom. nud., Cactaceae (Pasadena), p.106
C. bruneispinus Backeb., Die Cactaceae II: 1001
(1959). Replaced synonym: Cereus baumannii
var. colubrinus K. Schum.
C. chacoanus Ritter, Kakt. Südamer. 2: 671, Abb.
685 (1980). Type: Villa Montes, Prov. Gran
Chaco, Bolivia, Ritter 841 (U).
C. chacoanus var. santacruzensis (Backeb.) Ritter,
Kakt. Südam. 2: 672, Abb. 687 (1980).
Basionym: Cleistocactus santacruzensis
Backeb. Note: Ritter added further details
regarding the location, viz: “tiled roofs of house
in the capital”, Santa Cruz, Bolivia.
C. croceiflorus Ritter, Kakt. Südamer. 1: 272
(1979). Type: in the neighbourhood of Cerro
Gavan, Puerto Casado, Dept. Boquerón,
Paraguay, Ritter 1185 (U).
C. colubrinus var. flavispinus (Salm-Dyck) Borg,
Cacti p.150 (1937). Basionym: Cereus
colubrinus var. flavispinus Salm-Dyck.
C. colubrinus var. grossei Borg, nom. nud., Cacti
p.150 (1937).
C. flavispinus Backeb., Die Cactaceae II: 1000
(1959). Basionym: Cereus baumannii var.
flavispinus K. Schum.
C. grossei Backeb., nom. nud., Descr. Cact. Nov. I
p.17 (1957).
C. jugatiflorus Backeb., Die Cactaceae VI: 3685
(1962). Type: the cited illustration.
C. paraguariensis Ritter, Kakt. Südamer. 1: 273
(1979). Type: rocky hills near Paraguari, Dept
Paraguari, Paraguay, Ritter 1185b (U).
C. santacruzensis Backeb., Kakteenlexikon 89, fig.,
(1966). Type: Santa Cruz, Bolivia, Ritter 356
Echinopsis baumannii (Lem.) Anceschi & Magli,
Cactusinhabitat South America 2011/2013,
p.37 (2013). Basionym: Cereus baumannii
E. santacruzensis (Backeb.) Anceschi & Magli,
Cactusinhabitat South America 2011/2013,
p.40 (2013). Basionym: Cleistocactus
santacruzensis Backeb.
Stems erect or arching to 2m × 1–4cm, branch-
ing from the base with 12–17 ribs. Areoles ±1cm
apart bearing 15–20 acicular radial spines 1–2cm
long and 2–3 thicker central spines to 4cm. Flow-
ers red-orange, markedly “S”-shaped, 5–7cm × 7–
8mm, broadening to 1cm below the corolla.
Hypanthium with red scales each producing sev-
eral white hairs. Inner tepals red/orange, ex-
panding slightly to reveal the filaments and style.
The anthers produce red-magenta pollen. Fruit 1–
1.5cm diameter, globose.
Key to subspecies of C. baumannii
1. Stems >2cm diameter, spines > 1cm long
(AR, BO, S.PY) a. ssp. baumannii
Stems <2cm diameter, spines < 1cm long (BR,
NPY) b. ssp. horstii
Figure 1. Cleistocactus baumannii LM0507.01
north of Eyti, Bolivia, 980m. The form previously
known as ssp. santacruzensis.
Photograph: Martin Lowry
Bradleya 34/2016
a. ssp. baumannii
Stems erect, ascending or decumbent, > 2cm
diameter. Spines 10–15, the centrals to 3cm long
(Figure 1).
Etymology: named after Mr Napoleon Baumann
of Bolwilliers (sic) who gave Lemaire the plant he
Habitat & Distribution: Shrubland along the
Chaco margin amongst trees in loamy-sandy soils.
Its range is extensive reaching from La Rioja
Province, Argentina in the south to Santa Cruz de
la Sierra, Bolivia in the north and extending
eastwards from Villamontes in Bolivia, across
Paraguay to Asunción. 150–1050m.
Ecological regions: Dry Chaco, Chiquitano Dry
Forest and Humid Chaco. This is the only species
adapted to the harsh arid conditions of these
Occurrence: AR (CC, CD, FO, LR, SA, SE), BO
(CH, SC, TR), PY (AG, BQ, CE, CN, CR, GU, PG,
PH), UY. Map 1.
The plate chosen as the neotype for Cereus
baumannii Lem. is a copy of the illustration from
the protologue of Cereus tweediei Hooker. Kew
received plants from the Messrs Lee of the
Hammersmith Nursery and also grew plants from
seeds from a Mr Tweedie of Buenos Aires,
Argentina. It is not completely clear from
Hooker’s text which of these was used for the
drawing, however, I consider it more likely to have
been the plants from the Hammersmith Nursery,
which supplied Kew’s “largest and flowering
plant”, since the English description speaks of
“Our tallest plants ...”.
The decision to include Backeberg’s C.
santacruzensis and synonyms here is taken after
observing several populations in habitat. The only
character which appears to separate the two
forms is stem diameter, with those in the north
(Santa Cruz to Camiri), generally thinner than
those in the south (La Rioja and Córdoba).
Between these extremes (Villamontes to Santiago
del Estero) there are populations of intermediate
b. ssp. horstii (P.J. Braun) N.P. Taylor, Cact.
Cons. Init. 6: 15 (1998). Basionym: Cleistocactus
horstii P.J. Braun.
C. horstii P.J. Braun, Kakt. and. Sukk. 33: 204–
208, figs. (1982). Type: lowlands of Río
Amoguija, SW of Mato Grosso do Sul, Brazil,
150m, 1974, Horst & Uebelmann 373
Echinopsis horstii (P.J. Braun) Anceschi & Magli,
Cactusinhabitat South America 2011/2013
p.38 (2013). Basionym: Cleistocactus horstii
P.J. Braun
Stems decumbent to prostrate, to 1cm
Figure 2. Cleistocactus baumannii ssp. horstii at Fazenda Bala near Porto Murtinho, Brazil.
Photograph: Paulo Robson de Souza
154 Bradleya 34/2016
diameter, spines <1cm long (Figure 2).
Etymology: named to honour Leopoldo Horst,
Brazilian cactus explorer and friend of Pierre
Habitat & Distribution: lowlands west of the
Pantanal, 100m.
Ecological regions: Humid Chaco.
Occurrence: BR (MS), PY (CN). Map 1.
Although the protologue of Cleistocactus
horstii suggests a sparsely-spined plant, there are
several clones of HU 373 in cultivation which
demonstrate spination similar to that of the type
of the subspecies.
The remarkable similarity of Backeberg’s
illustration Abb. 921 in Die Cactaceae of what he
believed to be Gürke’s Cereus anguinus with the
photographs in the protologue of C. horstii P.J.
Braun leads me to suspect that these might
represent the same taxon. I refrain from making
them synonymous here for two reasons. First,
Gürke provided no explicit type location with his
description saying only that Schumann had listed
it as Anisits 28 on page 41 of the Nachträge of his
Gesamtbeschreibung. From there we learn that
Anisits collected the plant “bei Porongo”, an
Estancia in Dept. Presidente Hayes, Paraguay,
some 240km south west of Porto Murtinho, the
type locality of C. horstii, and much nearer to
other known locations of C. baumannii ssp.
baumannii. Secondly, at subspecies level, the
epithet anguinus has 3 years priority over horstii.
Making them synonymous would require a change
of name and introduce potential nomenclatural
2. Cleistocactus brookeae Cárd, Cact. Succ. J.
(US) 24: 144, figs., (1952) (as brookei’). Type:
“Pozo no. 4”, Camiri, Prov. Cordillera, Dept. Santa
Cruz, Bolivia, 900m, Mar 1949, A. Corro s.n. in
Cárdenas 4818 (LIL 531590, holo.); between
Zudañez and Incahuasi, Prov. Hernando Siles,
Dept. Chuquisaca, 1200m, Aug 1949, Miss W.
Brooke s.n. (US, para.).
C. brookeae var. flavispinus Ritter, Kakt.
damer. 2: 680, Abb. 664 (1980). Type: “some
km northwest from Boyuibe”, Prov. Cordillera,
Dept. Santa Cruz, Bolivia, 1963, Ritter 1129
Figure 3. Cleistocactus brookeae LM0512.02 east
of Monteagudo, Bolivia, 1300m.
Photograph: Martin Lowry.
Figure 4. Cleistocactus brookeae LM0636.01 at
Puente Azero, Bolivia, 1060m. The form previ-
ously known as ssp. vulpis-cauda.
Photograph: Martin Lowry.
Bradleya 34/2016
C. brookeae ssp. vulpis-cauda (Ritter & Cullmann)
Mottram, Cact. Syst. Init. 13: 10 (2002).
Basionym: C. vulpis-cauda Ritter &
C. vulpis-cauda Ritter & Cullmann, Kakt. and.
Sukk. 13: 38–40, figs. (1962). Type: Puente
Acero, Prov. Tomina, Dept. Chuquisaca,
Bolivia, Ritter 847 (ZSS, not found). Lectotype
(designated here): the specimen 124821 at
SGO labelled “Pte. Acero Nr. 12”.
C. wendlandiorum Backeb., Kakt. and. Sukk. 6:
113–117, figs. (1955). Type:fig.114, in Haage,
Freude mit Kakteen (1954), cited on p.117 of
the protologue.
A shrub with stems erect or pendant,
branching from the base to 2m × 2–4cm. Ribs 18–
24 bearing areoles 3–4mm apart. Spines 25–40,
undifferentiated, about 1cm long, yellow or white.
Flowers produced near stem apex, always vertical
regardless of stem orientation, 5cm × 8mm, red or
orange. Hypanthium geniculate above ovary;
perianth opening obliquely, not expanded. Style
shortly exerted; stamens basally connate; anthers
purple. Fruit globose, to 1cm diameter, purple
(Figures 3–5).
Etymology: named to honour Miss Winifred
Brooke who discovered the plant.
Habitat & Distribution: Steep cliffs and rocky
slopes close to rivers in the foothills of the eastern
Andes, 750–1300m.
Ecological regions: South Andean Yungas and
margins of Dry Chaco.
Occurrence: BO (CH, SC, TR). Map 2.
Although not rediscovered in habitat for many
years, since 2004 it has been observed at several
locations along the eastern margin of the Andean
foothills allowing a clearer assessment of its
variation. At the southern locations the plants are
generally erect, e.g. Ritter’s var. flavispinus,
whereas as one goes north they become
decumbent and eventually pendant. Hence the
former variety/subspecies vulpis-cauda is
included here since its main differentiating
character, pendant stems, is not unique and can
also be seen in several populations of the type.
The geniculate (bent like a knee) flower and
staminodial diaphragm in this species are
effective adaptations to ornithophily, both serving
to prevent insects from reaching the nectar
chamber. Similar adaptations are seen separately
in the two subspecies of C. winteri clearly linking
that species as a florally specialised member of
3. Cleistocactus buchtienii Backeb., Kaktus
ABC pp.189, 411 (1936) Type: between Arque and
Cochabamba, Dept. Cochabamba, Bolivia, 2000m,
not preserved. Neotype (designated here):
Angostura, 2560m; Oct 1947, Cárdenas 4820a
(LIL 531560).
C. angosturensis Cárd., Cact. Succ. J. (US) 28: 60,
figs. (1956). Type: Angostura, Mun. Tarata,
Prov. Arce, Dept. Cochabamba, Bolivia,
2550m, August 1955, Cárdenas 5055 (LIL
C. ayopayanus Cárd., Cact. Succ. J. (US) 28: 58,
figs. (1956). Type: between Morochata and
Independencia, Prov. Ayopaya, Dept.
Cochabamba, Bolivia, 2300m, October 1949;
Cárdenas 5053 (LIL, not found; US, not
recorded). Lectotype (designated here): fig. 35
of the protologue.
C. buchtienii var. flavispinus Cárd., Cact. Succ. J.
(US) 24: 182, fig. (1952). Type: La Angostura,
Mun. Tarata, Prov. Arce, Dept. Cochabamba,
Bolivia, 2570m, January 1948, Cárdenas
4820a (LIL 531560, holo; US 2342649, iso).
C. orthogonus Cárd., Cactus (Paris) xiv: 161
(1959). Type: between Chaqui and Betanzos,
Prov. Saavedra, Dept. Potosi, Bolivia, 2800m,
January 1956, A. Vidaurre in Cárdenas 5500
Figure 5. Cleistocactus brookeae GC857.02 from
east of Monteagudo, Bolivia, in culture.
Photograph: Graham Charles.
156 Bradleya 34/2016
(LIL 531581, US). Note: the altitude stated in
the protologue is not possible at the stated
location and must be an error since the
holotype label at LIL reads “3400m” which is
C. ressinianus rd., Cact. Succ. J. (US) 28: 55,
figs., (1956). Type: between Sucre and
Tarabuco, Prov. Oropeza, Dept. Chuquisaca,
Bolivia, 2700m, February 1949, Cárdenas
5051 (LIL 531567).
C. sucrensis Cárd., Cact. Succ. J. (US) 24: 148.
(1952). Type: Takos, “kilometre 7 from Sucre
to Tarabuco”, Prov. Oropeza, Dept.
Chuquisaca, Bolivia, 2750m, February 1949,
Cárdenas 4820 (LIL 531572, US).
C. tupizensis sensu Backeb., Kakteen-Freund 3:
124 (1934), non Cereus tupizensis Vaupel.
C. tupizensis var. sucrensis (Cárd.) Backeb., Die
Cactaceae II: 1012 (1959). Basionym:
Cleistocactus sucrensis Cárd.
Echinopsis buchtienii (Backeb.) Anceschi & Magli,
Cactusinhabitat South America 2011/2013
p.37 (2013). Basionym: Cleistocactus
buchtienii Backeb.
Stems erect to 2m, branching from the base
and higher, 5cm diameter. Ribs 12–18, low,
slightly grooved above the areole. Areoles <1cm
apart, tomentose, with 10–15 spines, barely
differentiable, acicular to 2.5cm long. Flowers
sparingly from upper part of stem, tubular, 5–8cm
× 9mm, directed upwards at ~45° to the stem,
deep red–purple with exerted green stigma and
purple anthers. Fruit globular, ±1cm diameter,
red. Seeds, 1.2mm, black, pitted (Figure 6).
Etymology: named to honour Otto Buchtien who
collected cacti in Bolivia in 1932 and 1934.
Habitat & Distribution: Sparsely wooded slopes
and plains in arid inter-Andean valleys in the
North. The southern form generally in more arid
locations, 2200–3700m.
Ecological regions: Bolivian Montane Dry Forest
and Central Andean Puna.
Occurrence: BO (CB, CH, PO, TR). Map 3.
The name as now used includes not only
plants from the region of the type locality but also
those from populations from much further south
for which the misapplied name C. tupizensis
(Vaupel) Backeb. was once used. Since the
Figure 6. Cleistocactus buchtienii LM0452.02
south of Kami, Bolivia, 3300m. The form previ-
ously known as C. ayopayanus.
Photograph: Martin Lowry.
Figure 7. Cleistocactus buchtienii tupizensis
GC892.01, Road to Culpina, Bolivia 3120m.
Photograph: Graham Charles.
Bradleya 34/2016
original observation by Mottram (2002), the
basionym of this combination has now been
conclusively demonstrated (Leuenberger, 2012) as
a synonym of Oreocereus celsianus (Salm-Dyck)
Riccobono. The plants to which the name C.
tupizensis have been misapplied show
considerable differences from the type. They are
generally shorter in stature, have more ribs and
more, often hyaline spines giving them a silver
appearance. The flowers are also much shorter (4–
5cm vs 8–9cm) with fewer hairs on the
hypanthium (Figure 7). These characters give the
plants an appearance similar to that of C.
hyalacanthus (K. Schum.) Rol.-Goss. and it may
be better to consider these southern populations
under that name rather than here.
Although Backeberg declared no type
specimen, the New Cactus Lexicon (NCL) records
Eggli as documenting that a specimen at ZSS
collected by Buchtien can serve as the type. I can
find no record of this specimen in the herbarium
database of ZSS hence, given the above
discussion, it seems desirable to provide an
unambiguous typification. Backeberg provided no
illustrations in the protologue that could serve as
a lectotype and subsequent illustrations, for
example, in Die Cactaceae, are insufficiently clear.
Of the subsequent names now considered
synonymous, only two are from within a
reasonable proximity of Backeberg’s quoted type
locality. These are C. angosturensis and C.
buchtienii var. flaviflorus, both names published
by Cárdenas with type sheets present at LIL.
Since the latter was obviously considered closer to
C. buchtienii Backeb. by Cárdenas, its specimen
is more representative and there is an excellent
photograph in the protologue so it is an
appropriate choice as neotype. Although the type
sheet does not contain flowers, a description in
Cárdenas’s handwriting states “wine red, long” on
the label and the photograph shows them rising
at an acute angle from the stem, all significant
characters in identifying this species.
4. Cleistocactus candelilla Cárd., Cact. Succ.
J. (US) 24: 146–147, figs., (1952). Type: La Tigre
(Pampa Grande), Prov. Florida, Dept. Santa Cruz,
Bolivia, 1400m, Oct 1950, A. Corro s.n. in
Cárdenas 4819 (LIL, not found; US, not recorded).
Lectotype (designated here): Fig. 94 of the
C. candelilla subsp. piraymirensis (Cárd.)
Mottram, Cact. Syst. Init. 13: 10 (2002).
Basionym: C. piraymirensis Cárd.
C. candelilla var. pojoensis Cárd., Cact. Succ. J.
(US) 24: 147 (1952). Type: Hacienda La
Habana (Pojo), Prov. Carrasco, Dept.
Cochabamba, Bolivia, 2500m; specimen not
designated. Lectotype (designated here): Fig.
95 of the protologue.
C. muyurinensis Ritter, Taxon 13: 114 (1964).
Type: Muyurina, Prov. Valle Grande, Dept.
Santa Cruz, Bolivia, Ritter 821 (U 117562B).
C. piraymirensis Cárd., Cact. Succ. J. (US) 33: 78,
figs. (1961). Type: Río Piraymiri, Prov. Valle
Grande, Dept. Santa Cruz, Bolivia, 1800m,
Sep 1959, Cárdenas 5526 (LIL, not found).
Lectotype (designated here): Fig. 40 of the
C. pojoensis (Cárd.) Backeb., Die Cactaceae II:
1003 (1959). Basionym: C. candelilla var.
pojoensis Cárd.
Echinopsis candelilla (Cárd.) Anceschi & Magli,
Cactusinhabitat South America 2011/2013
p.38 (2013). Basionym: Cleistocactus
candelilla Cárd.
Seticleistocactus piraymirensis (Cárd.) Backeb.,
Descr. Cact. Nov. 3: 13 (1963). Basionym: C.
candelilla var. piraymirensis Cárd.
C. vallegrandensis Cárd., Cact. Succ. J. (US) 33:
75–76, figs. (1961). Type: between Vallegrande
and Candelaria, Prov. Valle Grande, Dept.
Santa Cruz, Bolivia, 2600m, October 1959;
Cárdenas 5523 (LIL, not found). Lectotype
Figure 8. Cleistocactus candelilla BCMdV206.03
from south of Mataral, Bolivia, 1670m in culture.
Photograph: Martin Lowry.
158 Bradleya 34/2016
Figure 10. Cleistocactus candelilla LM0736.01
east of Pena Colorado, Bolivia, 1675m. A form
with exceptionally hairy flower tubes.
Photograph: Martin Lowry.
Figure 9. Cleistocactus candelilla LM0730.02
east of Mataral, Bolivia, 1365m.
Photograph: Martin Lowry.
Figure 11. Cleistocactus capadalensis LM0865.01
east of Capacdala, Bolivia, 2015m, near the type
locality. Photograph: Martin Lowry.
Figure 12. Cleistocactus capadalensis LM0768.01
west of the Rio Pilcomayo near Kollpa, Bolivia,
2120m. Photograph: Martin Lowry.
Figure 13. Cuttings of Cleistocactus chryso-
cephalus from near Inquisivi, Bolivia.
Photograph: Holger Kelsch
Bradleya 34/2016
(designated here): fig. 34, of the protologue.
Syn. nov.
A small shrub to 1m tall, branching from the
base. Stems, generally many, erect, ascending or
decumbent, 2–3cm diameter with 10–15 ribs
bearing areoles < 1cm apart. Spines
differentiated; radials, 12–18, yellow to white, up
to 1cm long; centrals, 1–4, yellowish brown, 10–
25mm. Flowers, straight, 3–6cm × 4–7mm
diameter; hypanthium magenta, sometimes
covered with white hairs to 1cm; outer tepals
somewhat expanded, magenta or yellowish, inner
tepals magenta. Fruit globose, orange-red to
purple, ±12mm (Figures 8–10).
Etymology: the Spanish vernacular name for the
plant meaning “small candle” and apparently
referring to the flame-like flowers.
Habitat & Distribution: dry semi-deciduous thorn
scrub in the higher interior valleys, 1300–2800m.
Ecological regions: Bolivian Montane Dry Forest.
Occurrence: BO (CB, CH, SC). Map 3.
A very frequent plant in the dry thorn-scrub of
the north-eastern Andes. In the west its range
overlaps with that of C. parviflorus (K. Schum.)
Rol.-Goss. where in many places they grow
sympatrically.It is possible that some inter-
gradation of the two species occurs giving rise to
plants with mixed characteristics (Figure 28).
5. Cleistocactus capadalensis Ritter, Kakt.
Südam. 2: 677 (1980). Type: slopes above Río
Pilcomayo, Capadala, Prov. Azurduy, Dept.
Chuquisaca, 1958, Ritter 828 (U).
Stems ascending or decumbent 1–1.5m long ×
3–5cm diameter branching from the base. Ribs,
13–18, bearing areoles 5–8mm apart. Spines 8–
12, bright yellow, the centrals, 2–3, thicker and
longer, to 3cm. Flowers straight, directed upwards
at a slight angle, 2.5–3.5cm × 8mm. Receptacle
green; hypanthium yellow-orange; inner petals
red. Stigma-lobes green, slightly exerted; anthers
magenta. Fruit globose, approx. 1cm diameter
with persistent floral remains, initially green
ripening to red. Seeds unknown (Figures 11 & 12).
Etymology: referring to the place at which Ritter
discovered the species: Capadala in the gorge of
the Río Pilcomayo.
Habitat & Distribution: Sparsely wooded slopes
on both sides of the Río Pilcomayo in sandy soil,
Ecological regions: Bolivian Montane Dry Forest.
Occurrence: BO (CH). Map 4.
Originally described without details of flower
and fruit this taxon was long considered a
synonym of C. buchtienii Backeb. until
rediscovered in 2012 close to the type locality by
the author and Mats Winberg. The description
above is expanded from that of Ritter to include
details of the flowers and fruit observed in
habitat. It was not recognized in the NCL and
consequently its conservation status was not
assessed. Given its small area of extent and area
of occupancy with the threats due to low level
subsistence farming it is probably best to consider
it as Vulnerable.
6. Cleistocactus chrysocephalus (Ritter)
Mottram, Cact. Syst. Init. 13: 11 (2002).
Cephalocleistocactus chrysocephalus Ritter,
Succulenta (NL) 1959: 107–111 (1959). Type:
gorge below Inquisivi, Prov. Inquisivi, Dept.
La Paz, Bolivia, 2000m, October 1954, Ritter
326 (ZSS).
A tall shrub of 2–5m branching both from the
base and higher. Stems 3–5cm diameter with 11–
14 ribs bearing areoles 5–8mm apart. Spines
differentiated; radials about 15, pale, about 7mm
long; centrals to 6, yellowish, 7–30mm long.
Flowers produced from a lateral pseudocephalium
up to 1m long and 3–4cm wide densely packed
with spines and bristles to 3cm long. The ribs in
this zone reduced and the areoles closer together.
Flowers ±straight, 5cm long, yellow to red. Fruit
greenish-red, 2cm long × 3cm broad. Seeds small,
0.7mm, black, shiny, nearly smooth (Front cover
picture and Figures 13–15).
Etymology: from Greek; “golden-headed”,
referring to the colour of the cephalium.
Habitat & Distribution: Sparsely wooded slopes of
seasonally dry valleys of the Yungas, 1900–
Ecological regions: Bolivian Yungas.
Occurrence: BO (CB, LP). Map 9.
Originally considered by Ritter as belonging to
a separate genus, Cephalocleistocactus, this
unusual plant has rarely been seen in habitat.
There has however, been one recent report
(Kelsch, 2016) that it still exists near the type
locality. Ritter’s criteria for separating the genus
were based on a longer and wider nectar chamber
than found in Cleistocactus and on the cephalium.
Although exceptional, the cephalium as described
by Ritter is not a true cephalium as in the sunken
cephalium of Espostoa Britton & Rose and should
still be considered a pseudo-cephalium. Indeed, it
is possible that it represents a closer link to
160 Bradleya 34/2016
Figure 14. Cleistocactus chrysocephalus from near Inquisivi, Bolivia, 2430m.
Photograph: Holger Kelsch
Figure 16. Cleistocactus dependens LM0668.01
east of Samaipata, Bolivia, 850m.
Photograph: Martin Lowry
Figure 15. Cleistocactus chrsocephalus Ritter 326
in culture. Photograph: Graham Charles.
Bradleya 34/2016
Vatricania Backeb. since recent molecular
evidence (Schlumpberger & Renner, 2012) has
placed this monotypic genus closer to
Cleistocactus than to Espostoa. Similar, though
less extreme, pseudocephalia are produced by a
small number of other species in the genus (C.
ritteri, C. variispinus and C. viridiflorus). All of
them are found in similar habitats in the Yungas
on the northern edge of the Andes (Map 9).
7. Cleistocactus dependens Cárd., Cact. Succ.
J. (US) 24: 143–144, figs. (1952). Type: “humid
rocky slopes”, between La Angostura and La
Negra, Prov. Florida, Dept. Santa Cruz, Bolivia,
1200m, Oct 1950, A. Corro in Cárdenas 4817 (LIL,
not found; US 2906781, iso; UC, paraiso).
Seticleistocactus dependens (Cárd.) Backeb.
Kakteenlexicon p.406 (1966). Basionym:
Cleistocactus dependens Cárd.
Stems dependent, procumbent or clambering
to 2m, 3–4cm diameter, branching basally, dull
green. Ribs 10–12, rounded bearing areoles 1–
1.5cm apart. Radial spines 8–12, 2–4mm, acicular,
grey. Central spines 1–3, 1–1.5cm, thicker and
darker. Flowers 4–4.5cm × 0.7cm, straight;
pericarpel dark purple, hypanthium strawberry
red, both with scales bearing conspicuous short
white hairs and bristles; distal perianth-segments
green. Filaments white, anthers purple. Style
greenish-white, slightly exerted. Fruit globular,
approx. 1cm diameter, deep red. Seeds 1mm long,
shiny, black, finely pitted (Figures 16 & 17).
Etymology: from the Latin dependens; referring to
the hanging nature of the stems.
Habitat & Distribution: Hanging from cliffs in the
gorges of the north-easternmost rivers of the
Andes, 800–1500m.
Ecological regions: Bolivian Montane Dry Forest.
Occurrence: BO (SC). Map 5.
Although the NCL originally considered it a
disjunct form of C. candelilla Cárd., it can easily
be distinguished by its much thicker, decumbent
to sprawling branches, rounded ribs, fewer,
shorter and thicker spines and flowers with green
inner tepals. It has recently (2011) been recorded
in the region of the type locality by the author,
John Carr and Moises Mendoza.
Backeberg transferred it to his genus
Seticeistocactus based on the presence of the
conspicuous hairs on the floral tube. However,
this feature is found sporadically throughout the
genus and cannot be used as a generic character.
8. Cleistocactus hildegardiae Ritter, Kakt.
Südam. 2: 685, figs., (1980). Type: Caña Cruz,
Prov. Mendez, Dept. Tarija, Bolivia, 1962, Ritter
1126 (U).
Borzicactus janae Halda & Horacek, Cactaceae
etc. (Bratislava) 13: 17–19 (2003). Type:
Estancia Caña Cruz, Prov. Mendez, Dept.
Tarija, Bolivia, 3200 m, December 2002;
Halda & Horacek s.n. (PRC 4723). Syn. nov.
A small shrub with erect stems freely
branching from the base. Stems to 20–40cm × 1–
2cm diameter with 15–19 ribs bearing close-set
areoles. Spines 18–28, acicular, yellow to red-
brown, radials numerous, 3–8mm; centrals 5 –8,
thicker than radials, 1–3cm long. Flowers tubular,
straight, 2.5–3.5cm × 5–6mm; magenta to purple,
outer tepals spreading. Fruit globose, 1cm
diameter (Figure 18).
Etymology: named to honour Frau Hildegard
Winter, Ritter’s sister who distributed his
collected seeds.
Habitat & Distribution: dry high altitude valleys
sides in stony soils, 2600–3300m.
Ecological regions: Bolivian Montane Dry Forest,
Central Andean Puna and margins of Southern
Andean Yungas.
Occurrence: BO (CH, TR). Map 5.
Figure 17. Cleistocactus dependens HNT57784 in
culture showing the hairs and bristles on the hy-
pantheum. Photograph: Graham Charles.
162 Bradleya 34/2016
Figure 19. Cleistocactus hyalacanthus
LM0602.01 north of Jucanas, Bolivia, 2260m
Photograph: Martin Lowry.
Figure 20. Cleistocactus hyalacanthus GC159.01,
the form previously known as C. jujuyensis, from
El Alisal, Quebrada del Toro, Salta, Argentina in
culture. Photograph: Graham Charles.
Figure 21. Cleistocactus laniceps LM0458.02
south-east of Independencia, Bolivia, 2110m
showing the woolly buds after which it is named.
Photograph: Martin Lowry.
Figure 18. Cleistocactus hildagardiae LM0546.02
in the Paicho valley north of Iscayachi, Bolivia,
3170m. Photograph: Martin Lowry.
Bradleya 34/2016
This is the smallest member of the genus and
at one time was regarded a synonym of C.
hyalacanthus (K. Schum.) Rol.-Goss. but now
considered distinct since the two species have
been found growing sympatrically at several
locations. Originally known only from its type
locality, the Río Paicho valley in Tarija, but
recently found by the author and Mats Winberg
on the north side of the Río Pilaya in Chuquisaca.
9. Cleistocactus hyalacanthus (K. Schum.)
Rol.-Goss., Bull. Soc. Centr. Agric. Hort. Acclim.
Nice 44: 33 (1904).
Cereus hyalacanthus K. Schum., Gesamtb. Kakt.
p.101 (1897). Type: Prov. Jujuy, Argentina,
Kuntze s.n., not preserved.
Cleistocactus hyalacanthus ssp. tarijensis (Cárd.)
Mottram, Cact. Syst. Init. 13: 11 (2002).
Basionym: C. tarijensis Cárd.
C. jujuyensis (Backeb.) Backeb., Beitr.
Sukkulentenk. Sukkulentenpflege 1942: 49–50
(1942). Basionym: C. strausii var. jujuyensis
C. jujuyensis var. fulvus Ritter, Kakt. Südamer. 2:
480 (1980). Type: Purmamarca, Prov. Jujuy,
Argentina, 1959, Ritter 59b (U).
C. strausii var. fricii (Därfl.) Backeb., Die
Cactaceae II: 1015 (1959). Basionym:
Pilocereus strausii var. fricii Därfl.
C. strausii var. jujuyensis Backeb., Cact. J.
(London) 3: 25 (1934). Type: Jujuy, Argentina.
C. tarijensis Cárd., Cact. Succ. J. (US) 28: 54
(1956). Type: low hills near Tarija city,
Cercado, Dept. Tarija, Bolivia, 1900m, Mar
1952; Cárdenas 5050 (LIL 531575, holo., US,
Demnosa straussii var. luteispina Frič, Möller’s
Deutsch. Gärtn.-Zeit. 44: 170 (1929). Type:
Echinopsis nothohyalacantha Anceschi & Magli,
Cactusinhabitat South America 2011/2013
p.38 (2013). Replaced synonym: Cereus
hyalacanthus K. Schum.
Pilocereus strausii var. fricii Därfl., Monats.
Deutsch. Kakteen-Ges. 2: 55, fig. (1930).
A low shrub with erect stems to 1m tall
branching from the base. Stems 4–6cm diameter
with 15–20 ribs bearing areoles 5–8mm apart.
Spines 20–30, unequal, up to 3cm long, white,
yellow or brown. Flowers in profusion from near
stem apex, directed slightly upwards, 3.5–4cm ×
5–7mm, slightly curved, hypanthium red, tepals
slightly spreading, magenta. Fruit globose, about
12mm diameter, red to orange. Seeds black, shiny
(Figures 19 & 20).
Etymology: from the Latin hyalinus and acanthus
meaning “glassy-spined”.
Habitat & Distribution: Sparsely wooded hills and
cliffs of the eastern Andean mountains, 1700–
Ecological regions: Bolivian Montane Dry Forest,
Central Andean Puna, High Monte and margins
of Southern Andean Yungas.
Occurrence: AR (JY, SA), BO (CH, TR). Map 8.
One of the oldest names in the genus but
recently much confused by application to the
plants discussed under C. buchtienii Backeb. as
the southern form. These two plants can look very
similar when not in flower. C. hyalacanthus
generally produces many flowers simultaneously
whilst on C. buchtienii there are normally only a
few produced together. The flowers of the latter
are also more straight, longer (>4cm) and purple
rather than red. When not in flower
distinguishing them is difficult, however spination
can help as plants of the southern C. buchtienii
tend to have centrals and radials of different but
more or less constant length whilst those of C.
hyalacanthus are quite variable. Otherwise
location may be the only clue; plants of C.
hyalacanthus are plants of the moister eastern
slopes whereas C. buchtienii prefers the drier
valleys further west.
10. Cleistocactus laniceps (K. Schum.) Rol.-
Goss., Bull. Soc. Centr. Agric. Hort. Acclim. Nice
44: 32 (1904).
Borzicactus laniceps (K. Schum.) Backeb., Cact.
Succ. J. (US) 3: 132 (1932). Basionym: Cereus
laniceps K. Schum.
Cereus laniceps K. Schum., Gesamtb. Kakt. p.93
(1897). Type: near Tunari, Dept. Cochabamba,
Bolivia, 1300m, Kuntze (B, not preserved?).
C. laniceps var. plurispinus Ritter, Kakt.
Südamer.2: 673 (1980). Type: below Inquisivi,
Prov. Inquisivi, Dept. La Paz, Bolivia, 1957,
Ritter 603a (U).
A shrub to 4m with erect stems branching
from the base or higher. Stems green, 3–5cm
diameter with ±9 ribs about 1cm high with cross-
furrows, the areoles large, 6mm, and well
separated. Radial spines few, short; centrals,
when present, to 2cm, yellow when new fading to
grey. Flowers from lateral areoles, funnelform,
3.5cm, red, receptacle with copious brown wool.
Fruit globose, ±1cm diameter (Figure 21).
Etymology: from the Latin for “woolly-head” and
referring to the hairy covering of the receptacle.
Habitat & Distribution: Dry valleys of the
northern Andes below the Yungas forests, 2100–
Ecological regions: Bolivian Montane Dry Forest.
164 Bradleya 34/2016
Occurrence: BO (CB, LP). Map 6.
A very distintive species because of its very
few spines and funnelform flower. These giving
rise to the, here discounted, suggestion that it is a
natural hybrid between C. luribayensis Cárd. and
Yungasocereus inquisivensis Ritter.
11. Cleistocactus luribayensis Cárd., Cact.
Succ. J. (US) 28: 59–60, figs., (1956). Type:
Luribay slopes, Prov. Loayza, Dept. La Paz,
Bolivia, 2600m, Apr 1955, Cárdenas 5054 (LIL,
not found; US, not found). Lectotype (designated
here): fig. 37 of the protologue.
C. glaucus Ritter, Taxon 13: 114 (1964). Type:
Tirata, Prov. Murillo, Dept. La Paz, Bolivia,
Ritter 112 (U 117611B).
C. glaucus var. plurispinus Ritter, Taxon 13: 114
(1964). Type: border of provinces Loayza and
Murillo, Dept. La Paz, Bolivia, Ritter 112a (U
C. granjaensis Ritter, Kakt. Südamer. 2: 687
(1980). Type: La Granja, Prov. Murillo, Dept.
La Paz, Bolivia, 3000m, 1953, Ritter 106 (U,
not found). Lectotype (designated here): the
specimen T4477 at ZSS collected by Ritter at
Aramani, Prov. Murillo, Dept. La Paz, Bolivia,
October 1954.
A shrub to 3m with numerous erect stems,
branching primarily from the base. Stems 4–6cm
diameter with about 19 ribs bearing areoles ±1cm
apart. Spines 16–22, white, yellow or pale brown,
5–15mm long, the centrals, 1 or 2, the longest.
Flowers tubular, ±straight, 3cm × 5mm, red. Fruit
globose to2cm diameter, salmon pink (Figure 22).
Etymology: referring to the place of its first
discovery, Luribay, a small town in southern
Department La Paz.
Habitat & Distribution: Northern dry inter
Andean valleys west of the Yungas, 2500–3600m.
Ecological regions: Bolivian Montane Dry Forest.
Occurrence: BO (LP). Map 6.
Included here are three taxa described by
Ritter which have not so far been rediscovered in
habitat and for which no material is known in
cultivation. The only illustrations Ritter
published for C. glaucus (Kakt. Südamer. 2: Abb.
667 & 668) do suggest a close relationship with C.
luribayensis. Unfortunately, Ritter himself never
published an illustration for C. granjaensis; he
does, however, refer to a photograph by Backeberg
(Die Cactaceae VI: 3689 Abb. 3350b) which he
concedes may be the correct plant observing
though that it does not branch as high as
illustrated when seen in habitat. Backeberg also
published side-by-side black and white
photographs (Die Cactaceae II: 1022 Abb. 954 &
955) of both of Ritter’s plants under their
collection numbers, which reveal significant
differences between them. He considered C.
granjaensis close to C. brookeae Cárd. Indeed, if
the plant illustrated in the colour photograph
(Abb. 3350b) really is Ritter 106 then it bears little
resemblance to plants of C. luribayensis seen by
the author at the type locality and would be
worthy of separate recognition.
12. Cleistocactus micropetalus Ritter, Kakt.
Südamer. 2: 675, figs., (1980). Type: Concepción,
Prov. Avilez, Dept. Tarija, Bolivia, 1958, Ritter
830, (U).
C. clavicaulis Cárd., Cactus (Paris) No. 80-81: 20
figs (1964). Type: near Angosto de Villa
Montes, Prov. Gran Chaco, Dept. Tarija,
Bolivia, 460m, May 1958, Cárdenas 5525 (LIL,
not found). Lectotype (designated here): the
photograph with the protologue.
C. crassicaulis Cárd., Cact. Succ. J. (US) 33: 77,
figs., (1961). Type: near Angosto de Villa
Montes, Prov. O’Connor, Dept. Tarija, Bolivia,
460m, March 1952, Cárdenas 6124 (LIL, not
Figure 22. Cleistocactus luribayensis LM0447.02
at Luribay, Bolivia, 2580m.
Photograph: Martin Lowry.
Bradleya 34/2016
found). Lectotype (designated here): fig. 38 of
the protologue.
C. crassicaulis var. paucispinus Ritter, Kakt.
Südamer. 2: 674–675 (1980). Type: bottom of
gorge N of Cajas, Prov. Mendez, Dept. Tarija,
Bolivia, Ritter 829a (U).
C. tominensis ssp. micropetalus (Ritter) Mottram,
Cact. Syst. Init. 13: 12 (2002). Basionym:
Cleistocactus micropetalus Ritter.
C. viridialabastri Cárd., Cact. Succ. J. (US) 35:
201, figs. (1963). Type: between Villa Montes
and Palos Blancos, Prov. Gran Chaco, Dept.
Tarija, Bolivia, 600m, May 1959; Cárdenas
5549 (LIL, not found; US, not recorded).
Lectotype (designated here): the photograph in
the protologue.
An erect shrub branching from the base with
stems 1.5–3m tall. Stems 6–8cm diameter with
16–18 ribs bearing areoles ±1cm apart. Spines
acicular, golden yellow; radials 5 or 6, 5–15mm
long; centrals at least 1, 2–4cm. Flowers straight,
tubular, 2–3cm, perianth rarely expanding;
hypanthium green becoming pinkish at anthesis,
covered by short white hairs. Fruit globose, 15–
20mm diameter, yellowish (Figures 23 & 24).
Etymology: from Latin referring to the tiny tepals.
Habitat & Distribution: Lowlands and inter
Andean valleys of southern Bolivia, west of the
Chaco margin, 750–2300m.
Ecological regions: Southern Andean Yungas, Dry
Chaco and Central Andean Puna.
Occurrence: BO (TR). Map 3.
The NCL treatment considers it a subspecies
of C. tominensis (Weingart) Backeb. However, I
consider it sufficiently different through its fewer
ribs and fewer, less conspicuous radial spines that
allow sight of the body justifying its recognition
as a distinct species. It has a well-defined
distribution in the Andean foothills of central
Tarija at altitudes generally less than those of C.
tominensis. Further, C. tominensis is less clearly
defined, particularly at its northern border, and it
is therefore better to maintain C. micropetalus as
distinct to avoid confusion.
Figure 23. Cleistocactus micropetalus LM0520.02
west of Palos Blancos, Bolivia, 1020m. The form
previously known as C. crassicaulis.
Photograph: Martin Lowry.
Figure 24. Cleistocactus micropetalus GC863.02
south of Palos Blancos, Bolivia, 880m.
Photograph: Graham Charles.
166 Bradleya 34/2016
Figure 28. Cleistocactus aff. parviflorus
LM0839.01 east of Pasorapa, Bolivia, 1370m. An
especially hairy-flowered form that may be a hy-
brid with C. candelilla.
Photograph: Martin Lowry.
Figure 27. Cleistocactus parviflorus LM0584.03
north of Vallegrande, Bolivia, 2000m.
Photograph: Martin Lowry.
Figure 25. Cleistocactus morawetzianus
LM0354.01 at Villa Azul, Peru, 2590m.
Photograph: Martin Lowry.
Figure 26. Cleistocactus morawetzianus in cul-
ture showing the characteristic downward curva-
ture of the flower. Photograph: Martin Lowry.
Bradleya 34/2016
13. Cleistocactus morawetzianus Backeb.,
Jahrb. Deutsch. Kakteen-Ges. 1: 77, fig. (1936).
Type: near Mariscal Cáceres, Ayacucho, Peru (fide
Backeb. Die Cactaceae 2: 1008 (1959)) (ZSS?).
C. morawetzianus var. pycnacanthus Rauh &
Backeb., Descr. Cact. Nov. I: 17 (1957). Type:
near Ayacucho, Peru, 2700m, 1954, Rauh K75
C. pycnacanthus (Rauh & Backeb.) Backeb.,
Kakteenlexikon p.88 (1966). Basionym: C.
morawetzianus var. pycnacanthus Rauh &
C. villaazuensis Ritter, Kakt. Südamer. 4: 1359
1360, fig. (1981). Type: Villa Azul, Dept.
Huancavelica, Peru, 2000m, September 1964,
Ritter 1296, (U).
A shrub to 2m with erect stems branching
from the base. Stems grey-green, ~5cm diameter
with 12–14 ribs bearing areoles ±1cm apart,
distinctly cross-furrowed. Spines 10–14, unequal,
the longest to 1.5–5cm, yellow at first becoming
grey. Flowers 6–8cm × 9mm, white or green
occasionally becoming tinged with pink at
anthesis, distinctly down-curved, style exerted;
self-fertile. Fruit globose with small scales, ~1cm
diameter, brownish-green. Seeds 1.2 × 0.8mm,
keeled, shiny black (Figures 25 & 26).
Etymology: named to honour Victor Morawetz, an
American philanthropic lawyer who financed two
of Backeberg’s expeditions to South America.
Habitat & Distribution: Dry inter-Andean valleys
of the Río Mantaro and Río Apurimac drainage
systems, 2200–3000m.
Ecological regions: Peruvian Yungas.
Occurrence: PE (AP, AY, HV). Map 7.
A very distinctive species since it is the only
one with long, down-curved greenish-white
flowers. However, there are some forms which
produce a pink or reddish colouration of the
hypanthium at anthesis and it is then possible to
confuse them with C. pungens Ritter. In these
cases the thick grey-green stems compared to the
thin dark green stems of C. pungens may be a
useful character. As far as I am aware it is the
only species which appears to be self-fertile.
It is rather surprising that this species and C.
pungens Ritt. are the only two taxa occurring in
Peru. Even more so since they are highly disjunct
being separated from the nearest of the remaining
species by a distance of at least 450km. The
nearest Bolivian species is C. viridiflorus Backeb.,
found in humid valleys north-west of La Paz.
Ecologically, the two areas are relatively similar
and both are termed “Yungas” by Olsen et al.
14. Cleistocactus parviflorus (K. Schum.) Rol.-
Goss., Bull. Soc. Centr. Agric. Hort. Acclim. Nice
44: 32 (1904).
Cereus areolatus K. Schum., Gesamtb. Kakt. p.100
(1897). Type: mid-Andean region of South
America, Kuntze s.n. (B†?).
C. parviflorus K. Schum., Gesamtb. Kakt. pp.100–
101 (1897). Type: near Parotani, SW of
Cochabamba, Bolivia, 2400m, Kuntze s.n.
Cleistocactus areolatus (K. Schum.) Riccob., Boll.
Ort. Bot. Palermo viii: 264–265 (1909).
Basionym: Cereus areolatus K. Schum.
C. fusiflorus Cárd., Cactus (Paris) No. 57: 252–
254, figs (1957). Type: near Saipena, Prov.
Valle Grande (sic, actually Prov. Caballero),
Dept. Santa Cruz, Bolivia, 1900m, April 1957,
Cárdenas 5491 (LIL, not found; US, not
recorded). Lectotype (designated here): the
photograph in the protologue.
C. herzogianus Backeb., Blätt. Kakteenf. Pt. 6, p.4
(1934). Type: “in Quebradas”, Prov. Arque,
Dept. Cochabamba, Bolivia, 2500m.
C. ianthinus Cärd., Cact. Succ. J. (US) 28: 57–58,
figs (1956). Type: “on the Río Caine mountain
slopes”, near Calahuta, Prov. Charcas, Dept.
Potosi, Bolivia, 2000m, November 1949,
Cärdenas 5086 (LIL, not found; US, not
recorded). Lectotype (designated here): fig. 33,
of the protologue.
C. parviflorus var. aiquilensis Ritter, Kakt. and.
Sukk. 14: 102–104, figs (1963). Type: Aiquile,
Prov. Campero, Dept. Cochabamba, Bolivia,
December 1954, Ritter 359, (erroneously as
539 in the protologue) (U).
C. parviflorus var. comarapanus Ritter, Kakt.
Südamer. 2: 683–684 (1980). Type: “east of
Comarapa”, Prov. Caballero, Dept. Santa
Cruz, Bolivia, 1954, Ritter 358 (U).
C. parviflorus var. herzogianus (Backeb.) Backeb.,
Descr. Cact. Nov. 3: 5 (1963). Basionym: C.
herzogianus Backeb.
Echinopsis parviflora (K. Schum.) Anceschi &
Magli, Cactusinhabitat South America
2011/2013 p.39 (2013). Basionym: Cereus
parviflorus K. Schum.
A shrub to 3m tall branching from the base
and higher. Stems spreading, ±3cm diameter with
12–14 ribs bearing close-set areoles. Spines
differentiated; radials 5–9, 3–7mm long, dark
yellow; centrals at least 1, longer and thicker.
Flowers 2.5–3.5cm × 8mm, hypanthium pale
yellow at first becoming red, perianth
unexpanded; inner tepals purple. Fruit globose,
±1cm diameter, yellow (Figures 27 & 28).
Etymology: from the Latin parvus and flos;
meaning “small-flowered”, the character which
168 Bradleya 34/2016
Schumann considered to distinguish it from the
other species he knew.
Habitat & Distribution: Seasonally dry inter
Andean valleys amongst shrubs and trees in stony
ground, 1300–3000m.
Ecological regions: Bolivian Montane Dry Forest.
Occurrence: BO (CB, CH, SC). Map 5.
As one of the oldest names in the genus and a
species with a large distribution it has acquired
several synonyms. The defining characters of the
species are its mesotonic branching habit, few
spines and short red flowers. This combination is
not found in the other species with which it is
often sympatric. In the west of its distribution it
occasionally grows in close proximity to C.
buchtienii Backeb., although the latter prefers
more exposed situations. Both species can form
large mesotonically branched shrubs, but C.
parviflorus can be distinguished by its fewer,
stiffer spines and shorter, red flowers. In the east
its range overlaps extensively with that of C.
candelilla Cárd. and the two can frequently be
found together. Generally it is easy to tell them
apart since C. candelilla is a much smaller,
basitonically branching, thinner-stemmed plant
with magenta flowers. On occasion, however, I
suspect they produce hybrids which, although
having intermediate characters, can be difficult to
recognise as such. A remarkable example is plants
found south east of Pasorapa, where the tall
yellow-spined plants have large magenta flowers
copiously covered in long white hairs.
15. Cleistocactus pungens Ritter, Taxon 13: 115
(1964). Type: Río Pampas, frontier of Depts
Ayacucho and Apurímac, Peru, 1957, Ritter 664
C. brevispinus Ritter, Kakt. Südamer. 4: 1360–
1361, fig. (1981). Type: Pachachaca gorge near
Abancay, Dept. Apurimac, Peru, 1964, Ritter
1297 (U).
A short shrub to 1.5m with stems branching
from the base. Stems grey-green, 2.5–3.5cm
diameter with 12–15 ribs, the areoles 5–7mm
apart. Spines reddish brown; radials 8–12, 2–
10mm long; centrals 2–5, 2–5cm, stronger.
Flowers 7–8cm × 6–8mm diameter, red. Fruit
globose, ±2cm diameter, pale red (Figure 29).
Etymology: from the Latin pungo, to prick. Ritter
Figure 29. Cleistocactus pungens LM0363.02
west of Chincheros, Peru, 2060m.
Photograph: Martin Lowry.
Figure 30. Cleistocactus reae LM0498.01 north of
Quime, Bolivia, 2760m.
Photograph: Martin Lowry.
Bradleya 34/2016
gave no explanation but presumably, referring to
the sharp spines.
Habitat & Distribution: Seasonally dry inter
Andean valleys of central eastern Peru, 1100–
Ecological regions: Peruvian Yungas.
Occurrence: PE (AP, AY, HV). Map 7.
Only this and one other species, C.
morawetzianus Backebg., are found in Peru. Both
species grow in the same region, along the valleys
of the lower Río Apurimac and its tributaries, and
although occasionally found sympatric they
generally occur at different altitudes with C.
pungens preferring the lower slopes below
2400m. When found together they are relatively
easy to distinguish, especially when in flower.
16. Cleistocactus reae Cárd., Cactus (Paris) 12:
251–252, figs (1957). Type: Quime, Prov.
Inquisivi, Dept. La Paz, Bolivia, 3000m,
November 1956, Cárdenas 5490 (LIL 531554,
holo., US iso.).
A many stemmed shrub branching from the
base. Stems erect, 1–2m × 3–4.5cm diameter with
18–20 ribs. Areoles only 5mm apart bearing 16–20
spines, radials 12–16, 5–10mm long, hyaline;
centrals 3 or 4, 1–3cm long, stronger, yellowish.
Flowers clustered near the stem apex, 5–7cm × 6–
9mm, straight, presented perpendicular to the
stem, hypanthium scales with brown hairs; tepals
strawberry red. Style not exerted. Fruit globose to
2.5cm diameter. Seeds shiny, dark brown, 1.5mm
long (Figure 30).
Etymology: named to honour the Bolivian
agronomist Julio Rea who worked at the
Universidad Mayor de San Simón in Cochabamba
from where Cárdenas made his descriptions of
Habitat & Distribution: Amongst grass and
shrubs in loamy soil on steep slopes of clearings
in the Yungas, 2700–3000m.
Ecological regions: Bolivian Yungas.
Occurrence: BO (LP). Map 6.
One of several micro-endemic species. It has
some similarities with the typical form of C.
buchtienii and grows in an area only 40km further
west at a similar altitude. At present it is little
known and further observations are required to
determine its status.
17. Cleistocactus ritteri Backeb.,Kakt. and.
Sukk. 10: 163–165, fig. (1959). Type: a flowering
plant of Ritter 325 growing in the collection of Dr.
Karius at Muggensturm (Baden). Ritter gave the
details of this collection as: at the Río La Paz,
Plazuela, Prov. Sud-Yungas frontier with Prov.
Loayza, Dept. La Paz, Bolivia, 1100m, 1953.
Lectotype (designated here): the illustration in the
Cephalocleistocactus ritteri (Backeb.) Backeb., Die
Cactaceae VI: 3691, figs. (1962).
A few-stemmed shrub, 1–2m high, branching
from the base. Stems ascending, occasionally
prostrate, thin, ±3cm diameter, with 12–14 ribs
bearing areoles 5mm apart. Radial spines to 30,
6–8mm long, very thin and fragile, white. Central
spines – when distinguishable – a little longer and
thicker, pale yellow. Flowers from areoles near
the stem apices producing additional conspicuous
white bristles, 3–4cm × 5mm, lemon yellow–
orange. Fruit green, 9–12mm diameter. Seeds
black, pyriform, 0.8mm × mm (Figure 31).
Etymology: named to honour Friedrich Ritter,
German explorer who spent many years searching
for cacti in South America and was the first to
discover the species.
Figure 31. Cleistocactus ritteri LM0502.02 from
the type locality near La Plazuela, Bolivia, 1070m
in culture. Photograph: Martin Lowry.
170 Bradleya 34/2016
Habitat & Distribution. On slopes beneath trees
in humid Yungas forest of the Río La Paz
drainage, 1000–1100m. Currently known only
from this one small area.
Ecological regions: Bolivian Yungas.
Occurrence: BO (LP). Map 9.
There is some debate regarding the validity of
this name since Backeberg explicitly cited the
type as a living plant in the collection of Dr Karius
of Muggenstürm, Baden thus rendering his
description invalid according to ICN Art 8.4.
However, according to Eggli et al. (1995), there
had been multiple specimens of Ritter 325 present
at ZSS since 1954 thus, in their opinion,
validating the name. Undoubtedly though, this
material was not consulted by Backeberg. Ritter’s
subsequent re-description in 1980 (Kakt.
Südamer. 2: 687–688) cannot be considered a
validation since he did not specify a replacement
type even though he had deposited material of the
type collection at ZSS. His illustration (Abb. 659)
was actually a reprinting of Backeberg’s Abb.
3353 from Die Cactaceae, presumably a
photograph of the living plant in the collection of
Dr Karius. This photograph (Abb. 3353) is
undoubtedly of the correct material but possibly
not available when Backeberg made his
description since he did not publish it in the
protologue using instead a photograph by Kilian.
A second photograph published by Backeberg in
Die Cactaceae (Abb. 3352) is clearly of the same
plant as in the protologue but taken from a
different angle. In view of this situation I have
chosen to select the illustration in the protologue
as the lectotype for this noteworthy species.
A third illustration (Die Cactaceae II: 1023
Abb. 956) was published prior to the description
where the text repeats Ritter’s description, in his
1956 catalogue, of the flower colour as orange.
This colour has never been seen on plants in
cultivation and was not found by the author on
plants at the type locality. A red or pink flowered
form sometimes encountered in cultivation is said
to be a hybrid of which the other parent is
18. Cleistocactus samaipatanus (Cárd.) D.
Hunt, Bradleya 5: 92 (1987). Basionym:
Bolivicereus samaipatanus Cárd.
Bolivicereus brevicaulis Ritter, Kakt. damer. 2:
703, figs. (1980). Type: Mairana, Prov. Florida,
Figure 32. Cleistocactus samaipatanus GC854.01
west of Ipati Jctn, Santa Cruz, 1014m
Photograph: Graham Charles
Figure 33. Cleistocactus samaipatanus GC918.04
from north of Vallegrande, Bolivia in culture.
Photograph: Graham Charles
Bradleya 34/2016
Dept. Santa Cruz, Bolivia, 1958, Ritter 839,
B. croceus Ritter, Kakt. Südamer. 2: 704–705, fig.
(1980). Type: “probably” Puente Azero, border
of Prov. Tomina and Prov. Azero, Dept.
Chuquisaca, Bolivia, Ritter 1471, (U).
B. rufus Ritter, Kakt. Südamer.2: 703–704, fig.
(1980). Type: Puente Azero, border of Prov.
Tomina and Prov. Azero, Dept. Chuquisaca,
Bolivia; Ritter 840, (U).
B. samaipatanus Cárd., Cact. Succ. J. (US) 23: 91
(May 1951). Type: near “El Fuerte” de
Samaipata, Prov. Florida, Dept. Santa Cruz,
Bolivia, 1890m, May 1948; A. Corro in
Cárdenas 4395 (LIL 531550, holo.; US).
B. samaipatanus var. multiflorus Cárd., Cact.
Succ. J. (US) 23: 93–94, fig. (May 1951). Type:
La Angostura, Prov. Florida, Dept. Santa
Cruz, Bolivia, September 1949, Cárdenas
4395a (LIL 531549, holo.; US).
Borzicactus samaipatanus (Cárd.) Kimnach, Cact.
Succ. J. (US) 32: 93 (1960). Basionym:
Bolivicereus samaipatanus Cárd.
B. samaipatanus var. multiflorus (Cárd.) Krainz,
Kakteen, Lief. 21: C Vb. (1962). Basionym:
Bolivicereus var. multiflorus Cárd.
Echinopsis samaipata (Cárd.) Anceschi & Magli,
Cactusinhabitat South America 2011/2013
p.39 (2013). Basionym: Bolivicereus
samaipatanus Cárd.
A sprawling, decumbent or erect shrub
branching from the base. Stems to 1.5m × 3–4cm
diameter with 14–16 low rounded ribs bearing
areoles 3–4mm apart. Spines 13–22, unequal, 4–
30mm, yellow or brown at first becoming grey.
Flowers 3.5 × 2cm, bright red; hypanthium S-
shaped, perianth expanded obliquely. Stamens
and style exerted with filaments and anthers
purplish. Nectar chamber closed by staminodial
hairs. Fruit ±globose, 9–11 × 7–9mm, reddish
Key to subspecies of C. samaipatanus
1. Stems sprawling or decumbent, spines few to
many, stiff (Andean foothills) (Figures 32 & 33)
a. ssp. samaipatanus
Stems erect, spines many, fine (Chiquitano
Dry Forest, Prov. Chiquitos) (Figures 34 & 35)
b. ssp. divi-miseratus
a. ssp. samaipatanus
Etymology: referring to the place of discovery of
the plant, Samaipata, then a small market town
in the Andean foothills west of Santa Cruz.
Habitat & Distribution: On cliffs and steep slopes
in seasonally dry inter Andean valleys of the
north-eastern Andes, 1300–2100m.
Ecological regions: Southern Andean Yungas,
Bolivian Montane Dry Forest, Dry Chaco.
Occurrence: BO (CH, SC). Map 4.
Cárdenas considered his Bolivicereus
samaipatanus as distinct based solely on the wide
opening of the perianth compared with the then
known members of the genus Cleistocactus. It is
now one of three taxa with flowers having
expanded perianths retained in Cleistocactus due
to the otherwise conforming characters,
particularly the lateral positioning of the flowers.
The species is quite widely distributed and has
been found growing sympatrically with six other
species of the genus without the appearance of
hybrids. Notably however, the one species with
which hybrids have been raised in cultivation,
Cleistocactus winteri ssp. winteri, to produce the
cultivar ‘Jupiter’ (Mottram, 1989), is not amongst
these six species.
The apparent disjunction in distribution in the
Andean portion of its range is primarily due to
lack of exploration in the area between
Vallegrande and Moteagudo. The mountain
ranges in the region seem highly suitable for the
species and it is likely to be found there
b. ssp. divi-miseratus (Cárd. ex Backeb.) M.
Lowry comb. et stat. nov. Basionym:
Bolivicereus samaipatanus var. divi-miseratus
Cárd. ex Backeb., Natl. Cact. Succ. J. 6: 9, fig.,
(March 1951) nom. inval. [Art. 35.1]; Die
Cactaceae 2: 972, 973, (Mar 1959). Type: Fig. 892
of a flowering plant on p.972 of the protologue).
Bolivicereus samaipatanus ssp. divi-miseratus
(Cárd.) Guiggi, Cactology 3: 6, (2012) nom.
inval. (Art. 41.5). Basionym incorrectly cited
as: B. samaipatanus var. divi-miseratus Cárd.
Borzicactus samaipatanus var. divi-miseratus
(Cárd. ex Backeb.) Krainz, Kakteen, Lief. 21:
C Vb. (1962). Basionym: Bolivicereus
samaipatanus var. divi-miseratus Cárd. ex
Etymology: referring to the place of discovery of
the plant, the Serrania Divi Miserato, a range of
hills northeast of Roboré, Bolivia.
Habitat & Distribution: Arid rocky outcrops and
hills amongst shrubs in the mountains east of
Roboré, 600–900m.
Ecological regions: Chiquitano Dry Forest.
Occurrence: BO (SC). Map 4.
Long considered as just a disjunct variant this
form has become adapted to a very different
environment than that of the Andean foorhills. It
172 Bradleya 34/2016
grows erect rather than decumbent, has more ,
generally hyaline spines and longer flowers which
lack the paler edge to the tepals.
Cárdenas' original description of the plant as a
variety has to be considered invalid since it was
published two months prior to the description of
the genus and type subspecies. The name remained
so until validated by Backeberg in 1959. Cárdenas
had intended the type to be a plant he collected in
February 1950 at 900m in the Sierrania Divi
Miserato of Province Chiquitos, Dept. Santa Cuz,
Bolivia and subsequently preserved in his
herbarium as Cárdenas 4569 (LIL 531545).
19. Cleistocactus smaragdiflorus (F.A.C.
Weber) Britt. & Rose, Cact. 2: 174 (1920).
Basionym: Cereus colubrinus var. smaragdiflorus
F.A.C. Weber.
Cereus colubrinus var. smaragdiflorus F.A.C.
Weber, Dict. hort. Bois p.281 (1894). Type:
from cultivated material, not designated.
Neotype, selected by Hunt & Taylor, Cact.
Syst. Init. 21: 6 (2006): Gürke, Bluh. Kakt. 2: t.
87 (1906).
Figure 34. Cleistocactus samaipatanus ssp. divi-miseratus Bercht 3781, Aguas Calientes, Santa Cruz, Bo-
livia, 253m. Photograph: Ludwig Bercht.
Figure 35. Cleistocactus samaipatanus ssp. divi-
miseratus Bercht 3781, Aguas Calientes, Santa
Cruz, Bolivia, 253m.Photograph: Ludwig Bercht.
Bradleya 34/2016
C. smaragdiflorus (F.A.C. Weber) Speg., Anal.
Mus. Buenos Aires xi: 482 (1905). Basionym:
Cereus baumannii var. smaragdiflora (sic)
F.A.C. Weber.
Cleistocactus azerensis Cárd., Cact. Succ. J. (US)
33: 74, figs., (1961). Type: Puente Pacheco,
Prov. Azero, Dept. Chuquisaca, Bolivia,
1100m, April 1959, Cárdenas 5522 (LIL,
531534, holo.; US). Syn. nov.
C. ferrarii Kiesling, Hickenia 2: 37, fig. (1984).
Type: Cerro de la Cruz, Sierras de Zapla,
Capital, Jujuy, Argentina, April 1974, Kiesling
1722 (SI).
C. parapetiensis Cárd., Cact. Succ. J. (US) 24:
182–183, figs., (1952). Type: on the road from
Camiri to Charagua, Prov. Cordillera, Dept.
Santa Cruz, Bolivia, 900m, March 1949,
Cárdenas 4821 (LIL, not found; US 2906758,
C. rojoi Cárd., Cact. Succ. J. (US) 28: 56, figs.
(1956). Type: between Entre Rios and Villa
Montes, Prov. O’Connor, Dept. Tarija, Bolivia,
600m, March 1952, Cárdenas 5052 (LIL,
531576, holo.; US, iso.).
C. smaragdiflorus var. flavispinus Borg, Cacti
p.151, (1937). Type: not designated.
C. smaragdiflorus var. gracilior Backeb.,
Kakteenlexikon p.89 (1966), nom. inval.. Type:
not designated.
C. smaragdiflorus fma. rojoi (Cárd.) Ritter, Kakt.
damer. 2: 678 (1980). Basionym: C. rojoi
C. villamontesii Cárd., Cact. Succ. J. (US) 33: 76–
77, figs., (1961). Type: Angosto de Villa
Montes, Prov. Gran Chaco, Dept. Tarija,
Bolivia, 460m, May 1958, Cárdenas 5524 (LIL
531598, holo.; US, iso.).
C. villamontesii var. longiflorior Backeb.,
Kakteenlexikon p.90 (1966), nom. inval. Type:
not designated.
A small shrub with erect, arching or
decumbent stems to 1m long. Stems 2–3cm
diameter with 12–14 ribs bearing brown-felted
areoles. Spines differentiated, radials 10–14, up
to 1cm long; centrals 4–6, 15–35mm long,
stringer, yellowish or brown. Flowers 4–5cm ×
±8mm, straight, hypanthium red with scales and
white bristles, perianth unexpanded, outer tepals
yellow, inner tepals vivid emerald green. Fruit
globose, ±1.5cm diameter (Figures 36 & 37).
Etymology: from the Latin smaragdus meaning
“emerald” and referring to the bright green tepals.
Figure 36. Cleistocactus smaragdiflorus
LM0523.01 east of Entre Rios, Bolivia, 1180m.
The form previously known as C. rojoi.
Photograph: Martin Lowry.
Figure 37. Cleistocactus smaragdiflorus
LM0243.03 south of Alemania, Argentina, 1300m.
Photograph: Martin Lowry.
174 Bradleya 34/2016
Habitat & Distribution: Amongst trees and
bushes in the seasonally dry hills and valleys in
the eastern foothills of the Andes, 500–1500m.
Ecological regions: Southern Andean Yungas and
Dry Chaco.
Occurrence: AR (CT, JY?, SA, TM); BO (CH, SC,
TR). Map 8.
Long considerd to be an Argentinian endemic
but now taken to include several forms (C.
azerensis, parapetiensis, rojoi and villamontesii)
in Bolivia described by Cárdenas. These forms
differ only marginally from the type in that their
stems tend to be thinner and their central spines
shorter. Their flowers are more or less identical to
the type form with the exception that the green
inner tepals may be shorter and fewer giving
them a slightly less contrasting colouration.
It can be found growing sympatrically with C.
brookeae Cárd. (C. vulpis-cauda Ritter &
Cullmann) in Bolivia and with C. baumannii
(Lem.) Lem. in Argentina and Bolivia.
It is worth noting that Backeberg’s illustration
(Die Cactaceae VI: 3694 Abb. 3354) of what he
thought might be C. parapetiensis Cárd. is in fact
Figure 38. Cleistocactus strausii LM s.n. north of Santa Ana, Bolivia, 2200m.
Photograph: Martin Lowry.
Figure 39. Cleistocactus strausii Preston-
Mafham 189 from east of the Condor Pass, Tarija,
Bolivia in culture. Photograph: Graham Charles.
Bradleya 34/2016
what he later described as Cephalocleistocactus
20. Cleistocactus strausii (Heese) Backeb.,
Kakteen-Freunde 3: 121 (1934). Basionym:
Pilocereus strausii Heese.
Borzicactus straussii (Heese) Berger, Entwickl.
Kakt. 53 (1926). Basionym: Pilocereus strausii
Cephalocereus straussii (Heese) Houghton, Cact.
Succ. J. (US) 1: 170 (1930). Basionym:
Pilocereus strausii Heese.
Cereus straussii (Heese) Vaupel, Monatschr. Kakt.
23: 37 (1913). Basionym: Pilocereus strausii
Demnoza straussii (Heese) Frič, ller’s Deutsche
Gärtn.-Zeit. 44: 170. (1929). Basionym:
Pilocereus strausii Heese.
Echinopsis nothostrausii Anceschi & Magli,
Cactusinhabitat South America 2011/2013
p.39 (2013). Replaced synonym: Pilocereus
strausii Heese. (Nom. illeg. [Art. 52.1]).
Pilocereus strausii Heese, Gartenflora 56: 410, fig.
49, (1907). Type: from cultivated material,
hort. Berlin, Bolivia, not preserved. Lectotype,
selected by Hunt & Taylor, Cact. Syst. Init. 21:
6 (2006): the fig. cited.
A few-stemmed shrub 1–3m tall branching
from the base. Stems stout, 4–8cm diameter with
25–30 ribs bearing close-set areoles. Spines
differentiated, centrals, about 4, to 2cm, acicular,
pale yellow; radials, 30–40, 15–20mm, very fine
like hairs, pure white. Flowers from upper third of
stem, horizontal, 8–9cm × 7–9mm, dark red-
purple, hypanthium covered with silky brown
hairs, perianth unexpanded with long-exserted
style. Fruit globose, about 2cm diameter, reddish-
brown with brown hairs (Figures 38 & 39).
Etymology: named to honour Leopold Straus, a
German merchant from Bruchsal, cactus hobbyist
and co-founder of the German Cactus Society
Habitat & Distribution: Steep cliffs and mountain
sides of the seasonally dry valleys of central
Province Tarija, 1700–2700m.
Ecological regions: Central Andean Puna and
margins of Southern Andean Yungas.
Occurrence: BO (TR). Map 5.
Probably the most easily recognised species of
the genus although it spent time in several genera
before its recognition as a Cleistocactus by
Backeberg in 1934.
In habitat it can occur sympatrically with both
C. micropetalus and C. hyalacanthus, either
singly or together, with no hybrids between them.
This clearly demonstrates their adaptation to
pollination by distinct species of hummingbird.
Like Darwin’s prediction for the orchid
Angraecum sesquipedale Thouars one could
postulate that there is a species of hummingbird
(rather than hawkmoth) with a beak sufficiently
long to reach the nectaries of C. strausii, which
has the longest flower of the three cacti, yet
cannot enter the curved flowers of C.
hyalacanthus nor the tightly closed flowers of C.
micropetalus. Presumably there are two other
hummingbird species with beak characteristics
allowing them access to the nectar in these other
Cleistocactus strausii has significant horticul-
tural value and is often grown outdoors in warm
climates where it can produce huge specimens. It
has also been hybridized with species from other
genera, notably Echinopsis to produce flowers
that open more widely but retain the colour of C.
21. Cleistocactus tominensis (Weingart)
Backeb., Kaktus-ABC p.190 (1936). Basionym:
Cereus tominensis Weingart.
Borzicactus tominensis (Weingart) Borg, Cacti
p.147 (1937). Basionym: Cereus tominensis
Cereus tominensis Weingart, Monatschr. Deutsch.
Kakteen-Ges. 3: 117–118 (1931). Type: dry veg-
etation in the valley of Tomina, Prov. Tomina,
Dept. Chuquisaca, 1800–2000 m, 1926–28,
Troll 174, not preserved.
C. mendozae Cárd., Cact. Succ. J. (US) 35: 202,
figs., (1963). Type: between Tomina and
Padilla, Prov. Tomina, Dept. Chuquisaca, Bo-
livia, 2200m, March 1958, Cárdenas 5550,
(LIL, not found; US, not recorded). Lectotype
(designated here): the photograph with the
Echinopsis tominensis (Weigart) Anceschi &
Magli, Cactusinhabitat South America
2011/2013 p.40 (2013). Basionym: Cereus tomi-
nensis Weingart.
An erect shrub to 1–2m tall branching from
the base. Stems ±5 cm diameter with 13–22 low,
transversely-furrowed ribs. Areoles brown-felted,
about 1 cm apart bearing 1–20 acicular spines;
centrals 1–3, 10–15mm long, yellow; radials 0–20,
shorter and finer. Flowers produced in profusion
from upper part of stem, straight, 2.5–3cm × 7–
9mm, hypanthium at first green becoming red at
anthesis, perianth unexpanded, tepals green.
176 Bradleya 34/2016
Figure 40. Cleistocactus tominensis LM0963A.01
south of Rendencion Pampa, Bolivia, 2260m. The
form previously known as C. mendozae.
Photograph: Martin Lowry.
Figure 41. Cleistocactus tominensis
BCMdV200.01 from north of Nuevo Mundo, Bo-
livia, 1250m in culture.
Photograph: Martin Lowry.
Figure 42. Cleistocactus variispinus LM0197.01
from north of Sorata, Bolivia, 3060m in culture.
Photograph: Martin Lowry.
Figure 43. Cleistocactus viridiflorus LM0444.01
near Acupata, Bolivia, 3140m.
Photograph: Martin Lowry.
Bradleya 34/2016
Style slightly exserted. Fruit globose, about 15mm
diameter, pale pink–brown (Figures 40 & 41).
Etymology: referring to the place of first discovery
of the species, Tomina, a small town in the eastern
part of Department Chuquisaca, Bolivia.
Habitat & Distribution: Deciduous shrub lands
and forest of the seasonally dry inter-Andean val-
leys along the Río Grande drainage, 1000–3100m.
Ecological regions: Bolivian Montane Dry Forest
and margins of Southern Andean Yungas.
Occurrence: BO (CB, CH, PO, SC). Map 4.
In the absence of a preserved type for this
species it is unfortunate that the illustration in
the protologue is of a small seedling that is not at
all representative of the mature plant.
The most striking character of this species is
the colour-changing flower. The buds start bright
green and remain so until anthesis at which time,
presumably to indicate readiness for pollination,
the hypanthium turns bright red. The effect is
particularly marked in the form Cárdenas called
C. mendozae and in the flowering season these
plants can be seen with many flowers at all stages
of development giving the impression of multi-
coloured flowers as noted in Weingart’s first de-
The species is quite variable with plants from
the bottom of the Río Grande valley having far
fewer, sometimes only one stem and often a com-
plete absence of radial spines (Figure 41). The
flowers of this form are, however, similar to the
type and behave in an identical manner. The de-
scription given above is amplified from the origi-
nal to reflect these differences.
22. Cleistocactus variispinus Ritter, Taxon 13:
114 (1964); Kakt. Südamer. 2: 689–691, figs.,
(1980). Type: Mina Cascabel, Prov. Lareca, Dept.
La Paz, Bolivia, 2000m, 1953, Ritter 108 (U
Cephalocleistocactus schattatianus Backeb.,
Descr. Cact. Nov. III: 5 (1963), nom. inval. [Art
A many-stemmed erect or decumbent shrub
freely branching from the base and lower half.
Stems 0.5–1 (-2)m × 3–5cm diameter with 14–18
ribs bearing areoles only 3–4mm apart. Spines
30–40, pale yellow to brownish-yellow, most <
10mm but occasionally 1 or 2 longer, 2–4cm. Flow-
ers from upper part of stem from areoles produc-
ing additional bristles and spines, 3.5–4.5cm ×
6–8 mm, hypanthium crimson red–orange, tepals
white with pale green tips, spreading slightly,
style not exserted. Fruit globose, 1.5–2cm diame-
ter, violet-red (Figure 42).
Etymology: from the Latin varius and spinus re-
ferring to the spines being of variable length. The
apparent correction used in the NCL is copied from
Eggli et al. (1995) where Ritter’s spelling was, in
error, considered a misprint (ICN Rec. 60G.1).
Habitat & Distribution: Valleys and slopes of
humid mountain forests above the Yungas, 3000
Ecological regions: Central Andean Wet Puna and
Bolivian Montane Dry Forest.
Occurrence: BO (LP). Map 9.
One of several species which produces addi-
tional long spines and bristles on areoles in the
flowering zone of the stem to produce a pseudo-
cephalium. The production of this pseudo-cephal-
ium was the reason why Backeberg either
transferred to, or described new names under Rit-
ter’s genus Cephalocleistocactus. Ritter
adamantly disapproved of these names since for
him the pseudocephalia of these plants were not of
the same form as those produced by his Cephalo-
cleistocactus chrysocephalus.
Ritter’s illustration (Kakt. Südamer.2: Abb.
677) shows a large sprawling mass of stems on a
hillside apparently bare of trees. The stems ap-
pear to be quite short and thick, unlike those of
plants currently in cultivation under this name
which are similar to the plant described by Backe-
berg as Cephalocleistocactus schattatianus. Ritter
also quotes the habitat altitude as 2000m which is
considerably below that at which I have seen
plants similar to those in cultivation. Ritter also
goes to great length to argue that Backeberg’s C.
schattatianus may not be the same as C. variispi-
nus. Thus there is still some confusion over the
correct application of this name that would bene-
fit from a careful exploration of the area around
the type locality.
23. Cleistocactus viridiflorus Backeb., Descr.
Cact. Nov.III: 5 (1963). Type: a plant of Ritter 323
in the collection of Saint-Pie. Ritter gives details
for this number as: Ayata, Prov. Muñecas, Dept.
La Paz, Bolivia, 3300m, Oct 1954., Lectotype, des-
ignated by Backeberg: Kakteenlexicon p.90, fig. 61
(left) (1966).
Cleistocactus palhuayensis Ritter & Shahori,
Kakt. Südamer. 2: 688, figs., (1980). Type:
lower Río Palhuaya, Prov. Muñecas, Dept. La
Paz, Bolivia, 1954, Ritter 324 (U, not found;
ZSS T4911 & T4912, iso.)
178 Bradleya 34/2016
C. palhuayensis var. camachoensis Ritter & Sha-
hori, Kakt. Südamer. 2: 688–689, figs. (1980).
Type: upper Río Palhuaya, Prov. Camacho,
Dept. La Paz, Bolivia, 3200m, Ritter s.n. (U,
not found). Note: Eggli (Englera 16: 401
(1995)) gives Ritter 835 as the probable type
number for this taxon, however, the location
details are different from those given in the
protologue. Also Ritter clearly states there
that this taxon has no number, however he
later (Kakt. Südamer. 4: 1637) assigns this
Cephalocleistocactus pallidus Backeb., Kakteen-
lexicon p.79, fig. (1966). Type: Palhuaya gorge,
Consata basin, Dept. La Paz, Bolivia, Ritter
An erect shrub, branching from the base.
Stems 1–1.6m × 2–4cm diameter with about 12–
17 low ribs. Areoles with tawny felt and 15–20 yel-
lowish spines, radials 15–18, < 1cm long, thin;
centrals 1 or 2, to 2cm long, thicker. Flowers al-
most apical, about 3.5 cm × 6–8mm diameter from
areoles producing additional longer spines, hy-
panthium with short brown hairs, basally reddish,
upper half pale yellow–green, tepals pale yellow–
green, barely expanded. Style not exserted. Fruit
globose, about 1.5cm diameter, reddish with
brown hairs (Figure 43).
Etymology: from the Latin viridus and flos;
Habitat & Distribution: Amongst trees and shrubs
on slopes of humid valleys in the northern Yun-
gas, 1400–3200m.
Ecological regions: Bolivian Yungas and Bolivian
Montane Dry Forest.
Occurrence: BO (LP). Map 9.
Previously referred, in the NCL, to C.
buchtienii Backeb., this species is not only highly
disjunct but grows in a very different environ-
mental niche and has much shorter green rather
than purple flowers.
This species has several characters in common
with the previous taxon which grows in a similar
environment nearby and they may only be ecolog-
ical variants of one species. They differ primarily
in growth form and branching pattern. Cleisto-
cactus variispinus regularly branches mesotoni-
cally and can eventually form a small tree
whereas this habit has not been seen in the cur-
rent species. I hesitate to synonymise them here
since I have not observed plants of this species or
its synonyms at their respective type localities.
In his 1956 catalogue (Winter, 1956) Ritter
claims that this species grows at the highest ele-
vation of all cleistocacti, where in winter the tem-
perature sinks below freezing at night.
24. Cleistocactus winteri D. Hunt, Bradleya 6:
100 (1988). Replaced synonym: Winteria au-
reispina Ritter.
Borzicactus auriespinus (Ritter) Rowley, Repert.
Pl. Succ. XXIV - 1973: 6 (1975), nom. inval.
[Art. 41.5].
Cleistocactus auriespinus (Ritter) D. Hunt (non C.
auriespinus Frič), Bradleya 5: 92 (1987), nom.
illeg. [Art 41].
Hildewintera auriespina (Ritter) Ritter, Kakt.
and. Sukk. 17: 11 (1966), nom. inval. [Art.
35.1, 41.5]. Intended replaced synonym: Win-
teria auriespina Ritter.
Hildewintera auriespina (Ritter) Ritter ex G.D.
Rowley, Repert. Pl. Succ. XVII in Regnum Veg.
54: 15. (1968), nom. illeg. [Art 52.1].
Loxanthocereus auriespinus (Ritter) Buxbaum,
Kakteen (H. Krainz) 57: CVb (1974), nom.
inval. [Art. 41.5].
Winteria auriespina Ritter, Kakt. and. Sukk.13:
4–8, figs. (1962), nom. incorr. [Art 11.4]. Type:
Yapacani gorge on overhanging rocks, Agua
Clara, Prov. Florida, Dept. Santa Cruz, Bo-
livia, July 1958, Ritter 846 (U 097816B).
Winterocereus auriespinus (Ritter) Backeb.,
Kakteenlexikon p.455 (1966).
A procumbent or pendant shrub with stems
branching from the base. Stems 1–2m × 3–7cm
with 14–19 ribs bearing closely set areoles. Spines
15–30, variable, golden yellow or white, 5–50mm
long. Flowers 4–6 × 5cm diameter, orange or red,
perianth expanded and nearly regular. Outer
tepals linear or spathulate, 2–3.5cm × 3–5mm,
spreading; inner tepals spathulate, white–pale-
pink, 3–8mm × 3–5mm, incurved against the
filaments. Stamens and style exserted, filaments
concolourous with outer tepals, anthers orange or
magenta. Fruit globose, 7–10mm diameter,
reddish green.
Habitat & Distribution: Sandstone cliffs in semi-
deciduous humid forest of the north-eastern
Andean valleys, 1300–1400m.
Ecological regions: Bolivian Montane Dry Forest.
Occurrence: BO (SC). Map 2.
The nomenclatural history of this species is
highly confusing and even propagated into
confusion over the names of both subspecies.
Unravelling the confusion took several attempts
over a number of years and I do not propose to
revisit the issues here simply referring the
Bradleya 34/2016
Figure 44. Cleistocactus winteri ssp. winteri
LM0021.01 at its only locality west of Mairana,
Bolivia, 1430m. Photograph: Martin Lowry.
Figure 46. Cleistocactus winteri ssp. colademono
LM0587.03 on Cerro El Fraile, Dept. Santa Cruz,
Bolivia, 1350m. Photograph: Martin Lowry.
Figure 47. Cleistocactus winteri subsp. colade-
mono Krahn 950 ( Type collection) from Cerro El
Fraile, Dept. Santa Cruz, Bolivia, in culture.
Photograph: Graham Charles.
Figure 45. Cleistocactus winteri ssp. winteri
LM0021.01 from west of Mairana, Bolivia, 1430m
in culture. Photograph: Martin Lowry.
180 Bradleya 34/2016
interested reader to two excellent papers by
Kiesling & Metzing (2004) and Metzing &
Kiesling (2007). As these authors say, the
confusion was not caused by any neglect of the
rules and recommendations “but rather, it shows
the complexity of the Code ... and its many
pitfalls”. Essentially, strict application of the Code
(McNeill et al., 2012) does not allow retention of
the generic name Hildewintera since Ritter did
not explicitly state the page number of his earlier
description of the genus Winteria. Thus, if one
wishes to consider these plants as distinct from
Cleistocactus, the earliest validly published
generic name is Winterocereus Backeb., which is
where Metzing and Kiesling place the two
subspecies as separate species. The taxa are
retained here in Cleistocactus based on their
many shared features including stem, flower, fruit
and seed characteristics, and pending further
molecular studies on the circumscription of the
genus. Their few differing floral traits, although
considerable, are taken to be derived from the
basic floral structure of Cleistocactus.
Key to subspecies of C. winteri
(Adapted from Kiesling & Metzing, 2004)
1. Stems almost rigid, erect when young,
pendulous later, spines golden yellow,
hypanthium curved without axial spur, nectar
chamber closed by diaphragm, outer tepals orange
(Figures 44 & 45) a. ssp. winteri
Stems flexible, always pendulous, spines like
hair, white, hypanthium geniculate with axial
spur, no diaphragm, outer tepals red (Figures 46
& 47) b. ssp. colademono
a. ssp. winteri
Etymology: apparently referring to the husband
of Frau Hildegard Winter from whose nursery
near Frankfurt she distributed the seeds Ritter
collected in South America (Mottram, 2006).
Nomenclatural issues aside, C. winteri is a
remarkable plant showing features highly
adapted to growth on cliffs in an occasionally very
wet habitat where for six consecutive months the
rainfall can exceed 80mm per month. The long
downward pointed spines allow this rainfall and
run-off to quickly drain from the plant. Further
adaptations are evident in the floral structure
where the reduced and in-curved inner tepals
protect the nectar chamber from unwanted insect
visitors. In addition, in the current subspecies, the
nectar chamber is further protected by a
diaphragm created by the connate stamen bases,
a feature also present in C. brookeae.
Even after nearly sixty years the taxon is still
only known from a single location where the old
road from Santa Cruz to Cochabamba runs
alongside the Río Yapacani north of Pampa
Grande, an area of about 0.06km2. I have been
observing the plants there for 20 years and
although the road has been much developed in
that time there are still many specimens visible
on the rocks.
b. ssp. colademono D. Hunt, Cact. Syst. Init. 20:
14 (2005). Replaced synonym: Hildewintera
colademononis Diers & Krahn.
Hildewintera colademononis Diers & Krahn,
Kakt. and. Sukk. 54: 221, figs (31 July 2003).
Type: Cerro el Fraile, Prov. Florida, Dept.
Santa Cruz, Bolivia, 2000; Krahn 950
(KOELN, holo.).
H. polonica E. Foik & V. Foik, Cactaceae etc.
(Bratislava) 13: 68–70 (15 July 2003), nom.
inval. [Art. 40.7].
H. polonica E. Foik & V. Foik, Cactaceae etc.
(Bratislava) 14: 43–50, (2004), nom. inval.
[Art. 38.13 & 41.5].
Borzicactus colademononis (Diers & Krahn) Lodé,
Cact.-Avent. Int. 97: 2 (2013, published 28-12-
2012). Basionym: Hildewintera colademononis
Diers & Krahn.
Cleistocactus colademononis (Diers & Krahn)
Mottram, CactusWorld 24: 21 (2006).
Basionym: Hildewintera colademononis Diers
& Krahn.
Winterocereus colademononis (Diers & Krahn)
Metzing & R. Kiesling, Taxon 56: 227 (2007).
Basionym: Hildewintera colademononis Diers
& Krahn.
Etymology: the local vernacular name of the plant
around Samaipata, deriving from Spanish and
meaning “tail of the monkey”, an allusion to its
many soft white spines.
The discovery of this plant in 2000–2003
created a sensation in cactus horticulture and the
plant has rapidly spread through collections
around the world. Its flowers are perhaps its most
appealing attribute. They have characters
combining those of both C. winteri ssp. winteri
and of C. brookeae, leading to the suggestion that
the plant might be of hybrid origin (Kiesling &
Metzing, 2004; Mottram, 2006; Charles, 2011).
The stems also resemble those of the former C.
vulpis-cauda Ritter & Cullmann (now considered
synonymous with C. brookeae Cárd.), adding
further weight to this suggestion. However,
Bradleya 34/2016
considering the features of all three taxa it is
potentially possible that C. winteri ssp.
colademono could be a contributor rather than a
recipient of genetic information and that C.
winteri ssp. winteri is actually the hybrid.
One of the main characters which led Kiesling
and Metzing (2004) to consider the current
subspecies as a separate species is its possession
of an axial spur on the hypanthium. They
considered this structure, sometimes associated
with sphingophily, to be, in this case, an
adaptation to ornithophily allowing humming-
birds to obtain nectar without damaging the lower
gynoecium. Although this structure is clearly
evident in the figures of their paper and they
considered it unique, careful examination of the
sectioned geniculate flower of C. wendlandiorum
illustrated by Backeberg show an identical
structure, more clearly visualized in an image by
Charles (2011). Externally the flowers of C.
vulpis-cauda Ritter & Cullmann show a
pronounced bulge at the knee and a section
reveals that this taxon too possess a rudimentary
axial spur. These observations give considerably
more weight to the argument that the two
subspecies of C. winteri D. Hunt should remain
within Cleistocactus.
Undescribed species
Although the area of distribution of the genus
is relatively well explored there are still some
areas that would benefit from further exploration.
In the past this has been difficult due to limited
access and poor road infrastructure, a situation
which is improving rapidly, particularly in
Bolivia. One particular area which has already
revealed an interesting, potentially new species,
is the Bolivian Yungas. Plants discovered in this
area have now matured in cultivation and
produced bright orange flowers (Figure 48).
Considering the disjunction in distribution of the
genus between the Peruvian and Bolivian species
it would interesting to explore this region in
greater detail. Further tantalizing photographic
images taken north-east of Independencia, north-
east of Vallegrande and east of Pulquina also
show plants which are difficult to place within any
of the species recognised here. It seems then that
we don’t yet know the complete picture for this
fascinating genus.
Distribution Maps
Figure 48. Cleistocactus sp. An undescribed
species from the Yungas east of La Paz, Bolivia.
Photograph: Graham Charles.
Map 1. (Central South America). Occurrence lo-
cations for Cleistocactus baumannii (red dots) and
its subspecies horstii (green dot). In this, and sub-
sequent maps, shading represents elevation in
meters. Cities with representative elevations for
this map are: La Paz: 3650m, Santa Cruz: 420m,
Filadelphia: 140m, Asuncion: 60m, Cordoba:
400m and Salta: 1200m.
182 Bradleya 34/2016
Map 2.(Eastern Bolivian Andes). Occurrence loca-
tions for Cleistocactus brookeae (yellow dots) and
the two subspecies of C. winteri: ssp. winteri (green
dot), ssp. colademono (red dot). Representative el-
evations are Cochabamba: 2570m, Santa Cruz:
420m and Tarija: 1900m.
Map 3. (South-western Bolivia). Occurrence loca-
tions for Cleistocactus buchtienii (northern, type
form: red dots, southern, “tupizensis” form: brown
dots), C. candelilla (magenta dots) and C. mi-
cropetalus (green dots). Representative elevations
are Cochabamba: 2570m, Santa Cruz: 420m,
Sucre: 2820m and Tarija: 1900m.
Map 4. (South-eastern Bolivia). Occurrence loca-
tions for Cleistocactus capadalensis (red dots), C.
tominensis (yellow dots), C. samaipatanus (green
dots) and the approximated type locality of its sub-
species divi-miseratus (purple dot). Representative
elevations are Cochabamba: 2570m, Santa Cruz:
420m, Sucre: 2820m and Tarija: 1900m.
Map 5. (Southern Bolivia). Occurrence locations
for Cleistocactus parviflorus (green dots), C. de-
pendens (red dots), C. hildagardiae (yellow dots)
and C. strausii (purple dots). Representative ele-
vations are Cochabamba: 2570m, Santa Cruz:
420m, Sucre: 2820m and Tarija: 1900m.
Bradleya 34/2016
Map 6. (Central west Bolivia). Occurrence loca-
tions for Cleistocactus laniceps (red dots), C. lurib-
ayensis (green dots) and C. reae (purple dots).
Representative elevations are La Paz: 3660m and
Cochabamba: 2560m.
Map 7. (Central Peru). Occurrence locations for
Cleistocactus pungens (red dots) and C. morawet-
zianus (green dots). Representative elevations are
Huancavelica: 3720m, Ayacucho: 2730m and Aban-
cay: 2475m.
Map 8. (Central Andes). Occurrence locations for
Cleistocactus hyalacanthus (red dots) and C.
smaragdiflorus (green dots). Representative eleva-
tions are Sucre: 2820m, Tarija: 1900m, Jujuy:
2175m, Salta: 1200m and Tucuman: 425m.
Map 9. (Bolivian Yungas). Occurrence locations for
Cleistocactus chrysocephalus (red dots), C. viridi-
florus (green dots), C. variispinus (purple dots) and
C. ritteri (yellow dots). Representative elevations
are La Paz: 3660m and Cochabamba: 2560m.
184 Bradleya 34/2016
The preparation of this synopsis would not
have been possible without the help of many
people. In particular I am grateful to Graham
Charles, Paul Hoxey, Mats Winberg and Holger
Kelsch for providing additional occurrence data
and to Graham Charles, Andreas Hofacker,
Zlatko Janeba and Paul Laney for providing
copies of older literature. Ludwig Bercht, Graham
Charles, Holger Kelsch and Paulo Robson de
Souza also provided additional images. I also
thank Roy Mottram and Nigel Taylor for
enlightening feedback on nomenclatural
complexities. It is also a pleasure to thank all
those friends who have travelled with me in the
field and helped make the collection of the data
such a rewarding experience.
Cactusinhabitat South America 2011/2013:
ANDERSON, E.F. (2001). The Cactus Family.
Timber Press. Portland, Oregon.
BACKEBERG, C. (1959). Die Cactaceae. Handbuch
der Kakteenkunde. Band II. Gustav Fischer.
BACKEBERG, C. (1966). Das Kakteenlexikon -
Enumeratio Diagnostica Cactacearum. Gustav
Fischer. Jena.
BRITTON, N.L. AND ROSE, J.N. (1920). The
Cactaceae. Descriptions and Illustrations of
Plants of the Cactus Family. Volume II.
Carnegie Institute. Washington.
BUXBAUM, F. (1956). La Division du Genre
Cleistocactus. Cactus (Paris) 51: 87–91.
BUXBAUM, F. (1958). The phylogenetic division of
the subfamily Cereoideae, Cactaceae.
Madrono 14: 177–216.
CHARLES, G.J. (2011). In the Glasshouse. The
Cactus Explorer 1: 13–15.
Eggli, U., Schick, M.M. & Leuenberger, B.E.
(1995). Cactaceae of South America: The
Ritter Collections. Englera 16.
EGGLI, U. & NEWTON, L.E. (2004). Etymological
Dictionary of Succulent Plant Names.
Springer. Berlin & Heidelberg.
EGGLI, U. & LEUENBERGER, B.E. (2005). The
Cárdenas Type Specimens of Cactaceae
Names in the Herbarium of the Instituto
Miguel Lillo, Tucumán, Argentina (LIL).
Willdenowia 35: 179–192.
EGGLI, U. (2005). Das grosse Kakteen-Lexikon.
Ulmer. Stuttgart.
FOIK, E. & FOIK, V. (2003). The discovery of a new
Hildewintera. Cact. Succ. J. (Los Angeles) 75:
MAGALLÓN, S. (2011). Phylogenetic
relationships and evolution of growth form in
Cactaceae (Caryophyllales, Eudicotyledoneae).
Am. J. Bot. 98: 44–61.
HUNT, D. & TAYLOR, N.P. (1987). New and
unfamiliar names of Cactaceae to be used in
the European Garden Flora. Bradleya 5: 91–
HUNT, D. (1999). CITES Cactaceae Checklist. RBG
Kew. London.
HUNT, D. (2003). Notulae Systematicae Lexicon
Cactacearum Spectantes III: Cleistocactus.
Cact. Syst. Init. 15: 7–8.
HUNT, D., TAYLOR N.P. & CHARLES, G. (2006). The
New Cactus Lexicon. dh Books. Milborne Port.
HUNT, D. (2013). The New Cactus Lexicon
Illustrations. dh Books. Milborne Port.
IPNI. (2012). The International Plant Names
Index. Published on the Internet [last accessed 5th August
IOS WORKING PARTY. (1986). The genera of the
Cactaceae: towards a new consensus. Bradleya
4: 65–78.
Kelsch, H. (2016). Following Friedrich Ritter’s
Footsteps in Inquisivi, Bolivia. The Cactus
Explorer 15: 31–36.
KIESLING, R. & METZING, D. (2004). An amazing
species of Hildewinteria (Cactaceae) -
characters and systematic position. Cact. Succ.
J. (Los Angeles) 76: 4–12.
LEMAIRE, C. (1844). Nouveautés: Cereus
Baumanni. Hort. Univ. 5: 126–127, 315. H.
Cousin, Paris.
LEUENBERGER, B.E. (2002). The misunderstood
Cleistocactus serpens, and C. longiserpens, a
new name for a species of Cactaceae from
Peru. Bot. Jahrb. Syst. 124: 13–30.
LEUENBERGER, B.E. (2012). The misnamed
Cleistocactus of Tupiza. Cact. Syst. Init. 27:
LODÉ, J. (2015). Taxonomy of the Cactaceae.
Editions Cactus-Adventures. Barcelona.
(2012). International Code of Nomenclature
for Algae, Fungi and Plants (Melbourne Code).
Regnum. Veg. 154. (Available online at:
Bradleya 34/2016
METZING, D. & KIESLING, R. (2007) Winterocereus
(Cactaceae) is the correct name for
Hildewintera. Taxon 56: 226–228.
MOTTRAM, R. (1989). New Cleistocactus cultivars.
Cact. Succ. J. (Los Angeles) 61: 154–157.
MOTTRAM, R. (2002). Notulae Systematicae
Lexicon Cactacearum Spectantes I:
Cleistocactus. Cact. Syst. Init. 13: 10–12.
MOTTRAM, R. (2006). Hildewintera a winter’s tail.
Cactus World 24: 19–22.
& KASSEM, K.R. Terrestrial Ecoregions of the
World: A New Map of Life on Earth (2001)
BioScience 51: 933–938.
OSTOLAZA, C. (1996). A closer look at the
conservation status of cacti in the vicinity of
Lima, Peru. Brit. Cact. Succ. J. 14: 158–174.
OSTOLAZA, C. (1998). The cacti of the Pisco, Ica and
Nazca valleys, Peru. Brit. Cact. Succ. J. 16:
S. (2003). Cacti of the Huara river valley, Peru
a conservation study. Brit. Cact. Succ. J. 21:
RITTER, F. (1980). Kakteen in Südamerika. Band 2
& 4. Self published. Spangenberg.
Molecular phylogenetics of Echinopsis
(Cactaceae): Polyphyly at all levels and
convergent evolution of pollination modes and
growth forms. Am. J. Bot. 99: 1335–1349.
SCHUMANN, K. (1903). Gesamtbeschreibung der
Kakteen. Neumann. Neudamm.
WINTER, H. (1956). Cactus seed catalogue.
Index to specific epithets
Numbers and letters refer to the numbered
sequence of taxa in the synopsis. Accepted names
are indicated in bold face.
brevicaulis 18a
croceus 18a
rufus 18a
samaipatanus 18a
samaipatanus ssp. divi-miseratus 18b
samaipatanus var. divi-miseratus 18b
samaipatanus var. multiflorus 18a
aureispinus 24a
aureispinus f. cristatus 24a
colademononis 24b
janae 8
laniceps 10
samaipatanus 18a
samaipatanus var. divi-miseratus 18b
samaipatanus var. multiflorus 18a
strausii 20
tominensis 21
strausii 20
chrysocephalus 6
pallidus 23
ritteri 17
schattatianus 22
anguinus 1a
baumannii 1a
colubrinus 1a
grossei 1a
smaragdiflorus 19
strausii 20
angosturensis 3
anguinus 1a
areolatus 14
areolatus var. herzogianus 14
aureispinus 24
ayopayanus 3
azerensis 19
baumannii 1a
baumannii ssp. anguinus 1a
baumannii ssp. chacoanus 1a
baumannii var. colubrinus 1a
baumannii ssp. croceiflorus 1a
baumannii var. flavispinus 1a
baumannii var. grossei 1a
baumannii ssp. horstii 1b
baumannii var. paraguariensis 1a
baumannii ssp. santacruzensis 1a
brevispinus 15
brookeae 2
brookeae var. flavispinus 2
brookeae ssp. vulpis-cauda 2
bruneispinus 1a
buchtienii 3
buchtienii var. flavispinus 3
candelilla 4
candelilla ssp. piraymirensis 4
candelilla var. pojoensis 4
capadalensis 5
chacoanus 1a
chacoanus var. santacruzensis 1a
chrysocephalus 6
clavicaulis 12
colademononis 24b
colubrinus 1a
colubrinus var. flavispinus 1a
186 Bradleya 34/2016
colubrinus var. grossei 1a
colubrinus var. smaragdiflorus 19
crassicaulis 12
crassicaulis var. paucispinus 12
croceiflorus 1a
dependens 7
ferrarii 19
flavispinus 1a
fusiflorus 14
glaucus 11
glaucus var. plurispinus 11
granjaensis 11
grossei 1a
herzogianus 14
hildegardiae 8
horstii 1b
hyalacanthus 9
hyalacanthus ssp. tarijensis 9
ianthinus 14
jugatiflorus 1a
jujuyensis 9
jujuyensis var. fulvus 9
laniceps 10
laniceps var. plurispinus 10
luribayensis 11
mendozae 21
micropetalus 12
morawetzianus 13
morawetzianus var. pycnacanthus 13
muyurinensis 4
orthogonus 3
palhuayensis 23
palhuayensis var. camachoensis 23
paraguariensis 1a
parapetiensis 19
parviflorus 14
parviflorus var. aiquilensis 14
parviflorus var. comarapanus 14
parviflorus var. herzogianus 14
piraymirensis 4
pojoensis 4
pungens 15
pycnacanthus 13
reae 16
ressinianus 3
ritteri 17
rojoi 19
samaipatanus 18a
samaipatanus ssp. divi-miseratus 18b
santacruzensis 1a
smaragdiflorus 19
smaragdiflorus var. flavispinus 19
smaragdiflorus var. gracilior 19
smaragdiflorus f. rojoi 19
strausii 20
strausii var. fricii 9
strausii var. jujuyensis 9
sucrensis 3
tarijensis 9
tominensis 21
tominensis ssp. micropetalus 12
tupizensis 3
tupizensis var. sucrensis 3
vallegrandensis 4
variispinus 22
villaazuensis 13
villamontesii 19
villamontesii var. longiflorior 19
viridialabastri 12
viridiflorus 23
vulpis-cauda 2
wendlandiorum 2
winteri 24a
winteri ssp. colademono 24b
strausii 20
strausii 20
strausii var. luteispina 9
baumannii 1a
candelilla 4
horstii 1b
nothohyalacantha 9<