Article Index by Title

Paul Hoxey

The first instalment of this article appeared in TSG September 2008 Vol. 14 No. 3 (see part 1). In this second instalment I will continue reviewing names associated with Cumulopuntia sphaerica.  A third and final part (see part 3) will follow to complete the review and I will propose a new species level classification with a full list of synonyms.

Cumulopuntia multiareolata (Ritter) Ritter

This plant was originally described by Ritter as Tephrocactus multiareolatus in 1964 and then subsequently moved into Cumulopuntia by Ritter in 1981.  The type locality is given as Convento in the province of Caravelí in the northern part of the department of Arequipa in habitats on the eastern edge of the desert belt.

The description calls for bluish green segments to 5cm tall and 3.5cm thick with large woolly areoles which extend to the base of the segment.  Spines number 8 to 20, central spines to 2cm long and finer radials to 1cm long.  

I encountered plants close to the town of Caravelí, due east and at similar altitude to Convento, growing in very arid conditions on the edge of the desert belt, just as Ritter describes.  The illustration (Fig. 1, originally shown on the front cover of the journal) shows a plant found in this area with very densely spined segments and with the large white areoles.  Generally the plants were well adorned with spines but not always as extensively as this example. Plants tended to form isolated small clumps on the rocky slopes but larger colonies grow on flat sandy areas, possibly due to the ease at which detached segments root.

Fig. 1

Fig. 1. Cumulopuntia multiareolata. PH760.04 Caraveli, Arequipa, 1860m.

This plant is notable for the high number of spines which have the untidy appearance of the type form from Arequipa.  Tentatively I place it as a synonym of Cumulopuntia sphaerica.

Nasca Valley Taxa

The Nasca valley, which starts west of the town of Nasca and rises to over 4000m into the Andes, is a rich source of Cumulopuntia with no less then three taxa having their type localities within it.  There is possibly a fourth taxon of uncertain status.  Unfortunately I have not travelled the full length of the valley and I have only visited the area around the town of Puquio where I have seen one of the four.  

Tephrocactus mirus Backeberg

The type locality for this plant is given as the Nasca valley at 1200m and with a description which calls for a close relative of Cumulopuntia sphaerica with small segment size, densely spined (up to 18) with abundant wool on young growth.

I have not visited the type locality but I have seen one population (PH576.02) located in the Río Omás valley at 1840m  which matches the description and the accompanying illustration in Rauh (1957:199) quite closely.  I tentatively place it here although it is a considerable distance (~200km) to the north of the type locality.  The plants consist of a sprawling clump of segments in the range 2 to 4cm in size, usually slightly elongated (Fig. 2).  The areoles on new growth are particularly woolly and nearly obscure the body.  New spines are straight, needle-like and brown but fade to grey on older segments.  Fruits are globular and much less spiny than the segments, with a few weak spines on the upper half.

Fig. 2

Fig.2. Tephrocactus mirus.? PH576.02 Rio Omas, Lima, 1840m.

Rauh considered Tephrocactus mirus to be close to Cumulopuntia kuehnrichiana and Ritter places it into synonymy beneath that species.  It is clearly a Cumulopuntia but it has never been formally placed in that genus at the species level, hence my use of the generic name Tephrocactus.  The large woolly areoles and straight needle like spines all point to a closer relationship with the small segmented forms of Cumulopuntia unguispina/Opuntia leucophaea rather than Cumulopuntia sphaerica.

Tephrocactus bicolor (Rauh & Backeberg) Rauh

Originally named in 1957 as a variety of Tephrocactus  fulvicomus which is probably a Cumulopuntia boliviana form, it was upgraded to a good species by Rauh a year later.  It is reported from between Nasca and Puquio at 3400/3500m.  I am unfamiliar with this species in habitat.  Iliff (2002:180) refers it to the Cumulopuntia sphaerica group but with reservations.  Its description calls for a loose clump of segments 3cm long and 2cm think, deep green but often reddish. Areoles small 3mm in diameter with yellow glochids.  Spines rigidly erect, 6-8, unequal length, the longest ~3cm long.  There is an illustration in Rauh (1958:224) of a dehydrated plant which is reproduced in Backeberg (1958:T29).

Cumulopuntia galerasensis Ritter

This species was described in 1981 by Ritter from material (FR1045) found at Galeras on the Nasca to Puquio road at 3500m.  His description calls for small dense cushions with grey-green segments 3-4cm long and 1.5-2cm thick. They are fairly strongly tuberculate with the areoles situated near the upper end of the tubercles.  Areoles are yellowish 2-5mm in diameter with yellow glochids.  Only areoles on the upper two thirds of the segments have spines.  Spines are straight and upright, orange-brown to brown in colour, 5-12 in number and 15-50mm in length.  At the lower side of the areole they are weaker, backward facing and paler. The locality and description both point to C. galerasensis being a re-description of Tephrocactus bicolor.

I have not visited the locality of Cumulopuntia galerasensis/Tephrocactus bicolor but Martin Lowry (personal communication) has seen this taxa in habitat growing as small densely spined clumps with tuberculate bodies.  The spines are limited to the top half of the segments with plentiful glochids on lower areoles. Red as well as yellow flowering plants were seen. The tuberculate nature of the segments, small areoles and the glochids point to it being related to C. corotilla.   Further study is required to determine the exact relationship of these Galeras plants.

Cumulopuntia sp. (Puquio)

From a small area just west of Lucanas to as far as Puquio, a Cumulopuntia species can be seen which does not fit comfortably with any other Cumulopuntia taxa.  I believe Ritter's concept of Cumulopuntia ignota, is in part, this taxa. His FR560 is reported from two localities; “south of Misti at 3400m” and “road to Lucanas at KM103”.  The first of these localities is possibly C. corotilla (see later discussion) but the second is clearly the Cumulopuntia species which I have seen near Lucanas where it grows with Oreocereus ritteri at 3930m.  I have also seen it at a second locality near Puquio at 3250m.

This plant forms low loose cushions of elongating segments, 2-3cm wide and 5-6 long, which are very slightly tuberculate but on fully hydrated segments this character is all but invisible (Fig. 3). They are slightly club shaped, wider at the upper end with a rounded top (Fig. 4).  They are usually grey/green but sometimes tinged with red. The areoles are small (but a little larger on the upper half of the segment) with a little white wool and distributed evenly over the segment.  The spines are light brown 3 to 8 in number, spreading evenly and limited to areoles on the top half of the segment only.  Lower areoles are naked and generally devoid of glochids but occasionally a few are present.   The fruits are similar to new segments but globular.  Only areoles around the rim have 2-3 spines, the rest are naked. Of two unripe fruits sectioned both contained one seed each.

Fig. 3

Fig. 3. Cumulopuntia sp. (Puquio). PH747.03 East of Lucanas, Ayacucho, 3930m

Fig. 4

Fig. 4. Cumulopuntia sp. (Puquio). PH747.03 East of Lucanas, Ayacucho, 3930m.

Ritter's classification of this plant as Cumulopuntia ignota is understandable as the brief description and sketch by Britton and Rose (1919:99) fits quite well but the segments are described as small, narrow and only 2-3cm long.  The type location for Opuntia ignota is at Pampa de Arrieros yet there is no evidence Ritter visited it and was acquainted with the true Opuntia ignota.  Until recently Pampa de Arrieros was only accessible by train and possibly not easy to reach.  Opuntia ignota as described by Britton and Rose is a different plant which I will discuss in the third instalment of this article.  We should disregard Ritter's interpretation of  the  name which leaves  the  plant at  Puquio without a valid published name.  It is possibly a relative of Cumulopuntia corotilla with which it shares the small areoles but lacks the numerous glochids.

Cumulopuntia corotilla (K.Schumann ex Vaupel) E. F. Anderson

Opuntia corotilla was described by Schumann in 1913 from material collected by Weberbauer in 1902.   Shortly after this publication Britton and Rose (1919:96) reduced it to a synonym under Opuntia sphaerica.  The combination as a good species in Cumulopuntia was made by Anderson (1999:324) although in the new Cactus Lexicon it was left as a synonym under Cumulopuntia sphaerica.

The type locality is given as between Airampal and Pampa de Arrieros at 3200-3400m on the Arequipa to Puno road. Despite the existence of a herbarium sheet which is illustrated in Iliff (188:2002) and accurate locality details it has remained a misunderstood and misapplied name.  Even Weberbauer (1945) himself may have not understood it correctly as he mentions it growing in other places besides the Arequipa to Puno railroad, including in the Chuquibamba valley and further south inland from Tacna.  At Chuquibamba I only found Cumulopuntia crassicylindrica and at Tacna I found the high altitude small-structured Cumulopuntia sphaerica form.  The name corotilla is a Quechua name for small, low growing, opuntioid-like cacti and most likely includes a number of different taxa.   Ritter followed Britton and Rose and considered Opuntia corotilla a synonym of Cumulopuntia sphaerica (Ritter used the name C. berteri for C. sphaerica; a name we should dismiss). (Please see TSG Vol. 3 No. 4 Dec. 1997 P201-3 Ed.) However to add to the confusion, he considered  Backeberg's concept of Cumulopuntia corotilla to be different and placed it under Cumulopuntia ignota. However Backeberg did have the correct plant under the name Cumulopuntia corotilla, although he illustrates a form devoid of central spines, Backeberg (1958:Abb 280 & 283).  Nearly spineless forms are seen in habitat and most likely cultivated material is also going to be less spiny.

The description summarized from Leighton-Boyce & Iliff (1973: 80) and Iliff (2002:188) calls for a small, loosely branched plant to 15cm high. Stem segments globular to longish-ovoid to 6cm long, lightly tuberculate.  Areoles fairly distant, round, up to 3mm across, somewhat woolly, with numerous erect glochids up to 3mm long, and 0-7 unequal spines on the upper half of the segments of which the  longest is sometimes over 3cm long.  The spination can be sporadic and sometimes entirely absent.  Flowers are at first cream coloured, later rose. Fruits have glochids and a few weak spines on the rim.  

As Iliff states it is quite a distinctive plant and the description has a number of features which immediately distinguishes it from Cumulopuntia sphaerica, including tuberculate segments, small widely spaced areoles, large numbers of glochids and white/rose flowers (all C. sphaerica taxa have consistently yellow flowers which fade to orange/red).  

I encountered this species on hills above the small village of Pampa de Arrieros at 3800m on the road out of Arequipa in February 2008 during the short rainy season.  The plants form low clumps no more than 3 or 4 segments  high.  The  segments are  globular to slightly  elongated to 6cm long, blue/grey in colour, occasionally slightly purplish, clearly tuberculate with numerous glochids. The spination between plants is very variable.  Fig. 5 shows a large fully developed segment with very well developed spination. Spines are straight, spreading and limited to areoles in the top two thirds of the segments whereas areoles below have glochids only. At the other extreme (Fig. 6) plants have segments with only weakly developed spination. The high number of glochids is a more consistent characteristic but the occasional plant was found with a relatively low number. Unfortunately no flowers were observed in February 2008 although developing fruits were present.  They are very similar to young segments but globular 20-25mm wide, with areoles contains glochids and the occasional weak spine around the rim.  

Fig. 5

Fig. 5. Cumulopuntia corotilla. PH710.03 Pampa de Arrieros, Arequipa, 3820m.

Fig. 6

Fig. 6. Cumulopuntia corotilla. PH710.03 Pampa de Arrieros, Arequipa, 3820m.

Further plants were found to the west between Huambo and Huacán, adjacent to the Cañón de Colca, in January 2009.  At this time of year the rains had yet to arrive and the plants looked very different to C. corotilla  with a deep purple body (Fig. 7). This population exhibited a very high number of glochids with variable spination which was often weak.  Ripening fruits contained dry seeds within the pulp-free cavity which is typical for Cumulopuntia (Fig. 8).  On this occasion a plants was excavated and the roots were examined. Surprisingly they are highly tuberous in nature and in excess of 30cm in length (Fig. 9).  The above ground growth is attached through a relatively thin neck. Initially I felt I had found a new distinct species but I wanted to compare it with the plants at the habitat at Pampa de Arrieros. These plants were also found to grow a large tuberous root and exhibited a deep purple colour on shrivelled bodies due to the lack of water. I now believe the Huambo plant is not new and should be included in Cumulopuntia corotilla, extending the range approximately 80km to the West.                                                

Fig. 7

Fig. 7. Cumulopuntia corotilla PH794.03 Between Huambo and Huacan, Arequipa, 3720m.

Fig. 8

Fig. 8. Cumulopuntia corotilla PH794.03 Between Huambo and Huacan, Arequipa, 3720m.

Fig. 9

Fig. 9. Cumulopuntia corotilla PH794.03 Between Huambo and Huacan, Arequipa, 3720m.

Cumulopuntia corotilla is clearly distinct from Cumulopuntia sphaerica (and all other Cumulopuntia) and not a synonym of that species.  It is unusual in growing a tuberous root and having white flowers, both characters usually associated with the genus Maihueniopsis.  I retain it in Cumulopuntia due to the dry fruit cavity and seed structure which more closely fit this genus.  Backeberg's Tephrocactus mistiensis is probably a spineless form of this species.

Opuntia campestris Britton & Rose

This name was erected by Britton and Rose for plants found at Pampa de Arrieros by Dr Rose in 1914. The description calls for a much branched, low forming plant, 30 to 60cm in diameter.  Joints are globular or a little longer than thick, 3 to 5cm long, with numerous prominent areoles and conspicuous turbercles when young.  Numerous yellow glochids.  Spines only on  upper areoles and missing  on lower areoles, the longest to 3.5cm.  Flowers rosy white to light yellow, 2-3cm long.

This is clearly a re-description of Cumulopuntia corotilla with the key characteristics of tuberculate segments, numerous glochids and rosy white flowers. The type location is also consistent with Cumulopuntia corotilla and it should be regarded as a synonym of C. corotilla.

To be continued (see part 3).

This article was originally published in Tephrocactus Study Group (the 'TSG Journal'), June 2009, Vol. 15, No. 2, pages 19-27. © TSG and Paul Hoxey