Article Index by Title

Paul Hoxey

The genus Cumulopuntia was erected by Ritter (1980:399) to include a small number of dwarf South American Opuntioids.  The key characteristic which distinguishes this genus from the closely related Maihueniopsis are the fruits which contain seeds within a dry cavity.  In the New Cactus Lexicon only four species (Cumulopuntia boliviana, chichensis, rossiana and sphaerica) are accepted with a further 24 names referred to these four. No less than 11 are considered synonyms of C. sphaerica.

A trawl through Iliff (2002) reveals a further seven names which are associated with C. sphaerica giving us a potential 18 synonyms.  As Iliff mentions (2002:143) this species group is not well known.  As far as I know no attempt has been made to critically study the group to determine if any of the names should be retained at some sort of botanical rank.   I have seen a number of populations in habitat, many of which correspond to validly published names, and I think it is useful to show the variation in the plants which are encountered under the broad concept of C. sphaerica in the NCL.

Cumulopuntia sphaerica in the broadest sense is a very wide ranging species found on the western side of the Andes in southern Peru and northern Chile.  The northernmost population (C. kuehnrichiana) is from the Rímac valley, near to Lima in the centre of Peru and the southern extreme is reached in the province of Coquimbo, Chile, a distance of over 2000km.  I have observed populations at over 70 localities at altitudes from sea level to 3830m.  The large distribution and altitude range is remarkable for a single species within the Cactaceae. The distribution contrasts strongly with C. boliviana which is a plant of the high Andes to the east and which is found at high elevations in Peru, Chile, Bolivia and Argentina.  The ranges just about overlap in the highlands of southern Peru and I know of one locality near Pampa de Arrieros, on the road from Arequipa to Puno where they grow together.

Not surprisingly for a plant with a wide distribution and so many synonyms there is a large amount of variation between populations.  However I have found variation within populations to be low and the plants generally uniform in characteristics at any given locality.  Perhaps this is due to a significant amount of vegetative propagation from joints which are often weakly attached to each other and readily root to form new plants when detached.

The list below gives the names which are referred to C. sphaerica in Hunt (2006) and Iliff (2002).  Not all have been transferred into Cumulopuntia and I give the earliest published name if that is the case. I also include C. galerasensis as I believe this too has become entangled within this web of names and should be covered in this discussion.

Cumulopuntia alboareolata (Ritter) Ritter
Tephrocactus bicolor (Rauh & Backeberg) Rauh
Opuntia campestris Britton & Rose
Opuntia corotilla K.Schumann ex Vaupel
Cumulopuntia crassicylindrica (Rauh & Backeberg) Ritter
Opuntia dimoropha Foerster
Cumulopuntia galerasensis Ritter
Opuntia ignota Britton & Rose
Cumulopuntia kuehnrichiana (Werdermann & Backeberg) Ritter
Opuntia leucophaea Philippi
Tephrocactus mirus Backeberg
Tephrocactus muellerianus Backeberg
Cumulopuntia multiareolata (Ritter) Ritter
Tephrocactus pseudorauppianus Backeberg
Cumulopuntia rauppiana (Schumann) Ritter
Cumulopuntia tubercularis Ritter
Cumulopuntia tumida Ritter
Cumulopuntia unguispina (Backeberg) Ritter
Cumulopuntia zehnderi (Rauh & Backeberg) Ritter

I will not be discussing further the names Opuntia dimorpha Foerster, Tephrocactus muellerianus Backeberg, Tephrocactus pseudorauppianus Backeberg and Cumulopuntia rauppiana (Schumann) Ritter because none have a well defined type locality or type.  All are accepted to belong to the C. sphaerica group and are best considered synonyms of C. sphaerica.

I highly recommend Iliff's “The Andean Opuntias” published in 2002 as part of the “Studies in the Opuntioideae” as an excellent reference to South American Opuntioids.  I have used it extensively whilst preparing this article, particularly for information on older names in publications which are not readily accessible today.

Cumulopuntia sphaerica (Foerster) Anderson (type form)

Opuntia sphaerica was described in 1861 by Foerster with a location given as the department of Arequipa, Peru.   As no type specimen was preserved Iliff defined the type using the illustration Fig. 113 in Britton and Rose (1919:96), reproduced here (Fig.1), of a plant collected above the town of Arequipa, in the department of the same name.  This is clearly a large growing form with very spiny segments with the spines on all the areoles. The illustration also shows new growth of smaller, elongated and much more weakly spined segments.  I have observed this growth pattern, which is typical for C. sphaerica, in several populations where plants have very green weakly spined segments which are of very recent growth. The segments then appear to undergo a secondary growth phase to reach maturity with stronger spine growth and the toughing of the epidermis with the development of a waxy coating, turning the plant a greyish colour.  Fig. 2 shows clearly the mature and immature growths from a plant in cultivation of ISI 1525, collected 23km west of Arequipa at 2800m. This plant conforms closely to the type.

Fig. 1

Fig. 1. Cumulopuntia sphaerica neotype. Reproduced from Britton & Rose P. 96. Fig. 113.

Fig. 2

Fig. 2. Cumulopuntia sphaerica ISI 1525. Good example of type form. Arequipa, Peru.

I found similar looking plants to the type at Minas Cerro Verde (PH709.04) just outside Arequipa on very dry hills at 2400m, growing as low loose clumps.  Figs. 3 and 4 show a relatively large clone with segments roughly spherical to 8cm in diameter (and suffering from sooty mould). The body is slightly glaucus with a waxy coating. The spines emerge from large areoles and completely cover the segment; they are recurved and somewhat twisted or contorted to give the plant an overall untidy and distinctive appearance in comparison to other C. sphaerica forms. Also growing at this locality we found a smaller growing form, with segments no more than 3cm in diameter, with short straight spines (Fig. 18, near end of article).  I will discuss this plant in more detail in the comments for the small jointed high altitude form of C. sphaerica.

Fig. 3

Fig. 3. Cumulopuntia sphaerica PH709.04. Near Minas Cerro Verde, Arequipa, Peru. 2400m.

Fig. 4

Fig. 4. Cumulopuntia sphaerica PH709.04. Near Minas Cerro Verde, Arequipa, Peru. 2400m.

I have only encountered the type form growing at two other localities; between Moquegua and Omate at 2430m (PH618.04) and at Minas de Toquepala near Moquegua at 2620m (PH615.02) although in both cases with slightly smaller segments. All three localities have a similar altitude and arid environments, being situated on the edge and to the east of the coastal desert belt.  They do not receive any coastal mist and instead rely on very limited rainfall.

Cumulopuntia tumida Ritter

Ritter (1981:1254) erected this name for plants found on the coast near Chala Vieja, in the north part of the department of Arequipa.  Ritter’s original description calls for a loose sprawling plant made up of short-ellipsoid segments 4-8(-10) cm tall and 4-6cm across.  Initially they are green turning to a more grey green with age.  Areoles are 3-5mm in diameter, extending to the base of the segment and orange-brownish in colour. Spines are white to 3cm long, with a brown tip and they taper down from a thick base to a point. Spines are usually absent on the areoles on the lower half of the segment.

I was unaware of anyone other than Ritter finding this plant in habitat and it was with very good fortune we found plants on the coast near the base of the Lomas de Atiquipa, just north of Chala Vieja which correspond very well with the description, see Figs. 5 and 6, my PH584.08.  We only encountered a small number of plants but all were very consistent in their characteristics.  In general characteristics including segment size, this plant most closely resemble the type form of C. sphaerica rather than the smaller growing C. unguispina found to the south along the coast.   The short white spination is slightly recurved but is otherwise very distinct from the type C. sphaerica.

Fig. 5

Fig. 5. Cumulopuntia tumida PH 584.08. The coast north of Atiquipa, Arequipa, Peru.

Fig. 6

Fig. 6. Cumulopuntia tumida PH 584.08. The coast north of Atiquipa, Arequipa, Peru.

This habitat is particularly interesting as the Lomas de Atiquipa is the most favourable locality for plants along the coast of southern Peru because of the amount of moisture available. The extensive lomas vegetation survives on sea mists, which for unknown reasons are at  their strongest along this small section of coast.  The Lomas de Atiquipa is very isolated from other lomas localities as the surrounding area is much more arid.  It is well known as an area with many endemic species of plants (including Eulychnia ritteri and Haageocereus chalaensis within the Cactaceae) and animals because of the isolation.

Cumulopuntia crassicylindricus (Rauh & Backeberg) Ritter

This very distinctive plant is found in a small population in a very arid environment in the base of the Río Majes valley at approximately 1000m in altitude some 80km inland where it shares the habitat with Islaya grandis and Haageocereus pluriflorus. It was originally named as a Tephrocactus by Rauh and Backeberg in Rauh (1957) and was moved into Cumulopuntia by Ritter in 1981.  The plants grow as loose clumps of segments to 10cm or more in length (Fig. 7). They are more elongated than either the type form of C. sphaerica or C. tumida.  The spines are mainly found on the upper half of the segments, a character shared with C. tumida but not C. sphaerica (type form). There is usually one very robust central spine and then 4 or 5 further shorter but equally strong radials. They are always straight. Spine colour varies from grey through to black.  The black colour may be due to growth of sooty moulds and appears to affect the areole wool too (Fig. 8).  When I visited the locality in March 2008, the plants were in full flower with fresh growth. The area was incredibly arid so I imagine the plants had taken advantage of a rare rain event some weeks or months previously to grow new segments.  The new segments are clearly visible in Fig. 7, being bright green, weakly spined and very distinct from the mature growth. Small residual leaves are present. The strong spination (Fig. 7) and grey glaucous body was a consistent feature on the mature growth of all the plants I saw.  Royston Hughes has shown me material purporting to be this plant in cultivation and it is relatively weakly spined with white spines. It is likely in a UK greenhouse we can't give it sufficiently strong sunlight to develop the same degree of spination as is seen in habitat.

Fig. 7

Fig. 7. Cumulopuntia crassicylindrica PH 762.03. Hacienda Ongoro, Rio  Majes, Arequipa, Peru. 930m.

Fig. 8

Fig. 8. Cumulopuntia crassicylindrica PH 762.03. Hacienda Ongoro, Rio  Majes, Arequipa, Peru. 930m.

Cumulopuntia kuehnrichiana (Werdermann & Backeberg) Ritter

This is the most northern form of C. sphaerica which grows in the Rímac valley inland from Lima.  It was described as a Tephrocactus by Werdermann & Backeberg in 1931 and moved into Cumulopuntia by Ritter in 1981. The type locality is the town of Chosica and I have seen plants close by, on slopes where the urban spread from Lima has yet to reach. It grows into large sprawling clumps in a similar way to C. tumida and C. crassicylindricus. The segments are slightly elongated to 8cm or so in length. The spination is generally on the upper half of the segment only and relatively fine. The plant has the appearance of a weakly spined C. sphaerica (type form).  Figs. 9 and 10 show a typical plant.

Fig. 9

Fig. 9. Cumulopuntia kuehnrichiana PH 780.01. Rio Rimac, Lima, Peru.

Fig. 10

Fig. 10. Cumulopuntia kuehnrichiana PH 780.01. Rio Rimac, Lima, Peru.

The taxa C. sphaerica (type form), C. tumida, C. crassicylindrica & C. kuehnrichiana make up a group of very similar plants, characterised by large segments which vary a little in the amount of elongation. There are much larger differences in the spination which can be used to distinguish between the populations more clearly.

Cumulopuntia unguispina (Backeberg) Ritter

This species was described by Backeberg in 1937.  The type locality is given as the desert of Joya, south west of Arequipa.  This desert belt is formed because it is too far inland for coastal mists to penetrate but also too far west for rains from the east. It is unlikely the plant grows in the total desert because of the aridity.  I have found these plants in several coastal localities in southern Peru from Camaná to Tacna, always in very arid conditions in the fog zone and never in large numbers. Ritter also reports plants from similar habitats west of the desert belt, Ritter (1981:1251).

The original description calls for segments up to 4.8cm in diameter, although Ritter (1981:1251) states a maximum of 2.5cm and all the plants I have found never exceeded this segment size. The areoles are very large in comparison to the size of the segments and perhaps cover in excess of 50% of the segment surface.  When young the segments have significant amounts of white areole wool.  The straight, needle-like, spines radiate out of the areole and are up to 25mm in length and brown when young, turning grey with age. The original description calls for up to 18 spines but this seems to be unusual rather than the norm, and a maximum of 8 to 10 was more typical. Spines are often limited to areoles in the top half of the segment. New growth is bright green and often has small residual leaves present beneath the areole. Mature growth develops a glaucus coating to give plants a grey/green appearance.  Figs. 11 and 12 illustrate a plant PH769.04 at Camaná.

Fig. 11

Fig. 11. Cumulopuntia unguispina PH769.04. Camana, Arequipa, Peru.1000m.

Fig. 12

Fig. 12. Cumulopuntia unguispina PH769.04. Camana, Arequipa, Peru.1000m.

I consider the plants commonly found in northern Chile coastal localities to belong here. They have similarly sized small joints and large areoles which is remarkably consistent over the whole range. In the far north of Chile there is a gap in the distribution with the Peruvian populations but this is due to increased aridity of the region. They still exist at a few favourable localities south of Iquique but in the past the distribution must have been continuous from Coquimbo, Chile to Camaná, Peru.  

Figs. 13 and 14 show a plant at El Cobre, to the south of Antofagasta, Chile, which conforms very closely with C. unguispina from southern Peru.  Note the residual leaves on recent growth. In more favourable localities plants can make larger clusters but retain the small segment size.

Fig. 13

Fig. 13. Cumulopuntia unguispina PH655.03. El Cobre, Antofagasta, Chile. 830m.

Fig. 14

Fig. 14. Cumulopuntia unguispina PH655.03. El Cobre, Antofagasta, Chile. 830m.

This dwarf headed form of C. sphaerica is the most often seen in cultivation, which I suspect originate from Chilean rather then Peruvian localities. Segments are easily detached and rooted. I find it impossible to grow a plant to any size as I find they disintegrate when I attempt to re-pot them. However flowering is possible on plants once they achieve chains of 4 or 5 heads high.

Cumulopuntia sphaerica (Foerster) Anderson (high altitude form)

On the road which climbs out of the Río Majes valley towards Cotahuasi C. crassicylindrica can be found up to approximately 1600m in altitude. There is then a gap where no C. sphaerica forms can be found until the altiplano is reached. Here at 3830m, growing with Oreocereus leucotrichus I encountered C. sphaerica (Figs. 15 and 16).  I found further high altitude populations above Arequipa at 2830m and 3200m, Moquegua at 3700m and close to the Chilean border at 3470m.  Royston Hughes has also encountered similar plants in the Colca Canyon region at similar altitudes. In all cases the plants have small segments, large areoles and short, straight, spination consistent with C. unguispina. Fig.17 shows red flower buds but these open to reveal yellow flowers typical of the C. sphaerica group. The habitats of this form are very different to C. unguispina, as they receive regular annual rainfall but are colder due to the higher altitude. The desert belt in southern Peru which extends into northern Chile isolates the coastal habitats from the high altitude ones and forms a natural barrier between the two populations.

Fig. 15

Fig. 15. Cumulopuntia sphaerica (high altitude form) PH596.02. Road to Cotahuasi, above Chuquibamba, Arequipa, Peru. 3830m.

Fig. 16

Fig. 16. Cumulopuntia sphaerica (high altitude form) PH596.02. Road to Cotahuasi, above Chuquibamba, Arequipa, Peru. 3830m. 

Fig. 17

Fig. 17. Cumulopuntia sphaerica (high altitude form) PH596.02. Flower buds.

Surprisingly a name has not been erected for this high altitude form in Peru but in Chile we have two possibilities:

Opuntia leucophaea Philippi was described in 1891 with a type from Usmagana in Tarapacá, the northernmost province of Chile. Philippi describes a plant with globular segments to 2.5cm in diameter with prostrate branches and a saffron yellow flower with a short ovary. As Iliff points out (2002:207) this plant must be a C. sphaerica form.

Cumulopuntia tubercularis Ritter was described in 1980 from Chusmiza also in the province of Tarapacá.  Ritter also gives an altitude of 3000-3400m. He associates this plant with C. berteri (his name for C. sphaerica from coastal Chile).

Both localities are close together at approximately Lat S19.75°, Long W69.23° inland from Iquique and 200km south of the Peruvian border. Unfortunately no illustrations exist for either plant and it would be beneficial to return to the localities to record what grows there. From the information we have, it appears Cumulopuntia tubercularis is a re-description of Opuntia leucophaea and both belong to the high altitude form of C. sphaerica and extend the range significantly southwards.

As mentioned earlier it was interesting to find a small growing C. sphaerica with 3cm high segments (Fig. 18) together with the large typical form just outside Arequipa.  Unfortunately we didn't have time to explore fully to see if there were any intermediates or if the two forms grow side by side without intermingling. I am placing this plant with the high altitude form rather than with C. unguispina as the location is east and inland of the desert belt.

Fig. 18

Fig. 18. Cumulopuntia sphaerica (high altitude form) PH709.05. Near Minas Cerro Verde, Arequipa, Peru.


Britton and Rose (1919) The Cactaceae Volume 1
Hunt, D (2006), The New Cactus Lexicon
Iliff, J (2002) The Andean Opuntias, Studies in the Opuntioideae, Succulent Plant Research Volume 6
Rauh, W (1957) Beitrag zur Kenntnis der peruanischen kateenvegetation
Ritter, F (1980)  Kakteen in Sudamerika Band 2
Ritter, F (1981)  Kakteen in Sudamerika Band 4

To be continued (see part 2).

This article was originally published in Tephrocactus Study Group (the 'TSG Journal'), September 2008, Vol. 14, No. 3, pages 35-49. © TSG and Paul Hoxey