The second and concluding part of a survey of the fraileas of southern Brazil. Photography by the author.

Fig. 26 Two specimens of Frailea buenekeri ssp. densispina growing in habitat	Fig. 27 An old specimen of Frailea buenekeri ssp. densispina that has become more or less columnar

Frailea buenekeri W R Abraham. This species occurs as two subspecies. The typical subspecies (ssp. buenekeri) grows near the town of Vila Nova do Sul, while the subspecies densispina Hofacker & Herm grows more to the south, to the west of the town of Lavras do Sul and east of the small village of Ibaré. Unfortunately I did not find plants of the typical subspecies, but I saw subspecies densispina in a number of places in the area where it occurs, growing in granitic rock outcrops, in pockets of soil between the rocks, together with mosses, lichens, grasses and other vegetation (Figs. 26 & 27). It is a beautiful little plant of no more than about 3cm in diameter, with a dark green to winered epidermis, many ribs divided into low tubercles, and areoles with a dozen or so creamy yellow to white radial spines appressed against the body of the plant and pointing downwards. Below each areole there is a faint, dark, crescent-shaped mark. The plants are globular, becoming columnar in age. Another characteristic of this taxon is the tuberculate spines, which have small projections along them, thus creating the illusion of being slightly feathered. Near to the places where I found F. buenekeri ssp. densispina, I also found forms of F. pygmaea.

Perhaps to compensate for the fact that I did not find the typical subspecies, I did find a yet more spectacular subspecies of F. buenekeri, that I have provisionally named subspecies cen­trispino. (Figs. 28-29 & rear cover). This plant was found more to the east than any of the other subspecies, near the town of Santana da Boa Vista. It has a more dense cover of pure white spines, and is distinguished by the presence of a single central spine, which is stronger than the radial spines and varies from white to yellowishbrown in colour. Also, the tubercles on the spines are more developed, causing a greater impression that the spines are feathered. The overall appearance of the plants makes them superficially resemble the forms of Mammillaria duwei that have central spines. When I found the first plant of subspecies centrispina, my first impression was that it was a very young seedling of Parodia scopa. The plant was somehow strange for a P. scopa, but it was not until I found other specimens and examined them at close range - on my hands and knees - that I realized that it was a Frailea and not a Parodia at all! At the same place where I found F. buenekeri ssp. centrispina I also found a form of F. pygmaea.

Fig. 28 A beautiful new discovery: Frailea buenekeri ssp. centrispina n.n. growing in habitat	Fig. 29 Frailea buenekeri ssp. centrispina n.n. growing among mosses in habitat

Frailea castanea Backeberg is perhaps the most attractive species of the genus. The plants resemble a miniature Euphorbia obesa or Astrophytum asterias, with its body flat with the ground, and having many low ribs with strings of areoles bearing very short black spines (Fig. 30). The number of ribs varies between different populations, as also does the shape of the flowers, even in the same population (Fig. 31). The colour of the epidermis is usually dark reddish-brown, matching remarkably well the colour of the soil where the plants grow. This, and the fact that the plants retract into the soil during dry periods, makes them quite difficult to spot (Fig. 32). The species grows in areas of shallow reddish soil with rock outcrops of iron-rich sandstone. I found several popula­tions in the area between the towns of Quaraí and Alegrete, and the plant was quite common in places, with hundreds of indi­viduals to be seen. Growing together with F. castanea I often found F. pumila, and in a few places also F. phaeodisca.

Fig. 30 A plant of Frailea castanea growing among branches of the creeping fern, Selaginella sellowii.	Fig. 31 Two plants of Frailea castanea flowering in habitat. Note the different shapes of the petals

Fig. 32 An almost hidden plant of Frailea castanea	Fig. 33 Frailea cataphracta growing in a sandy habitat

Fig. 34 A plant of Frailea cataphracta with a more greenish colour	Fig. 35 A very dark-coloured, almost black plant of Frailea cataphracta

Frailea gracillima (Lemaire) Britton & Rose. Here is a species that is very widespread and morphologically quite variable (Figs. 39 & 40). It is very common in the mountainous region of the southeast of Rio Grande do Sul, coming from the vicinity of the capital city, Porto Alegre (from where the synonymous Frailea alacriportana Backeberg & Voll was described), all the way to the south and as far west as Santana do Livramento. It grows in gravelly soils, under shrubs and amidst grasses, or in humusrich, moss-filled pockets of soil in the rock outcrops. Frailea gracillima is the typical columnar species of the genus, its body always more elongated than wide. The ribs are tuberculate, and the central and radial spines have different colours. The central spines are dark reddish-brown to black, while the colour of the radials varies from white to yellow. Some forms of F. gracillima are rather small, with pencil-thin stems no longer than a couple of centimetres in height, with few ribs and long central spines (for instance, the form described as Frailea lepida Buining & Brederoo), while other forms are much more robust, with stems several centimetres long (up to 15cm) and up to 5cm in diameter, with numerous ribs and short spines.

Fig. 36 Plants of Frailea fulviseta growing between the leaves of a Dyckia (Bromeliaceae). These are yellow-spined variants	Fig. 37 A white-spined variant of Frailea fulviseta

Fig. 38 A group of plants of Frailea fulviseta. The plant in the foreground is a three-headed specimen	Fig. 39 A typical plant of the columnar Frailea gracillima, a thin plant with white radial spines, a few dark, long central spines and dark, crescent marking below each areole

Frailea mammifera Buining & Brederoo is divided into two subspecies, the typical subspecies (ssp. mammifera) growing in Rio Grande do Sul north of the town of Dom Pedrito, and subspecies angelesiae R Riesling & Metzing growing in Dep. Concordia, Prov. Entre Rios, in the northeastern part of Argentina. The place that I visited was the type locality of this charming species, which is globular, up to 3cm in diameter, and has around 16 tuberculate ribs, the tubercles being very prominent and conspicuous, hence the name (Figs. 43 & 44). The colour of the plants is green, and below each areole there is a dark crescentshaped mark. The areoles have about eight spines, all radial, orangeyellow in colour and tuberculate, conferring them with a slightly feathered appearance. The plants grow in granitic rock outcrops, in the pockets of soil between the rocks. I did not find many plants of this species at the locality that I visited, but I do not know if the reason is because it is simply a rare plant, or if the plants just happened to be "invisible" and hidden in the soil at the time when I visited them. In overall morphology, F. mammifera is quite similar to F. buenekeri, and it does not grow very far from where F. buenekeri ssp. densispina also occurs. These two species share the feature of having tuberculate spines, so perhaps they are closely related.

Fig. 40 Two specimens of Frailea gracillima growing in habitat, the one on the right bearing a ripe fruit	Fig. 43 A mature and a young specimen of Frailea mammifera growing at the type locality north of the city of Dom Pedrito
Fig. 41 A typical specimen of Frailea gracillima ssp. horstii, with a thicker body, more ribs and shorter spines than ssp. gracillima. This specimen was damaged and produced two heads. It is very plump because it rained a few days before the picture was taken	Fig. 44 A small but mature specimen of Frailea mammifera where the large tubercles are clearly visible
Fig. 42 A huge specimen of a Frailea gracillima ssp. horstii from a population near the town of Encruzilhada do Sul. This plant had just flowered profusely, as the flower remains visible at the crown of the plant testify. Plants at this location reach 15cm in height and 5cm in diameter, being the biggest Frailea that I saw	Fig. 45 A typical specimen of Frailea phaeodisca, with its dark areoles and white spines, growing at Minas do Camaquã near the city of Caçapava do Sul

Fig. 46 Frailea phaeodisca at Minas do Camaquã near the city of Caçapava do Sul	Fig. 47 Two plants of Frailea phaeodisca growing in their sandy habitat

Fig. 48 A specimen of the form of Frailea phaeodisca that was described as Frailea perbella. The plants of this form have a more elongated body, areoles with more black felt and stronger, interlacing bright white spines	Fig. 49 Specimen of Frailea phaeodisca from a population where the plants have characteristics more or less intermediate between the typical form and the form named as perbella

Fig. 50 A white-spined specimen of Frailea pumila	Fig. 53 Two specimens of a form of Frailea pumila with reddish-brown spines. These plants were growing in a shady place between two big rocks, and as a result they have a more feeble spination than normal, as well as being a little bloated. The whitish threads visible in the picture are desiccated stems of the creeping fern, Selaginella sellowii
Fig. 51 A large plant of Frailea pumila with cream-yellow spines	Fig. 54 Typical examples of Frailea pygmaea, with flatfish stems and whitish spines
Fig. 52 Plants of the golden-yellow form of Frailea pumila growing among mosses	Fig. 55 A specimen of a long-spined form of Frailea pygmaea

Frailea pumila is one of the largest of the globular species, attaining a diameter of up to five centimetres. It has many low, tuberculate ribs, while the areoles have whitish felt and about 18 radial spines that are more or less curved and point in all directions, and between one and four central spines, which are not very different from the radials. Old plants become slightly columnar. This species is quite common in the southwestern portion of the state of Rio Grande do Sul, growing from Santana do Livramento westwards to the border with Argentina, and all the way north up to the cities of Santiago and São Borja. I often found it growing in the rocky beds of creeks (called arroyos in this part of Brazil) or in areas of flat rock outcrops subjected to seasonal flooding. The plants were most plentiful in the places which were obviously flooded during the episodes of intense rain. In these places F. pumila grows in the crevices of the rocks together with mosses and lichens, the plants firmly attached to the rocks by their strong tuberous roots. Frailea pumila shares its habitat with a few other Frailea species, namely F. castanea, F. phaeodisca and F schilinzkyana (F. perumbilicata).

Fig. 56 Specimens of Frailea pygmaea with dark crescentshaped marks below the areoles	Fig. 57 A yellow-spined variant of Frailea pygmaea

Frailea pygmaea (Spegazzini) Britton & Rose. As the name implies (pygmaeus meaning dwarf), this is perhaps the smallest of all Frailea species, some forms having heads no wider than one or two centimetres. It is also quite widespread, growing in Brazil, Uruguay and Argentina. In Rio Grande do Sul I found it growing in the mountainous region in the southeast of the state, but also westwards up to the region around the city of Santana do Livramento. I found numerous populations of F. pygmaea, and it was often quite common. However, this is another very variable species, even more so than F. gracillima, and while F. pygmaea is quite uniform in the centre of its area of distribu­tion in Rio Grande do Sul (in the more southeastern regions of the state, from the city of Bagé eastwards and southwards), in the northern and western edges of its range there are quite distinct forms. In accordance with the variability of the species, many names have been created for different forms of F. pygmaea, some of them correctly published but the majority invalidly.

Typical F. pygmaea has a flattened, light green body, low ribs which are tuberculate (more noticeable when the body is dehydrated), the tubercles being flattish and with areoles bearing whitish or yellowish wool and a few short, bristly, glassy, white radial spines, and rarely also one or two small central spines (Fig. 54). Near the city of Caçapava do Sul I found plants with a more globular body and areoles having longer white spines, the base of the spines being orangeyellow in colour (Fig. 55). Southwest of Caçapava do Sul, west of the town of Lavras do Sul there are plants which have distinctive dark crescentshaped marks below the areoles (Fig. 56). A bit more to the west, north of the city of Dom Pedrito, I found plants with yellow spines and wool (Fig. 57). While different from each other, all these variants are not really that far from the typical form.

Fig. 58 Frailea pygmaea ssp. albicolumnaris, a stouter form distinguished by the glaucous greyishgreen epidermis and the tendency to become columnar	Fig. 59 A large form of frailea pygmaea found east of the city of Santana da Boa Vista. Although columnar, these plants are not the same as ssp. albicolumnaris, being more similar to typical F. pygmaea, however bigger and columnar in shape

So far the most puzzling form of F. pygmaea that I found was a population growing just north of the city of Quaraí in the south-western region of the state (Fig. 60). The overall appearance of the plants and the locality where I discovered it agree with Ritter's protologue of Frailea aureispina, a taxon that is put into the synonymy of F. pygmaea. I compared the plants that I discovered with the description and illustrations provided by Ritter (1979: 209) and have no doubt that what I found is the same as F. aureispina. However, this taxon is the one that least resembles F. pygmaea:. the dark felted areoles are similar to the areoles of F. phaeodisca, and the yellow to orange spines are stronger than those normally found in F. pygmaea. In this area no other forms of F. pygmaea grow, but F. schilinzkyana (F. perumbilicata) is quite common. This latter species also has strong spines, which vary from yellow to dark brownish in colour, but other than in these characteristics it does not resemble the plants that I believe to be F. aureispina. It is possible that the plants that I found and that Ritter named F. aureispina are not related to F. pygmaea at all, its characteristics being the result of natural hybridization between F. phaeodisca and F. schilinzkyana (F. perumbilicata), both of which grow nearby. This is just a hypothesis, but clearly this is a plant that needs to be studied in more detail.

Fig. 60 A form of Frailea pygmaea (?) found just north of the city of Quaraí, that is perhaps the plant described by Ritter as Frailea aureispina. These plants may be the result of hybridization between Frailea phaeodisca and F. schilinzkyana (F. perumbilicata). The specimen in the photograph had its head eaten and was growing a new head	Fig. 61 A two-headed specimen of F. schilinzkyana(F. perumbilicata), a species that although similar to Frailea pygmaea, differs in having a darker epidermis and stronger spination

Frailea pygmaea usually grows in the crevices of rocky outcrops, between the pebbles near these outcrops or in grassy, gravelly areas where the thin soil depth minimizes any competition from other plants. Other species that I found growing in the same places as F. pygmaea include F. gracillima and F. buenekeri, but in the case of the plants that I identified as F. aureispina, they were growing together with F. castanea.

Frailea schilinzkyana (Schumann) Britton & Rose. The form of this species that grows in Rio Grande do Sul was given the name Frailea perumbilicata by Ritter (1979: 206-207). Hunt (2006: 125) mentioned under F. pumila that its "differentiation from F. schilinzkyana is not altogether clear". However, F. schilinzkyana (F. perumbilicata) grows sympatrically with F. pumila in several places in the region between the cities of Santana do Livramento and Quaraí in the state of Rio Grande do Sul, where both species are clearly distinct. Alternatively, F. perumbilicata could be a species distinct from F. schilinzkyana, because while the distinction between F. schilinzkyana and F. pumila is not clear, F. perumbilicata is positively distinct from F. pumila.

Fig. 62 A mature and a young specimen of F. schilinzkyana (F. perumbilicata) growing near the city of Santana da Boa Vista. The very long bristles in the spent flower of the mature specimen is a characteristic of this species

This view is also supported by the different descriptions of F. schilinzkyana and F. perumbilicata given in Hunt (2006: 124-125). Until the question of the identity of F. perumbilicata is resolved, I will refer to it as F. schilinzkyana followed by the name F. perumbilicata in parentheses.

Frailea schilinzkyana (F. perumbilicata) is found in the region around the city of Santana do Livramento in southwestern Rio Grande do Sul, and also westwards to the city of Quaraí. Plants of this species somewhat resemble F. pygmaea, differing by a more globular body which is dark greyishgreen to brownish, a concave apex (reminiscent of F. cataphracta), flat ribs which are scarcely tuberculate, areoles with strong spines which are dark yellowish to reddishbrown and even black in colour and the more conspicuous dark bristles in the areoles of the flower tube (Figs. 61-64). It grows in the crevices of sandstone rocky outcrops, and it is locally fairly common. F. castanea, F. phaeodisca and F. pumila also grow together with F. schilinzkyana (F. perumbilicata).

The taxonomy of Frailea is far from being thoroughly resolved, and detailed studies should be conducted involving more populations and greater population sampling in order to assess the intra- and interpopulational patterns of morphological and genetic variation of the plants in the field. A modern revision of the genus should only be attempted after such a thorough survey. Taxonomic questions aside, Frailea is a genus that contains many beautiful and interesting little plants, as hopefully the pictures and descriptions in this article have shown. The plants of this genus do not take up much space, and deserve to be grown more widely.

LITERATURE CITED:

Britton, N L & Rose, J N (1922) The Cactaceae, Vol. 3. Carnegie
Institution, Washington. Gerloff, N (1997) Frailea cataphracta grows in Brazil as well. Brit. Cact. Succ. J. 15(4): 189-192. Hunt, D R (2006) The new Cactus Lexicon. DH Books, Milborne
Port. Metzing, D & Kiesling, R (2006) Notes on the diversity, biology, and taxonomy of Frailea (Cactaceae). Bradleya 24: 115-128. Ritter, F (1979) Kakteen in Sudamerika. Band 1. Privately published, Spangenberg.

Marlon C Machado

Universitat Zürich, Institut für Systematische Botanik, Zollikerstrasse 107, CH-8008 Zurich, Switzerland Email: machado@systbot.unizh.ch

Fig. 63 A specimen of F. schilinzkyana (F. perumbilicata) with a flower bud	Fig. 64 A specimen of F. schilinzkyana (F. perumbilicata) with a ripe fruit

Back cover: Frailea buenekeri ssp. centrispina nom. prov. in fruit A sensational novelty related to Frailea buenekeri subsp. densispina discovered by Marlon Machado in Rio Grande do Sul, Brazil. All spines are strongly pubescent, even those of the fruit, a character shared by ssp. densispina to a lesser degree, which also lacks the long central spines. The population suffers from the grazing of sheep, and many of the plants had had their heads eaten, but for­tunately most are able to regenerate new heads. The discovery of this new plant is described in this issue in the second, concluding part of a review of southern Brazilian fraileas by Marlon Machado, who is also the photographer.