SULCOREBUTIA, FOOD FOR TAXONOMISTS ? Johan Pot

Many lovers seem to have an opinion about the nomenclature of their . But do they really know the right name? Who will determine this and in particular, how is it done ? We haven’t heard the last word on this subject.

Creation of the image this from the name. And I heard it perso- My cactus hobby was scarcely born, when nally from Backeberg. He cannot be I met Karel. He presented himself as a wrong, as he found them himself. You say very experienced collector. He invited me ? You mean Echinocactus? But emphatically to visit him. And not to hesi- these plants are lobivias and there is an tate in asking any question. As I was eager end to it !” to learn, I accepted the invitation gladly. I didn’t understand much of it. Had Jaap Within a week I had entered his sanctum. really had contact with the famous Backe- It was indeed a paradise. Many plants berg ? On the other hand Gijs didn’t look were in bloom. In every flowerpot there stupid. I chose to keep silent as I didn’t was a label with a name on it. This was want to be thought a dummy. A few mi- quite a different experience to “120 cacti nutes later the guests left. Karel muttered in colour”. to me in offended tones “This Jaap, he Shortly before the visit somebody asked must always know better! The only thing me, if I already had the “hand of the ne- that really matters is that you understand gro” (Maihuenopsis clavarioides). I hadn’t me. Isn’t that right?” a clue, but Karel showed me the Not lobivias, but sulcorebutias took up all immediately. He was definitely an expert. my attention during the next years. But “But”, he said, “my great love are the during my search for possible correct jonias.” These I also knew nothing of but names I have heard such conversations Karel showed me them and he was entran- over and over again. Apparently we cactus ced as he described the beautiful flowers. lovers feel a deep satisfaction in such At that very moment Jaap and Gijs talks, in which expertise is not really re- dropped in by chance. quired. I believe that German-speaking “These are not jonias”, Jaap corrected. people use the word “Bierernst” in such “These are lobivias. To be more precise cases. I was struck by the fact that many Lobivia jajoiana from Bolivia.” Karel statements made no reference to suppor- looked bewildered but kept silent but Gijs ting information. Observations were har- raised his eyebrows: “Actually they are dly done, but this was no obstacle to ha- echinopsis from Argentina.” As if stung ving a firm opinion. by a wasp Jaap turned. “They do come for For the record I should say that names sure from Bolivia. After all you can derive used in this article correspond to what hobbyists on the continent use to say. This wer originated from outside the areole, it does not mean that the original authors had to be related to Mamillaria “without will always still support these names. any doubt”. The shape of the corolla and Therefore author citations are omitted in the pericarp however prevented clas- this paper. sifying it here. Some years later Schumann withdrew . Rebutia 28 years later on Spegazzini (1923) defi- A small plant with the name Echinopsis ned the genus Aylostera. The decisive minuscula was brought onto the market feature was a partial fusion of the style by Pierre Rebut. It probably came and the tube. Using this characteristic one originally from Argentina. Later on such could clearly distinguish Aylostera from plants were indeed found in the province Rebutia. (Fig 2 and Fig 3) of Tucuman. Have the observations of Schumann been In 1895 the genus Rebutia was defined by checked ? I suppose so. Though I was K. Schumann. (Fig 1) Schumann had ob- never able to confirm that Rebutia did not served that the new plant did not bloom bloom from the areole. But I have heard from the areole, so it could not be an echi- amateurs discussing seriously about pistils nopsis. The plant itself resembled an echi- which had grown together to the tube op- nocactus or malacocarpus, but as the flo- posite free pistils.

Fig. 1 Fig. 2 Flowersection Rebutia minuscula

Fig. 3 Flowersection Fig. 4: candelaris (Photo: Craig Howe) Aylostera schatzliana JK423 http://cactiguide.com/cactus Berger (1929) did not mention a genus proach is a natural way of acting, but of Aylostera. He recognized only one genus course it contains the risk of overlooking a Rebutia. He wrote: “Small plants, roughly significant characteristic. globular, reminiscent of mamillaria, with In the same way relationship between tubercles in spiral rows and small spines. plants will not only be accepted, but de- Flowers from the older areoles, often ori- nied as well. Many people do accept Re- ginating close to the base, small, funnel- butia minuscula and Aylostera pseudomi- shaped, with slender tube, open by day. nuscula to belong to the same genus. But These small plants from the mountains are who will for example expect a closer rela- not to be classified in Echinocactus nor in tionship between Rebutia minuscula of Echinopsis.” only a few centimetres diameter and I have some problems with this statement. height and a 5 meters high Browningia Schumann used the observation that the candelaris ? (Fig. 4) Nobody, don’t you plant did not bloom from the areole to think ? More about this later. distinguish the genus. The same characte- During the winter of 1929-1930 the plant ristic was denied by Berger. What did he collector José Steinbach sent Werdermann observe instead to be still able to reco- a plant, which “probably had been disco- gnize a genus Rebutia, distinguishable vered in the wider surroundings of Cocha- from other genera? We never will know. bamba (Bolivia) at an altitude of 2500 m”. Berger mentioned 6 species: Rebutia mi- Unfortunately Steinbach died shortly af- nuscula, R. deminuta, R. pseudominuscu- terwards. Werdermann identified the plant la, R. pygmaea, R. fiebrigii and R. stein- as “doubtless” being related to Echinocac- mannii, of these the first four species came tus minusculus. “Schumann had defined, from Argentina and the others from Boli- based on its special flowering characteris- via. tics, the genus Rebutia, which he wi- Does the characteristic, that pistil and tube thdrew later on, and classified this plant as are partially grown together, make any Echinopsis.” But Werdermann himself sense? One taxonomist will find it impor- argued to reintroduce the genus Rebutia tant, the other will dismiss it. It was Rit- for this “well characterized” group. It may ters opinion (1980), that the fusion hardly be just me, but I have no idea what Wer- had taxonomical meaning, as it would dermann meant with “well characterized”. have developed simultaneously in diffe- Did he possibly understand Rebutia in the rent separated lines. same way as Berger ? Often the characteristic is ignored. How Werdermann (1931) described the would early taxonomists have decided “interesting” plant, which strongly whether a plant belonged to Rebutia or resembled Lobivia boliviensis qua habitus, not ? I suspect that plants have been but nevertheless was called Rebutia stein- thought to be related after a rough obser- bachii based upon one single flower. A vation without a real check. Is this second opinion was impossible, as the strange ? No, because such things still plant had died when the publication was happen. It would be really strange, if one made. By the way, to my mind a plant can used a formal checklist of characteristics be called interesting if it is little known for every new plant to decide to what ge- and in the mean time is or will be desi- nus it belongs. The rough and ready ap- rable. It must be more than just a study- object. The possession of interesting in 1849, was also classified in this genus. plants works as a lure for collectors by the The result of this remarkable sequence of way. Thinking back, I guess, that in this events was nevertheless correct. Not only sense Karel did not have many interesting was the genus poorly defined plants. Who indeed wants a “jonia”? by the summary description of Backeberg, the plant description also was not clear, as Weingartia the name had been used twice, for com- As far as I know all authors mentioned pletely different plants, which had died above were professional botanists. But long ago. Boom (1962) dedicated a very amateurs also made their contributions worthwhile article to the correct name of heard. Very well known was Curt Backe- the plant. He ended his account with the berg. According to Wikipedia he met the remark “Look at what the consequences Czech plant collector Alberto Vojtěch Frič may be if the international Rules for Bota- by chance in 1927. His stories stimulated a nical Nomenclature are not employed in desire for adventure in Backeberg who the right way; it is definitely necessary, decided to import cacti himself. He was so that everybody, who is occupied with the strongly fascinated by these plants that he of cacti (and of course of all published a lot, for example the standard other plants) acquaint themselves with the work of six parts „Die Cactaceae“. proper use of these rules. Especially in the In 1933 Backeberg defined the genus Spe- case of cacti much incompetent work has gazzinia with the species fidaiana (Fig. 5) been done in this area.“ Then a list of and neumanniana (Fig. 6). Backeberg had names followed. In case of ambiguous observed a hairless scales tube with these names was added „quoad descr.“ plants, while the flowers kept open until they had wilted. This last characteristic Echinocactus cummingii Salm-Dyck has not been noticeable with my own non Hopffer, Cat. Hort. Dyck.: 174 plants. Had Backeberg really compared (1849), (1850), Allg. Gartenz. 11, 225 (1843). the tube with the one of Rebutia minuscu- Lobivia cumingii (Hopff.) Br. & R., The la? Albert Hofman (personal statement) Cact., 5, (1922), quoad descr. assumes that the first description of the Oroya cumingii Kreuz., Verzeignis: 39 species fidaiana (fidana Hunt 2006) has to (1935). be taken as a first description of the genus cumingii (Hopff) Backbg., also. Well, I have seen quite some cacti Kakt. ABC: 298 (1935), quoad descr. with this label, but with many of them I Weingartia cumingii (Hopff.) Werd. ex really had doubts that these were the Van Oosten, Succulenta 21: 129 (1939), plants meant by Backeberg. Are experts quoad descr. more flexible than me? Or is there insuffi- Weingartia neocumingii Backbg., Kakt. cient quality of reference data and. Sukk., 1: 2 (1950). (incomplete, inaccurate), as a result of Gymnantha cumingii (Hopff.) Ito, Expl. which classifications are also unreliable. Diagr.: 53 (1957), quoad descr. Another genus Spegazzinia already Gymnocalycium neocumingii (Backbg.) existed at the time. Therefore Werder- Hutch., Cact. & Succ. J. (U.S.): 29 (1957). mann changed the name of the genus into Gymnocalycium cumingii (Br. & R.) Weingartia (1937) without supplying extra Hutch., Nat. Cact. & Succ J.: 14 (1959). information. Later on Echinocactus cum- mingii, which had been described already I can imagine that not only the interested amateur will be lost here, the professional- Backeberg. Is this really important ? taxonomist will also perhaps do a double- Because herbarium material of the type- take. How could the name of a species be plant of Werdermann no longer existed, in connected to different genera so often in 1999 a neotype was deposited in the her- only 35 years ? Searching for explana- barium of the Städtische Sukkulenten- tions, one is reminded of the speed with sammlung in Zürich. This suggests it which Jaap solved taxonomical problems. should be a plant from the original popula- Personally I would guess that the poor tion. Would this plant completely corres- quality of information is the cause of this pond with the description of Werder- chaos. mann ? David Hunt (2006) explained: “The type Sulcorebutia (which does not have to be an average or Backeberg defined the genus Sulcorebutia ‘typical’ specimen of the species or other (1951). Here is an attempt to interpret the taxon concerned) gives the botanist an Latin text : „Plants caespitose, with rather absolutely fixed point of reference from small offsets, ribs tuberculous; tubercles which to judge whether other specimens to lobivoid, axeshaped (!), with crack (!); which the name has been applied are cor- flower funnel form, originating from the rectly identified or not.” Then the neotype circle round expanded crack, with the was superfluous, if it was identified, based colour of the morning sun, with scales, on the plant of Werdermann, to belong to hairless (!) fruit still unknown – Bolivia, the same species. However if the identifi- near Colomi (Cochabamba) in an altitude cation was different, the deposit would of 3400m (Cardenas). Typus: Rebutia lead to a paradox. Steinbachii Werd.” To Backeberg the crack (sulco) at the The characters followed by “(!)” will have upper side of the tubercle was a distinguis- been the main criteria. It is nice to con- hing characteristic of the genus, therefore clude, that this text differs from the one in the name Sulcorebutia. Moreover he men- “Die Cactaceae” (1959). tioned the lobivoid appearance of the plant Werdermann described the colour of the and the hairless flower. The totally free flower to be red, Backeberg mentioned pistel was mentioned in the English lan- “the colour of the morning sun” like the guage comment. (Fig. 8) one of Rebutia violaciflora. (Fig. 7) Pre- Although Cárdenas sent Backeberg a viously I had heard the explanation, that couple of plants, he rejected this genus. some taxonomists were rather flexible in There was a rumour that he objected in the their interpretation of the colour red. Later first place to the amateurish procedure of on Pip Smart told me, that Martin Cárde- Backeberg. Anyway, he rejected Sulcore- nas from Cochabamba had looked for butia and described for example Rebutia years for a plant like the one of Werder- arenacea and R. glomeriseta in 1951, R. mann, with positively real red flowers. He tiraquensis, R. totorensis and Aylostera never found one. Obviously his quest Krugerii in 1957 and even Weingartia brought him to the area east of Cochabam- torotorensis in 1971, all of which were ba. Backeberg got his plants from Cárde- recombined to Sulcorebutia. I never un- nas. It seems plausible, that the type-plant derstood why Cárdenas believed to reco- defining the genus comes from a popula- gnize in this plant an aylostera (with fused tion different to the one described by pistil and scales on the tube covered by Fig. 5: Weingartia fidaiana Fig. 6: Weingartia neumanniana hairs). into 30 sub characteristics. The presence John Donald (1971) made a stand against of these sub characteristics was measured. the opinion of Cárdenas. First he men- It resulted in the following table: tioned that deliberate hybridization bet- L 15 10 10 1 ween Sulcorebutia and Rebutia had W 10 15 13 4 brought no descendants. One can pollinate Sulcorebutia and Lobivia with Chamae- S 10 13 15 4 cereus, but not with Rebutia. Also one can R 1 4 4 15 pollinate Weingartia and Sulcorebutia L W S R with each others, but again not with Rebu- The main characteristics were (1) structure tia. Also hybridization between Rebutia of the rib, (2) tubercle, (3) position of the and Lobivia is not possible. areole, (4) structure of the areole, (5) ap- Cárdenas had criticized the European pearance of the flower, (6) structure of the taxonomists because of the lack of field receptacle, (7) insertion of the filaments, experience with Rebutia, as a result of (8) fruit and (9) seed. which they were not in a position to deve- I’m afraid I do not understand this expla- lop a proper classification. Such remarks nation of Donald very well. are rather familiar to me in circles of cur- Franz Buxbaum (1967) did recognize the rent day specialists. Donald however had genus Sulcorebutia. He believed he was observed hundreds of plants transported able to map the connections of taxonomi- directly from their natural surroundings, cal units as determined by ancestry. Of so he had a right to speak. In his opinion course these units had to be monophyletic. the separations between species was He intended that a group of taxonomical blurred and he recognized clines, for units should have a common ancestor, example Sulcorebutia candiae, S. me- from which all the members of subgroups nesesii, and S. xanthoantha and Sulcore- - but no other groups - are descended. butia kruegeri, S. arenacea, S. caineana Buxbaum compiled the following clad and S. breviflora (brachyantha). At last gram – in this case a phylogenetic tree. Donald compared sulcorebutias, rebutias, (Fig. 9) weingartias and lobivias with each others based on 9 main characteristics, divided Shifting insight Sulcorebutia steinbachii as a species of It is not clear to me how Ritter recognized reference. Actually I did not find any clear that for example Aylostera krugerii was a motivations for these suggestions. By the sulcorebutia, if he only used the characte- way, according to Donald Weingartia ristics of the genus Sulcorebutia men- neumanniana was found in the area of the tioned by Backeberg. Maybe he did Bolivian-Argentinian border, though something like this: B looks like A, C Backeberg himself stated the area near looks like B, D looks like C, and so D Humahuaca, 125 km more southward. looks like A? Ergo D is related to A ? Friedrich Ritter (1980) concluded that the Donald also accepted a considerable list of observation of Backeberg concerning the sulcorebutias, doubtless by a process of tubercle was based on an illusion. There shifting insight. This led automatically to was no crack at all, only a fold. This phe- an extension of interpretation, which was nomenon was found in other genera as followed by an amendment to the genus well, for example in Weingartia. But Rit- (1972). In this the crack on the tubercle, ter did recognize the genus Sulcorebutia what gave the genus its name, was still because “the short, light edged scales of mentioned. the flower are not these of Rebutia“. Fred H. Brandt (1977) surprised friend and Moreover he had observed, that the foe alike by the joining of Weingartia and areoles (?) on the flowers of Rebutia (as Sulcorebutia. He wrote: “As a result of far they were not hairless) except wool, observations spanning many years as well almost always bear several to many as investigations of the plants and seeds of bristles. Ritter believed Rebutia to be a these two genera I have come to the con- genus on its own, to which also Aylostera clusion that these genera are one closed with its hairy scales belonged. In spite of unit and thus have to be joined.“ Conse- this remark I cannot build on the conside- quently all sulcorebutias were recombined. rations of Ritter. Nevertheless they were It is a pity that I could not recover details good enough for a series of recombina- of his observations. Nor do I know what tions, like Sulcorebutia arenacea, kruge- he investigated. When I was introduced rii, tiraquensis, totorensis, which obviou- shortly afterwards to well-known sulk- sly were accepted by everybody. Of specialists I was pressed emphatically to course it was annoying, that this seemed to forget Brandt. Had he committed a here- be overruled by the activities of Brandt. sy ? Inquiry brought no more than a re- Were the arguments of Ritter more con- currence of the statement. vincing? Donald and Brederoo studied for a consi- derable time to decide if or how to sepa- rate the genera. It was not very easy to obtain an unambiguous judgement. Per- haps one had to think of three strongly related groups: (I) the group of the sou- thern weingartias with Weingartia fidaia- na as a species of reference, (ii) the group of the northern weingartias W. neocumin- gii as a species of reference (Fig. 10) and (iii) the group of the sulcorebutias with Cover : Fig. 7 - Sulcorebutia steinbachii JK 94

This article originally written in Dutch by Johan Pot was published in the journal Succulenta 90:4 (2011) (p. 190 -196) Until today, this English version (translation by Johan Pot) was unpublished.

Published with the permission of the author.