THE GENERIC CLASSIFICATION OF THE MEXICAN CRASSULACEAE

Jorge Meyrán

2A. Juarez 42, Col. Sanalvaro 02090, Mexico, D.F.

The classification and delimitation of the gen­era of the Crassula family has always been dif­ficult to establish, since their boundaries lack pre­cision and the characters of some species of various taxa overlap the characters of others. Froedcrstroem cites Linnaeus, who points out that Crassula, Cotyledon, Sempervivum and Rhodiola are so closely related to Sedum that they can be combined into one genus or divided into five, if in a given moment the boundaries between them could be established. Linnaeus es­tablished six genera in the present family, of which five have just been cited, and lacking only 777-laca.

The Crassula family was discussed by Schoenland in 1890 in Die Natürlichen Pflanzenfamilien of Engler & Prantl, giving the most definitive treatment until the revision by Berger appeared.

At the beginning of this century Rose estab­lished various genera, some of them with little consistency. Hence, Rose in Crassulaceae, as Britton & Rose later in Cactaceae, tended to mul­tiply the number of genera. The opposite ten­dency also appeared with Otto Kuntze proposing for Crassulaceae only one genus: Sedum.

In 1930 Alwin Berger presented a magnificent revision of this family in the second edition of Engler & Prantl. He divided it into six subfam­ilies: Crassuloideae, Kalanchoideae, Cotyledonoideae, Sempervivoideae, Sedoideac and Echeverioideae. By removing the genus Echeveria from Cotyledon, the groups seem more consis­tent and better established because the first four subfamilies are almost completely confined to the genera of the Old World (with Crassula prin­cipally African and excepting Tillaea, which is cosmopolitan), and with Echeverioideae having the genera of the New World. Only Sedoideae extends into both regions, though predominating in the northern hemisphere.

Walther considers the contribution of Berger excellent, with a criterion for the delimitation of the genera, but does not agree on all points, such as Graptopetalum united with Sedum since he considers it to be in Echeverioideae, and Villadia and Altamiranoa located in the latter subfamily since he feels they should be in Sedoideae due to their central, terminal inflorescence. Berger accepts Pachyphytum as a genus, but places Dud­leya, Stylophyllum and Thompsonella as sections of Echeveria, which Walther does not consider appropriate.

Monographs have continued to appear on var­ious genera: Sedum by Praeger (1921), Froederstroem (1930-35), Clausen (1942, 1969, 1975); Echeveria by Walther (1972); and from the Old World Kalanchoe by Hamet and Sempervivum by Praeger (1932).

In 1954 Jacobsen returns to discussing the en­tire Crassula family, though superficially, in Handbuch der Sukkulenten Pflanzen (which ap­peared in English in 1960), following Berger on Crassula and in which Moran contributed with Dudleya.

At present those who have most studied this family in America are Reid Moran, Charles Uhl, Myron Kimnach and Robert Clausen, among others.

Among the Mexican Crassulaceae, various genera are endemic such as Pachyphytum, Thompsonella, Cremnophila and Tacitus; in oth­ers the number of species is predominantly Mex­ican, such as Echeveria, Graptopetalum, and Vil­ladia; others, such as Dudleya, share their distributions with other countries or are widely disseminated over a great part of the world like Sedum and Tillaea.

Before discussing the Mexican genera, it is proper to take some thought. One group of bot­anists has inclined to gather the species into a great number of genera, possibly because by an­alyzing and dividing the total the parts can be better understood, thus enabling one to identify the species more easily. To determine more and more precisely the characteristics which should define each genus, they had to increase the num­ber of genera. Naturally there comes a contrary reaction, trying to reduce the number of genera, although having to modify the original concept of each genus, leaving the limits more vague. This imprecision makes the identification of the plants difficult, though this, in part, may be re­solved by increasing the subdivisions of each ge­nus.

In the first case we have the example of Sedum, from which Rose separated the genera Clementsia, Corynephyllum, Cremnophila, Sedastrum, Britton Gormania, Britton & Rose Sedella and Jepson Congdonia. In the second case we have Kuntze, cited above, with only one genus for the entire family.

Fig. 1. Cremnophila nutans inflorescence.

Subfamily Crassuloideae

The genus Crassula presents opposite leaves, inflorescence in terminal cymes or thyrsoid, flowers rarely solitary and axillary, sepals, petals and carpels usually five, stamens the same num­ber.

Tillaea was established as a genus by Linnaeus; not only did Britton & Rose accept it but addi­tionally Britton created the genus Tillaeastrum, based on the flowers being solitary in each axil and the carpels having few to many seeds, while Tillaea presents various groups of flowers, though at times they may be solitary and the carpels holding one or two seeds. The differences are small and soon Tillaeastrum passed into syn­onymy. Schoenland and Jacobsen consider Til­laea to be in the genus Crassula and have placed it in the section Tillaeoideae of that genus. Rzedowski treats it as a genus.

Fig. 2. Echeveria difractens.

Fig. 3. Echeveria viridissima.

Subfamily Kalanchoideae

The genus Kalanchoe has leaves opposite or in whorls of three, often with petioles or with amplexicaul bases, with adventitious sprouts (in Bryophyllum), inflorescence usually terminal or rarely axillary, paniculate-cymose; flowers erect or pendant, sepals, petals and carpels four, sta­mens eight. The three genera of the subfamily Kalanchoideae, Kalanchoe, Kitchingia and Bry­ophyllum, were united by Hamet into the genus Kalanchoe. The only wild species in Mexico are K. pinnata and K. tubiflora, plants naturalized in many parts of tropical America, originating in tropical Africa or Madagascar.

Subfamily Sedoideae

In the subfamily Sedoideae, characterized by its principally alternate leaves, by its petals dis­tinct or almost so, slender and widely spreading, and its inflorescence usually terminal, are three Mexican genera: Villadia, Lenophyllum and Sedum.

In 1903 Rose established two new genera: Villadia and Altamiranoa, the first with inflores­cence in raceme, spike or compact panicle, while the second has inflorescence in cymes or cincinni. Due to their terminal inflorescence, which in­dicates affinity with Sedoideae, Walther elimi­nated them from the subfamily Echeverioideae, where they had been placed by Berger. Because the differences were of little importance, Walther, and soon Jacobsen, united the two genera, choosing Altamiranoa. Baehni and Macbride (1954) and Clausen (1940) chose Villadia and much later Jacobsen changed, adopting the latter course.

Fig. 4. Graptopetalum fruticosum inflorescence.

Villadia is characterized by caespitose plants, roots at times tuberous, leaves at times amplexicaul, semi-amplexicaul, or sessile, linear to oval spatulate to slightly acute, inflorescence terminal in a raceme, spike or cyme, flowers small and petals united at the base forming a short tube.

Lenophyllum Rose is a genus from northeast­ern Mexico, with one species in the south of Tex­as, characterized by plants glabrous, caespitose, with few leaves, large and opposite or verticillate below, the upper ones smaller and alternate, fleshy and more or less concave on their upper surface; inflorescence terminal, erect, with flowers soli­tary or few in racemes, cincinni or spikes. The petals are erect, without contact at base with neighbors, the point alone elongated or recurved. Surely this last was the reason for which Berger placed it in Echeverioideae. No disputes are found concerning this genus, or perhaps there are none because it is little studied.

The genus Sedum L. consists of succulent plants, the majority perennial, some annual or biennial, glabrous or at times pilose, leaves flat or cylindrical, entire or nearly so, usually alter­nate, on rare occasions opposite or verticillate, inflorescence commonly cymose, terminal, at times lateral, flowers pentamerous, at times 4-merous, rarely 3-6-7 parted, petals usually sep­arated from the base or almost so, spreading, at times erect, stamens double the number of petals, at times the same number.

Fig. 5. Lenophyllum weinbergii.

In the description of the genus, the vagueness of the characteristics is obvious, and when some botanists wished to be more precise, it was nec­essary to create new genera, some monotypical or of few species, such as Sedastrum Rose, Tetrorum Rose, Congdonia Jepson, and others.

With the discovery in recent decades of new species, especially from Mexico and China, the lines of demarcation of Sedum have become more indefinite. The section Pachysedum, with a lat­eral inflorescence, is more closely related to some Echeverioideae than to the majority of the Mex­ican members of Sedum. Sedum is the largest genus of this family and Moran has called it the traditional depository of the anomalous species of Crassulaceae. When a species does not fit well in the other genera, the easiest and most practical disposition is to put it into Sedum; there are species with five stamens, with inflorescence in raceme, with erect petals, with urceolate corolla, etc.

Fig. 6. Pachyphytum bracteosum.

Fig. 7. Pachyphytum viride inflorescence.

The genus Cremnophila proposed by Rose for Sedum nutans in 1905 was suppressed because of its petals being distinct and spreading, the rea­son for which it was retained in Sedum by Praeger (1921), Berger (1930), Walther (1931), Froederstroem (1935) and Clausen (1943, 1959). It was restored as a genus by Moran in 1978, group­ing Sedum nutans with Echeveria linguaefolia principally for its lateral, pendant inflorescence in the form of a narrow thyrse and for its large, thick leaves. Walther believed E. linguaefolia to be the most primitive member of Echeveria and suggested that the transition to Sedum should be around this plant. The cytogenetic evidence con­firms the similar morphology of these two species, which suggests strongly that they must be­long to one genus, and since not falling within either Sedum or Echeveria they must remain to­gether, but separate from both.

Subfamily Echeverioideae

*B81340-4-8.jpg

Fig. 8. Sedum allantoides.

Fig. 9. Sedum hultenii.

Walther restricts the final subfamily, Eche­verioideae, to plants with lateral inflorescence, with flowers whose petals are almost always unit­ed at the base, principally erect, at least in the lower half, frequently thick. Even so, the bound­aries are not strong and distinctive. In this sub­family we have seven genera: Thompsonella, Cremnophila, Graptopetalum, Tacitus, Pachy­phytum, Echeveria and Dudleya.

Thompsonella Britton & Rose is a genus clear­ly distinct for its lateral inflorence, erect, in a thyrse below, rising into a spike above, with flow­ers having a short tube and petals which are con­nate in the lower half and spreading in the upper half and having stipitate carpels at the base. The leaves have one central vein and various shorter laterals (like Tacitus and some species of Grap­topetalum).

Cremnophila Rose already has been treated when discussing the subfamily Sedoideae and should be placed in Echeverioideae principally for its lateral inflorescence.

The genus Graptopetalum Rose, although it is variable in its vegetative aspects, is sufficiently homogeneous in certain floral characteristics such as the erect or ascending sepals, the petals which form a tube in the lower part and spread in the upper half, tapering throughout their length (not widening toward the middle as in other genera) and are variously marked with dark red, usually in transverse stripes, the orifice of the tube open (which distinguishes it from Tacitus). The sta­mens tend to become reflexed at the end of anthesis and the styles are short, rarely more than 1 mm long. The inflorescence is a cyme with short uniparous branches.

The genus Tacitus Moran presents an inflo­rescence similar to that of Graptopetalum, with few flowers, but the floral characters are different: sepals reflexed, petals spreading abruptly from the upper end of the tube of the corolla, much wider (two or three times) above than at the base, having no markings or stripes, the upper orifice of the tube is closed by growths on the base of the segment extended. The stamens remain as­cending and the styles measure from 2lh to 4% mm long.

Fig. 10. Tacitus bellus inflorescence.

Some botanists have wished to include Tacitus within Graptopetalum. To do that they had to modify the concept of the latter genus, making it more imprecise, above all in the floral char­acters. The species of Graptopetalum form a ho­mogeneous group which to me seems illogical to break up, and moreover the characteristics of Tacitus are so special that it cannot enter any other genus.

Pachyphytum Rose is a genus quite well de­fined, characterized by its very thick leaves, lat­eral inflorescence, appressed sepals shorter or longer than the petals, those erect, very briefly joined at the base, separated below (as in Lenophyllum), with two appendages at the sides of the epipetalar filament.

Lately species have appeared with similar ap­pendages which by all their other characters be­long to Echeveria, which made one think that that genus had lost one of the characters which was considered exclusive. Nonetheless, these ap­pendages seem to be clearly a fold of the lower edges of the petals; this point of view is supported by the fact, noted by Leinfellner (communicated by Moran) that the vascular bundles are inverted in the said appendages in a way that their ho­mology with the appendages of Echeveria viridiflora and E. dactylifera is not yet demonstrated. Taking into account the total array of characters, Pachyphytum is a sufficiently well defined genus.

Echeveria DC is a genus of rosulate plants with leaves not amplexicaul, with one vein near the base, with inflorescence in cincinnus, raceme, or panicle with cincinni, the leaves of the floral stem never amplexicaul, rarely cordate at the base, normally with brilliantly colored flowers, sepals frequently widely spreading at anthesis, com­monly with thick petals, erect, nearly always im­bricate, keeled on the dorsal side and grooved basally, large nectaries, thick and truncate. Plants principally from Mexico, Central America, from northern South America to Argentina, and one in Texas.

*B81340-4-11.jpg

Fig. 11. Thompsonella platyphylla.

Fig. 12. Villadia elongata.

Fig. 13. Villadia batesii.

There were created various genera, such as Oliveranthus Rose, Urbinia Rose, and Courantia Lemaire, which later were returned to Echeveria because the differences were not important.

Dudleya Britton & Rose is characterized by amplexicaul leaves with various veins more or less equal near the base, inflorescence lateral in cincinnus or branched with terminal cincinni, leaves of the floral stem amplexicaul or cordate, commonly with pale flowers, at times red or yel­low; sepals never spreading, appressed and equal, slender, convolute petals, neither prominently keeled nor grooved, nectaries small. Though the Mexican species of Echeverioideae and other Crassulaceae can hybridize easily, this has not taken place with Dudleya despite various at­tempts. The plants derive from Baja California, Sonora, and southwestern United States.

The genus Stylophyllum Britton & Rose was established because its petals are spreading from the middle part, and the genus Hasseanthus Rose because of its subterranean, unbranched, cor-miform main stem, and for its spreading petals and follicles, whereas in Dudleya the petals are erect except at the point and the follicles are erect and appressed. These differences are not suffi­cient to keep them separate and they were trans­ferred by Moran as subgenera of Dudleya.

Commentary

It is as bad to multiply the number of genera excessively as to reduce them exaggeratedly. Lin­naeus founded his system based principally on the floral structure, leaving in second place the characters of the vegetative organs. But since then have come persons, at times establishing sys­tems, who have created new genera based on differences in seed morphology, or on the pos­sibility that the evolutionary line of a group of plants which is not grounded in the fossil record, is derived in the past from one group or another, or such speculation.

To finalize, if a family is divided exaggeratedly into numerous genera, it may be criticized by one group of students; but if it is not divided and few genera are established, it is necessary to sub­divide later to have good classification and to facilitate the identification of species. That is to say, the problem is whether to subdivide before establishing a genus or to subdivide afterward.

If a genus is divided afterward, I believe it is difficult to remember the names. Exaggerating ad absurdum, if all the Crassulaceae were Sedum, it would be difficult to recall all of the 1,500 species which the family holds; alternatively, the different generic names facilitate their memory, aiding when they are based on some important character, as in Pachyphytum, Graptopetalum and Tacitus, or because they recall an important per­son such as Atanasio Echeverria, Manuel Villada, etc.

Another method, which appears the correct one to me, is not to exaggerate in any way the two tendencies, to take an average course; if a genus has sufficient characteristics which make it logical, compact, congruent, it is better not to modify it (as Graptopetalum); if various genera present characters common among them and in­termediate species exist, it is convenient to re­unite them into one (as Sedum); and lastly, if distinctive, very special characters are presented, accept that genus as monotypic (as Tacitus). Moran has made the following comment: that everyone wishes to take the middle course, but none can agree on where is the middle!

Key to the Genera of the Mexican Crassulaceae

A. Stamens same number as petals and carpels .......... Crassula

AA. Stamens double that of petals and carpels.

B. Flowers tetramerous; corolla more or less tubular; leaves principally opposite .......... Kalanchoe

BB. Flowers mostly pentamerous.

C. Inflorescence usually terminal, at times lateral; petals usually free or almost so, rarely connate below and spreading above and then inflorescence terminal .......... (Sedoideae)

D. Inflorescence terminal; petals forming a tube below.

E. Petals spreading in upper half; leaves alternate, linear to oval spatulate .......... Villadia

EE. Petals erect, only the points spreading or recurved; lower leaves opposite, concave on upper surface .......... Lenophyllum

DD. Inflorescence terminal or lateral, petals distinct or nearly so, usually spreading .......... Sedum

CC. Inflorescence always lateral and axillary; petals united at base, erect or spreading above; leaves more or less rosulate .......... (Echeverioideae).

F. Inflorescence in a thyrse or spike.

G. Inflorescence erect; petals connate in lower half, rotate in upper, carpels stipitate .......... Thompsonella

GG. Inflorescence in a narrow thyrse, pendant; petals distinct or connate below .......... Cremnophila

FF. Inflorescence a cyme, cincinnus, raceme, or panicle with cincinni.

H. Inflorescence a cyme; petals forming a tube below middle, then spreading.

I. Petals not widened in outer half, marked with dark red; orifice of tube of corolla open; stamens reflexed at end of anthesis; styles 1 mm long .......... Graptopetalum

II. Petals widened 2-3 times in outer half, not marked; orifice of tube of corolla closed; stamens rising at end of anthesis; style 2-4 mm long .......... Tacitus

HH. Inflorescence a cincinnus, raceme, or panicle with cincinni; petals erect excepting at point (except in Stylophyllum and Hasseanthus).

J. Inflorescence a solitary cincinnus; petals with appendages at the sides of the epipetalar filaments .......... Pachyphytum

JJ. Inflorescence a cincinnus, raceme, or panicle with cincinni; usually without appendages.

K. Petals imbricate, leaves not amplexicaul, one vein in base of each leaf; sepals commonly spreading; flowers of brilliant colors .......... Echeveria

KK. Petals convolute; leaves amplexicaul, various veins in base of each leaf; sepals principally appressed, rarely ascending; flowers of pale colors .......... Dudleya

References

Britton, N. L., and J. N. Rose. 1905. Crassulaceae. North Am. Fl. 22:7-74.

Clausen, R. T. 1959. Sedum of the Trans-Mexican Volcanic Belt. Cornell Univ. Press, Ithaca.

--------. 1975. Sedum of North America north of the Mexican Plateau. Cornell Univ. Press, Ithaca.

Froederstroem, H. 1936. The genus Sedum, Part 4. Acta Horti Gothob. 10(Append.):l-181.

Jacobsen, H. 1960. A handbook of succulent plants. Blandford Press, London.

Moran, R. 1978. Resurrection of Cremnophila. Cact. Suc. Jour. 50:139.

--------, and J. Meyran. 1974. Tacitus bellus. Cact. Suc. Mex. 19:75.

--------, and C. H. Uhl. 1964. The inflorescence of Echeveria. Cact. Suc. Jour. 36:167.

Praeger, R. L. 1921. An account of the genus Sedum as found in cultivation. J. Roy. Hort. Soc. 46:1-314.

Uhl, C. H. 1970. Chromosomes of Graptopetalum and Thompsonella. Amer. Jour. Bot. 57:1115.

--------. 1976-1985. Chromosomes of Mexican Se­dum. Rhodora 78:629 (1976); 80:491 (1978); 82: 377(1980); 87:381 (1985).

--------, and R. Moran. 1953. The Cytotaxonomy of Dudleya and Hasseanthus. Amer. Jour. Bot. 40: 492.

Walther, E. 1931. Alwin Berger's Crassulaceae. Cact. Suc. Jour. 2:383, 408.

--------. 1937. Thompsonella. Cact. Suc. Jour. VIII: 100.

--------. 1972. Echeveria. Calif. Acad. Sci., San Fran­cisco.

The foregoing is a translation from the Spanish of an article which appeared in Cactaceas y Suculentas Mexicanas, XXXIII, No. 4 (October-December 1988). The translation was made by Bob McClurkin with a most necessary assist by his friend, Marlene Rainman, to whom the writer wishes to express his very deep gratitude.

© Cactus and Succulent Journal US, 1991