DAVE FERGUSON’S VIEW ON MAMMILLARIA CLASSIFICATION
writted in 09-2000 Slightly modified in wording, but with no rearrangement on Oct. 20, 2004
Here are some observations and opinions on Mammillaria classification.
David Hunt’s classification of the genus has been the most accepted for a number of years now. Hunt’s classification of Mammillaria has been modified a bit here and there over the years, but not much, and I don’t agree with many things about it.
At the beginning of the new Pilbeam book on Mammillaria, Luthy’s classification of Mammillaria into subgenera, sections, and series can be seen as compared with that of Hunt. In my opinion Luthy is getting closer to a logical and natural classification than Hunt did. There are, however, some problems with Luthy’s classification as well. Some of these are pointed out by John Pilbeam at the beginning of his book; some seem just too ingrained in tradition to go away easily.
One of the problem areas in Luthy’s classification is the separation of the series Longiflorae (sensu Hunt, plus a few additions) into two sections called Krainzia and Archiebnerella. Both of these sections in Luthy’s scheme include species which are not at all related to the Longiflorae (i.e. M. barbata, M. wrightii, M. albiflora, M. herrerae, & M. humboldtii of Mamillopsis and the Lasiacanthae). The remainder of his Archiebnerella are much to closely related to one another to be split into separate sections and should be in the series Longiflorae of his section Krainzia (= Longiflorae). This would leave his Archiebnerella with no members of its own! One species, M. aurilanata, which is very close to M. zephyranthoides, is still left entirely out of this relationship by Luthy.
Another problem is very closely related, and involves the scattering of various species of Mamillopsis (i.e. Dolicothele & kin) among several series and sections, when they should (at least in my opinion) all be brought together under one very distinct subgenus. These include the M. barbata/wrightii complex (which is lumped in with the unrelated M. tetrancistra and M. guelzowiana by Luthy), M. oteroi (lumped with the unrelated series Sphacelatae), M. glassii (into the unrelated series Bombycinae), M. senilis (which remains in its own section), and M. beneckei (which remains in its own subgenus). After checking dates of descriptions it turns out that Mammillopsis is the oldest subgeneric name for Dolicothele. [sort of like the Thelocactus into the Leuchtenbergia situation], so all the species of Dolicothele would become Mammillopsis.
I mostly like the subgenus Cochemiea as recognized by Luthy, but some of the series and sections are still too close to Hunt’s arrangement. Also, Luthy splits the subgenera Cochemiea and Mammillaria in an unnatural way, and many species which should be included in subgenus Cochemiea are listed under subgenus Mammillaria as Luthy’s section Cylindricothelae.
The series under the Cylindricothelae still show strong influence from Hunt’s classification. The species “groups” under the various series are mostly fairly natural, but these “groups” are composed themselves in unnatural ways. Thus, the series Stylothelae [just what is the type species of the Stylothelae anyway??????], Lasiacanthae, Pectiniferae (should be a subset of Bombycinae), Bombycinae, and Proliferae need some serious rearranging. The series Sphacelatae should actually be moved to the Section Cochemiea where the series Ancistracanthae is most closely similar.
M. pottsii, which is actually close to M. bombycina, is still misplaced in the series Leptocladodae, but the remainder of the Leptocladodae species are finally placed in a reasonable position near the Heterochlorae and Polyacanthae.
Here is my approach to Mammillaria. I am hoping to receive constructive feedback and/or criticism.
This is not meant to be an official publication, but rather is a working list, with the goal being eventually a more natural arrangement of the genus Mammillaria than those which have been previously published.
For the various infrageneric groupings higher than species, in a few instances priorities may be violated. I have not fully researched publication dates and synonymies of sections, series, nor species groups, so some names may be incorrect, or may exist but remain unknown to me. I am following most closely Hunt’s sectional and series names in preference to those of Luthy. I believe Hunt’s are more recognizable, and I suspect they are properly and legally applied. As I don’t know the type species of some of the sections and series, I could have assigned names to an improper group of species (but I hope no such goofs were made).
I have listed the majority of recognized names. A few with which I am not very familiar are bound to be incorrectly assigned, but I think I’ve gotten most of them placed where they really belong. The synonymies may seem sweeping, but I think in most cases are reasonable. Even if this “lumping” is not agreed with, it serves to show very close kinship of the plants involved.
A) MAMMILLARIA subgenus MAMMILLARIA –
Seeds brown, roughly teardrop-shaped, with usually linear lateral hilum, with testa cells tabular pitted and bordered by a very strongly undulating raised line; body of plant usually with some milky sap at least toward core of plant. Flower and fruit is generally characteristic, but minor variations are often useful in identifying or grouping plants.
I) sect. Mammillaria
stems generally depressed globose, sometimes hemispheric, with relatively few spines and tubercles (stem usually easily visible through spines), plant body with milky sap to surface.
a) series Mammillaria (= Macrothelae)
1) M. mammillaris group – Includes M. mammillaris (incl. ? ekmanii, simplex , etc.), M. nivosa. Placement of M. nivosa here may be artificial?
2) M. heyderi group – Includes M. coahuilensis (incl. albiarmata), M. heyderi (= applanata, incl. bullingtoniana, gaumeri, hemispherica), M. peninsularis, M. roseoalba (often misapplied to plants of M. magnimamma type).
3) M. uncinata group – Includes M. brandegeei (incl. arida, baxteriana, bellisiana, bocensis, craigii, evermanniana, dawsonii, gabbii, gatesii, glareosa, johnstonii, lewisiana, macdougalii, marksiana, marshalliana, movensis, pacifica, petrophila, sonorensis, tesopacensis, etc.), M. gummifera (incl. crassa, durangicola, grusonii, pachycilindrica, papasquiarensis, tortulospina, wagneriana, zeyeriana, etc.), M. uncinata (incl. boelderliana).
4) M. petterssonii group – Includes M. gigantea (incl. ? antesbergeriana, armatissima, hastifera, ocotillensis, ? saint-pieana), M. orcuttii, M. petterssonii (incl. apozolensis, fulvaflora, hamiltonhoytea, huiguerensis, pilensis, saltensis), M. rubrograndis.
b) series Polyedrae
1) M. magnimamma group – Includes M. carnea (incl. sartorii, tenampensis), M. lloydii, M. magnimamma (incl. angularis, aramberri, bernalensis, bucareliensis, centralifera, centricirrha, compressa, crassimammillis, esseriana, macracantha, priessnitzii, rioverdense, saxicola, ?seitziana, tolimensis, vagaspina, vallensis, zuccariniana, etc.), M. melanocentra(incl. euthele, freudenbergeri, meiacantha, ? parrasensis, runyonii, winterae, zahniana), M. mystax (incl. atroflorens, casoi, crispiseta, erythra, huajuapensis, mixtecensis, ? multiseta, variaculeata, etc.), M. polyedra.
2) M. karwinskiana group – Includes M. karwinskiana (incl. confusa, conzattii, ebenacantha, eichlamii, jozef-bergeri, ?knippeliana [nomen dubium], ? multiseta, neomystax, praelii, strobilina, voburnensis, etc.).
II) sect. Supertextae (= sect. Leucocephalae)
Stems mostlyhemispherical to conical, with many tubercles; often much wool in areoles;
spines usually obscuring plant body (except in M. sempervivi), radials often nearly white, typically with two or four, usually short, often darker centrals, not hooked, one pointing up and one down (but number and arrangement varies some); fruits red, clavate.
a) series Leucocephalae – Stems usually milky to surface. More northern. – Includes M. chionocephala (incl. caerulea, ritteriana), M. geminispina (incl. albata, brauneana, bravoae, giseliana, hahniana, klissingiana, leucocentra, mendeliana, werdermanniana, woodsii ), M. formosa (incl. arroyensis, avila-camachoi, cadereytensis, ? celsiana, infernillensis,
leucocentra, morganiana, microthele, muehlenpfordtii, parkinsonii, perbella, queretarica), M. sempervivi (incl. ?amajacensis, pseudocrucigera), M. standleyi (incl. auricantha, auritricha, bellacantha, canalensis, cobrensis, floresii, hertrichiana, laneusumma, lanisumma, lindsayi, mayensis, miegiana, montensis, nolascana, taylorianum, ? xanthina).
b) series Supertextae – Stems usually milky only in core. More southern. Includes M. columbiana (incl. bogotensis, chiapensis, hennesii, ruestii, soehlemannii, yucatanensis), M. crucigera (incl. buchenaui, grandinosa, huitzilopochtli), M. pseudoperbella, M. supertexta (incl. acultzingensis, albidula, albilanata, collina, conspicua, dealbata, dixanthocentron, ?donatii, dyckiana, elegans, fauxiana, flavicentra, haageana, ignota Repp., igualensis, juxtlahuacensis, kunthii, lanata, lanigera, longicaudata, lupina, martinezii, meissneri, monticola, niduliformis, noureddineana, oaxacana, palmillensis, reppenhagenii, san-angelensis, schmollii, tegelbergiana, vaupelii), M. tlolocii.
III) sect. Subhydrochylus
Plants usually columnar, with numerous tubercles, with sap milky usually only in core, spines usually slender, smooth, and glassy; centrals often (not always) hooked; fruits usually not bright red, but varying shades of green through brown, pink, or purplish, often translucent. The series in this section are weakly defined. Seed is characteristically lighter in color and more strongly sculptured than in previous sections (though this needs further documentation, and the trait varies from series to series in the previous sections).
a) series Elongatae (= Leptocladodia sensu Hunt, minus its type species M. pottsii) – smallish, mostly cylindrical; spines not hooked. Includes M. densispina (incl. buxbaumiana), M. elongata (incl. amgiomea. densa, echinaria, echinata, intertexta, mieheana, rufrocrocea, stella-aurata, straminea, subcrocea, tenuis, var. viperina [not M. viperina]), M. eriacantha, M. microhelia (incl. droegeana sensu Repp., miroheliopsis).
b) series Polyacanthae – larger, tending toward cylindrical (sometimes subglobose), central spines often (not always) hooked. Distinction of species in this series is confused and they could conceivably ALL be the same species! Includes M. guerreronis (incl. aguilensis, aureispina, backebergiana, claviformis, compacticaulis, duoformis, ernestii, erythrocalix, hamata, heeriana, magnifica, matudae, minor, recta, rectiformis, rossiana, subhamata, serpentiformis, virginis, xaltianguensis, xuchiapensis, zapilotensis), M. rekoi (incl. albrechtiana, leptantha, krasuckae, mitlensis, pseudorekoi, pulliamata, rekoiana, sanjuanensis), M. spinosissima (incl. auricoma, bella, centraliplumosa, crassior, gasterantha, gracilis, haasii, kladiwae, ? machucae, meyranii, nunezii, pilcayensis, pitcayensis, tepoxtlana).
c) series Heterochlorae – larger yet, otherwise much like the Polyacanthae, except usually hemispherical to subglobose stems; spines not hooked. Very close to and probably not deserving of a separate series ranking from the Polyacanthae. Includes M. discolor (incl. esperanzaensis, halbingeri, longispina, multispina, ochoterenae, pachyrhiza, schmollii), M. rhodantha (incl. apamensis, aureiceps, bonavitii, clacantha, fera-rubra, fuscata, mccartenii, mollendorffiana, mundtii, parensis, pratensis, pringlei, sulphurea, verticealba, wiesingeri), M. polythele (incl. durispina, hidalgensis, hoffmanniana, kelleriana, kewensis, neophaeacantha, subdurispina, tetracantha, xochipillii).
d) series Decipientes – includes M. decipiens (incl. albescens, camptotricha).
B) MAMMILLARIA subgenus NOV. (= vetula group)
Seeds and fruit like subg. Mamillopsis, but seeds with undulating line along margin of testa cells more as in subg. Mammillaria. Flowers and plants similar to Stylothelae and to M. decipiens. Doesn’t fit neatly anywhere. Includes M. vetula (incl. gracilis, kuentziana, magneticola, pulchella), ? M. schwarzii .
C) MAMMILLARIA subgenus MAMILLOPSIS (= DOLICOTHELE)
Characterized by the soft stems with a “chin” above the areole; fruits very juicy with thin translucent walls, usually rounded; seeds roughly teardrop-shaped, mostly more laterally compressed than in subgenus Mammillaria, brown to nearly black with well developed pits in testa pattern cells, the pattern cells bordered by nearly straight raised lines. In M. beneckei the seeds are very irregular, but still conform to these criteria.
I) sect. Barbatae
Distinguished by shorter flower tube, often fringed sepaloids; petaloids usually not rich yellow, but some other shade.
a) series Barbatae – With fimbriate tepal margins. Includes M. barbata (incl. chavezei, garessii, meridiorosei, morricalii, orestera, santa-clarensis, viridiflora, wilcoxii, etc.), M. wrightii (incl. wolfii).
b) series Carretiiae [ser. nov.] – With tepal margins not noticeably fimbriate. Includes M. carretii (incl. saffordii), M. glassii (incl. as censionis, nominus-dulcis, siberiensis), ? M. oteroi. M. oteroi is perhaps (?) better placed in sect. Dolicothele, but the small flowers with a shorter tube indicate placement here.
II) sect. Dolicothele (incl. Oehmea)
Distinguished by flower mostly fragrant; usually long flower tube; petaloids usually rich yellow. Includes M. baumii, M. beneckei (has similarities in flower to M. senelis), M. longimamma, M. melaleuca, M. sphaerica, M. surculosa.
III) sect. Mamillopsis
Weakly defined from Dolicothele by zygomorphic structure of flower with tube possessing obvious “scales” (i.e. fleshy reduced sepaloids); and by the apparent lack of fragrance in the flower; flower yellow (rare), or near white (rare) to pink or red. Otherwise the same. Includes only M. senilis.
D) MAMMILLARIA subgenus COCHEMIEA
I have at times tentatively split the hummingbird pollinated Cochemiea sensu strictu from the remaining “Bartschella” species, but have never thought this a valid separation at this ranking. The separation would be more of a compromise to tradition. It has always seemed to me that the Ancistracanthae plus Bartschella [sensu strictu] plus Sphacelatae were more closely akin to Cochemiea than the resultant group as a whole is to other Mammillaria series. However, the Lasiacanthae does show some strong similarities to the Ancistracanthae, and provides something of a bridge or link to other groups.
I) sect. Krainzia (= Longiflorae)
a) series Krainzia (= Longiflorae) – Includes M. aureilanata, M. guelzowiana, M. heidiae, M. longiflora (incl. stampferi), ? M. tepexicensis, M. tetrancistra, M. zephyranthoides.
b) series Napinae – Includes M. deherdtiana (incl. dodsonii), ? M. hernandezii, M. napina.
c) series Saboae – Includes M. saboae (incl. goldii, haudeana), ? M. sanchez-mejoradae, M. theresae.
II ) sect. Solisia
Seeds brown to black with basal hilum, with testa cell dividing lines undulate as in subgenus Mammillaria.
a) series Solisia (= Bombycinae & Leptocladodia) – In many ways intermediate to subgenus Cochemiea and subgenus Mammillaria. Seed shape and habitus is most like subgenus Cochemiea. Plants with hard bodies and round-apices on tubercles, often (always?) with traces of milky sap or the internal structures associated with it; flowers characteristic; fruits dimorphic, some green, never lengthening and eventually drying between tubercle bases; some fruits lengthening but usually not exceeding spines, sticky, and hard to remove entire, usually white to pink. Includes M. bombycina (incl. perezdelarosae), M. carmenae (incl. dasyacantha, lauii, subducta), ? M. marcosii, M. mercadensis (incl. berkiana, brachytrichion, fuscohamata, gilensis, guillauminiana, jaliscana, kleiniorum, rettigiana, posseltiana), M. moelleriana (incl. cowperae), M. pectinifera, M. pottsii, M. solisioides. Widespread east from Sierra Madre Occidental, south to Oaxaca.
b) series Stylothelae (proper usage??; what is type species?) – Plants with soft bodies; ??? milky sap, or not ???; flowers more delicate in structure; fruits elongating, mostly shades of green, pink, or purplish, in some (? all) with characteristic odor (similar to many subgenus Mammillaria fruit in smell). Includes M. bocosana (incl. ?aurihamata, eschauzieri, knebeliana, multilanata, splendens), M. erythrosperma (incl. multiformis), M. fittkaui (incl. ?limonensis), M. glochidiata, M. zeilmanniana. This is a tentative, perhaps unnatural grouping. Central Mexico.
III) sect. Lasiacanthae
(incl. most of Crinitae & Proliferae) – Close to the section Cochemiea, but smaller, mostly depressed globose plants with mostly shorter flowers with capitate stigmata. Spines less often hooked, often “pubescent” to “plumose”. Concentrated east of Sierra Madre Occidental and north of Trans-volcanic region.
a) series Lasiacanthae – Plants mostly with body firm, mostly depressed globose; fruits normally bright shades of orange, red, or pink, elongate, firm, with basal pore; seeds roughly lightbulb-shaped (i.e. narrowed at hilum end). Includes M. gasseriana (incl. chica, lengdobleriana, magallanii, neobertrandiana, roseocentra, viescensis), M. herrerae (incl. albiflora), M. lasiacantha (incl. denudata, egregia), M. lenta, ? M. luethyi, M. nazasensis, M. pennispinosa, M. sinistrohamata (incl. zacatecasensis).
b) series Mammilloydia – Plants with body firm; with spines in axillary areoles; fruits spindle-shaped, mostly firm, green to pinkish or lavender (but very soft and bright red in M. ortiz-rubiona; elongate and bright orange-red in M. stella-de-tacubaya); seeds unusual for sbg. Cochemiea, slightly flattened and rather smooth, with hilum relatively small, with pitting often reduced (M. humboldtii) to nearly absent (M. candida & M. ortiz-rubiona). Includes M. candida (incl. estanzuelensis), M. ortiz-rubiona (separated due to distinct fruit, slightly distinct seed, distinct gypsum habitat, and distinct appearance of plant in habitat), M. humboldtii, ? M. plumosa, M. stella-de-tacubaya. Another possibly unnatural grouping, but all except M. plumosa are at least grossly very similar.
c) series Proliferae. Plants soft; spines less densely covering body; fruits mostly red, often not exceeding spines, soft and insubstantial; seeds as in series Lasiacanthae. Includes M. albicoma, M. crinita (incl. aureoviridis, ? aurihamata [all specimens I have seen under this name were M. bocosana with yellow spines], calleana, duwei, erectohamata, felipensis, leucantha, mathildae, moeller-valdeziana, monancistracantha, nana, painteri, purberula, pubispina, M. scheinvariana, schelhasei, ? seideliana, tezontle, trichacantha, variabilis, wildii) M. picta (not very distinct from M. crinita; incl. anniana, aurisaeta, rayonensis, sanluisensis, schieliana, subtilis, unihamata, viereckii, weingartiana), M. prolifera (incl. arachnoidea, ?glomerata, grisea, haitiensis, multiceps, perpusilla, pusilla, texana, zubleri), M. schiedenana (incl. dumetorum, giselae).
IV) sect. Cochemiea
Plants mostly with cylindrical bodies (sometimes other shapes), often clustering; central spines typically hooked; fruit mostly shades of bright orange, red, or pink, usually elongate and slightly clavate, often (? always) with an open basal pore when pulled off plant; seeds grey to black with a basal hilum, in most narrower at hilum end and somewhat lightbulb-shaped, pitted. Concentrated west of Sierra Madre Occidental, and also (Sphacelatae) Oaxaca/Puebla area.
a) series Sphacelatae – Flowers small, usually magenta to purple; fruit and seeds as in series Ancistracanthae and Lasiacanthae; southern distribution. Includes ? M. kraehanbuehill, M. sphacelata (incl. tonalensis), M. viperina.
b) series Bartschella – Plants with soft epidermis, soft, gray-green, usually depressed globose to depressed conical when growing in habitat; flowers with long narrow spreading stigmata; fruits with large basal pore; seeds large, rounded, with obvious secondary sculpturing in shallow pits. Includes M. boolii (incl. insularis), M. schumannii. Baja and Sonora.
c) series Ancistracanthae – Plants mostly relatively small with cylindrical branching stems, but in some depressed globose or geophytic; hooked central spines common, and normal in most species; flowers with long narrow spreading stigmata; fruits with small basal pore; seeds smaller with light bulb shape, and with pits deep and not obviously sculptured. Includes M. albicana (incl. dolorensis, sleviinii), M. armillata (incl. cerralboa), M. blossfeldiana (incl. goodridgei, rectispina, shurliana), M. capensis, M. dioica (incl. angelensis, estabanensis, multidigitata, neopalmeri, phitauiana), M. grahamii (incl. arizonica, auricarpa, microcarpa, milleri, oliviae), M. louisae (distinction from M. mazatlanensis needs further study; plants seem quite different), M. mainiae, M. mazatlanensis (distinction from M. grahamii is dubius; incl. alamensis, albissima, bullardiana, diabloa, gueldemanniana, guirocobensis, hutchisoniana, inae, littoralis, marierana, monocentra, occidentalis, patonii, sheldonii, sinaloensis, swinglei), M. thornberi (incl. fraileana, yaquensis). Wide distribution mostly along Pacific Slope.
d) series Cochemiea – As in Ancistracanthae, except plants usually larger; flowers zygomorphic with longer hypanthium and usually red; stigmata relatively long but often not spreading. Includes M. halei, M.pondii (incl. maritima), M. setispina, M. poselgeri. Baja California.