| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Plant Biology, Cornell University, Ithaca, New York 14853; and Sonoma State University, Rohnert Park, California 94928
Received for publication November 18, 1997. Accepted for publication August 13, 1998.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSSPECIES AND OBSERVATIONSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Altamiranoa • chromosomes • dysploidy • hybrids • Villadia.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSPECIES AND OBSERVATIONSDISCUSSIONLITERATURE CITED |
|---|
, 1905
). At the same times he proposed Altamiranoa for 12 Mexican species, three described as new and nine formerly in Cotyledon, Sedum, or Umbilicus. These are all Sedum-like plants with petals united at least near the base, but Villadia has a narrow thyrse of several few-flowered cincinni, sometimes reduced to a raceme or spike, while Altamiranoa has a broad cyme of a few several-flowered cincinni.
In a generic revision of the family, Berger (1930)
recognized both genera, adding to Altamiranoa eight Peruvian species formerly in Cotyledon or Sedum. He assigned both genera to his subfamily Echeverioideae, together with other American, and especially Mexican, genera having connate petals, but these two and Lenophyllum Rose differ from the rest in having terminal rather than lateral inflorescences. Baehni (1937)
and Clausen (1940)
combined Altamiranoa with Villadia, Clausen making it a section; and most authors since then have used Villadia in this expanded sense.
Authors generally have agreed on the limits of Villadia in the sense of Rose, with its distinctive thyrsoid to spicate inflorescence, but they have disagreed about Altamiranoa, which has a broad cymose inflorescence, as usual in Sedum, and has no clear boundary from Mexican Sedum. In his cladistic tree based on chloroplast DNA, van Ham (1995)
had only a few Mexican species and none of Villadia proper, but he placed Villadia (Altamiranoa) batesii (which we call Sedum goldmanii) closest to S. bourgaei Hemsl. of sect. Fruticisedum. Though it is not remarkably different, we believe that in this family of uncertain generic limits Villadia can best stand as a genus of its own, apart from Sedum. Altamiranoa in the sense of Rose clearly is polyphyletic, and Fröderström (1935)
already returned some species to Sedum; we now believe that all species of Altamiranoa are best dispersed in Sedum (Moran, 1996
). The present chromosome study adds some support to this treatment.
Although we would formally restrict Villadia to Rose's original definition, in this paper we also consider the species of the Altamiranoa group, which we refer to Sedum. Villadia (sensu stricto) has perhaps 25 species, some not yet named. Most are Mexican, but one extends into the Big Bend region of western Texas and another into Guatemala, and two or three others grow in the Andes of Peru. The Altamiranoa group has about seven species in Mexico and others, poorly known, from Peru to Argentina. Chromosome numbers in both groups are very diverse, with a great deal of dysploidy and some polyploidy, and they have been helpful in sorting out the species.
Although natural hybrids among the Mexican Crassulaceae are uncommon, Uhl (1992)
has made hybrids that directly or indirectly interconnect more than 200 species of what we would liberally call about nine genera, including both Villadia and Sedum (with Altamiranoa), showing these species to be part of a gigantic comparium. Study of chromosome pairing in these hybrids gives useful information about similarities in the parental genomes, about ploidy, and about the nature of the changes in their chromosomes during evolution.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSPECIES AND OBSERVATIONSDISCUSSIONLITERATURE CITED |
|---|
|
|
SPECIES AND OBSERVATIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSPECIES AND OBSERVATIONSDISCUSSIONLITERATURE CITED |
|---|
Villadia acuta Moran and C. H. Uhl (1991b)
is herbaceous, with close-set, spreading, sharply acute leaves and a spike of white flowers, the corolla with a long tube and very acute, spreading lobes. It is known only from the type locality in south-central San Luis Potosí. It has 16 pairs of chromosomes (Fig. 1), as does the related V. pringlei of Chihuahua and Durango.
|
|
is a small herb with an open thyrse of pale greenish-yellow flowers having bowl-shaped corollas. It is known from only one plant from a cliff on the western brink of the Sierra Madre Occidental of Durango, near the Sinaloa border. It has 15 pairs of chromosomes (Fig. 3), as do V. albiflora and V. painteri, neither of which seems closely related. A hybrid between V. aperta, as pollen parent, and Sedum greggii Hemsl. (n = 33) had 20 or more unpaired chromosomes at metaphase I of meiosis (Fig. 36); the 15 chromosomes from the Villadia parent are noticeably larger than the small Sedum chromosomes.
Villadia aristata Moran (1990)
has simple stems with close-set ascending leaves and a dense thyrse of white or pinkish flowers, the corolla ellipsoid, with segments rounded and denticulate at the tip and with a conspicuous dorsal spur. It grows in the Sierra Madre Oriental from southeastern Coahuila and southern Nuevo León and Tamaulipas for nearly 250 km south-southwest to the Sierra de Alvarez, east of the city of San Luis Potosí. Seven collections each had ten pairs of chromosomes (Fig. 4), as found also in a very different collection from Querétaro somewhat resembling V. cucullata.
Villadia cucullata Rose of northeastern Mexico has tall often solitary stems with crowded ascending leaves and with a conspicuous dense spike of red to orange-yellow flowers, the corolla ellipsoid, nearly closed, its segments denticulate. In subsp. cucullata (including V. jimulcensis Nesom), which grows mostly within about 75 km of Saltillo in southeastern Coahuila, the corolla segments are rounded and cucullate. Our five collections all had n = 11 (Fig. 5), a number unique to this species.
V. cucullata subsp. apiculata Moran and C. H. Uhl (1998)
differs in its strongly keeled corolla segments, narrowed at the apex, and each with a subdorsal spur. It ranges from southwestern Tamaulipas south to northern Hidalgo. A plant of the type collection from San Luis Potosí and another from Hidalgo both also had n = 11 (Fig. 6). In a plant from Tamaulipas, however, most cells at metaphase I (Fig. 25) had 21 pairs, plus an unpaired chromosome of standard size. Apparently it is a tetraploid that lacks a chromosome (4n - 1).
Villadia sp. aff. cucullata has tuberous roots and is generally similar to V. cucullata but differs in its brownish corolla and in its chromosome number. It has n = 10 (Fig. 7), as in the white-flowered V. aristata. We know it from a single plant from northwestern Querétaro, and we need more information.
Villadia guatemalensis Rose (including V. levis Rose) is a coarse plant, sometimes bushy, with minutely papillate stems and spreading acute leaves, the corolla greenish-yellow to white and often strongly flecked with red, with spreading segments. Nine collections from northern and central Oaxaca and two from southern Guatemala all had n = 21 (Fig. 8). However, two collections from southeast of Oaxaca city had n = 20, with one pair noticeably larger than the others and perhaps corresponding to two pairs in the other collections (Fig. 9). One other collection was meiotically irregular and probably triploid. In a hybrid between the 20-chromosome race and V. nelsonii (also n = 20) metaphase I had 15–20 paired elements and 0–8 unpaired (approximately 15 + 8 in Fig. 37). Hybrids of the same collection of V. guatemalensis with the 26-chromosome race of Sedum greggii (Moran 14730) and with the tetraploid S. (Villadia) grandisepalum (n = 44) had many unpaired chromosomes at meiosis (Uhl, 1994
).
Villadia imbricata Rose is a distinctive low herb with crowded mammillate leaves and a short spike or raceme of small white flowers, the corolla elliptic, of thick segments. Five collections from southern Puebla and northeastern Oaxaca had n = 12 (Fig. 10), as did two from cultivation, one of them with two small extra B-chromosomes (Fig. 11).
Villadia incarum (Ball) Baehni and J. F. Macbr. is a little-known Peruvian species, with terete leaves and a short raceme of yellowish or reddish flowers. We had three collections under this name. At metaphase I n = 88–89, with chromosomes that are variable in size and relatively large for Villadia (Fig. 13).
Villadia laxa Moran and C. H. Uhl (1995)
is a slender herb with papillose-roughened stems and with a raceme or lax thyrse of rosy flowers, the corolla with spreading segments. It is the northwesternmost Villadia, native to the Sierra Madre Occidental of northern Sinaloa and adjacent southern Sonora and southwestern Chihuahua. A plant of the type collection had n = 14, a number found also in the very different V. recurva of Guerrero and Oaxaca.
Villadia minutiflora Rose is a slender herb with papillose stems and bristly-ciliate leaves and sepals, the corolla small, white or flushed with red, with spreading segments. It is known only from the Sierra de Juárez just north of the city of Oaxaca. Plants from two localities had n = 21, a number found also in the very different V. guatemalensis.
Villadia misera (Lindl.) R. T. Clausen (= V. parviflora Rose) has tuberous roots and short, simple stems that die to the ground after flowering, with many ascending leaves and a narrow thyrse of small and inconspicuous flowers, the corolla white, with erect segments. It is widespread in eastern Mexico, typically in barren places. From western San Luis Potosí southeastward for 330 km across the altiplano and the dry western side of the Sierra Nevada Oriental to the west side of the Cofre de Perote volcano in Puebla, 14 collections had n = 33 (as also one other from cultivation), with chromosomes mostly about the same size (Fig. 14). Five more collections from Querétaro and adjoining Hidalgo had n = 42–44 (Fig. 15) and four others from scattered northern localities in San Luis Potosí and Aguascalientes had n = 48–49 (Figs. 16, 17), all with chromosomes differing substantially among themselves in size. These numbers in V. misera may represent tetraploids and hexaploids of some sort, derived from a common ancestor with the rather similar V. painteri, a diploid with n = 15 (Fig. 19), which grows just to the northwest. However, we have no hybrids of V. painteri or V. misera that might shed light on polyploidy here. Further study might support separating these chromosome races taxonomically.
V. sp. aff. misera is known to us from a single collection from extreme southeastern Coahuila. It is rather similar to V. misera, but it came from farther north, and it had far fewer chromosomes, n = 9, the lowest number found in the genus (Fig. 12).
Villadia nelsonii Rose is bushy and papillose, with scattered, broad, uniquely spatulate leaves and with a spike or spicate thyrse of white flowers, the corolla segments spreading (Moran, 1970
). Six collections from central and southern Oaxaca all had n = 20 (Fig. 18). Hybrids with V. albiflora (n = 15) and V. guatemalensis (n = 20) are mentioned above. A hybrid with V. recurva (U1441, n = 14) formed 12–14 paired elements and 0–8 unpaired at metaphase I (13 + 3 in Fig. 38). One with a topotype of Altamiranoa necaxana Fröd., which we consider merely a near-glabrous form of the mostly puberulent Sedum jurgensenii (U1555, n = 23), had 17–19 paired elements and 2–7 unpaired. Intergeneric hybrids of V. nelsonii with Cremnophila nutans (Rose) Rose (n = 33) (ca. 13 paired + 18 unpaired elements in Fig. 39), with Pachyphytum hookeri (Salm-Dyck) A. Berger (n = 32) (ca. 16 + 19 in Fig. 40), with Sedum craigii R. T. Clausen (n = 30) and with S. palmeri S. Watson (n = 34) all showed less chromosome pairing (Uhl, 1994
).
V. painteri Rose is rather similar to V. misera, but the corolla is slightly smaller and has obtuse segments. We refer our plants here tentatively because we have had no live material from the type locality (Barranca de Guadalajara) for comparison. We had six collections, slanting 250 km southeastward across the dry altiplano from south-central Zacatecas, through the northeastern tip of Jalisco, to Guanajuato city. All had n = 15 (Fig. 19), as in the very different V. albiflora and V. aperta. Villadia misera, with n = 33, 42–44, and 48–49, may have evolved from tetraploids and hexaploids related to V. painteri.
Villadia patula Moran and C. H. Uhl (1991a)
has somewhat the habit of V. misera, with many ascending leaves and a slender thyrse. However, the flowers are larger, with an open corolla and with long slender styles; and the plant seems closer to V. guatemalensis. We studied four collections from mountains along the east side of the central altiplano, from the type locality in southwestern San Luis Potosí, for nearly 200 km southeastward across eastern Guanajuato, to northeastern Querétaro. All had n = 13 (Fig. 20), a number unique to this species.
Villadia pringlei Rose, from the Sierra Madre Occidental in Chihuahua and Durango, has a dense spike or thyrse, the corolla white with segments strongly outcurved at the tips. It somewhat resembles V. acuta of southern San Luis Potosí 600 km to the southeast but is smaller and has a more branched inflorescence. A single collection from northwestern Durango had n = 16 (Fig. 21), as in V. acuta.
Villadia recurva Moran, Kimnach, and C. H. Uhl (1998)
is erect or somewhat decumbent and branching, the corolla white or purplish red, campanulate with widely recurving segments. It grows in the Sierra Madre del Sur of eastern Guerrero and western Oaxaca. Three collections had n = 14 (Fig. 23), a number found also in the very different V. laxa of northwestern Mexico. Hybrids of V. recurva with V. albiflora (n = 15) and with V. nelsonii (n = 20) are cited above, and a hybrid with the tetraploid Sedum (once Villadia) grandisepalum R. T. Clausen (n = 44) is mentioned below. Intergeneric hybrids with Graptopetalum fruticosum Moran (n = 31) (ca. 43 elements in Fig. 41), with Pachyphytum hookeri (n = 32) (ca. 36 elements at metaphase I in Fig. 42, 22 elements at metaphase II in Fig. 43), and with Sedum greggii (n = 33) (30 elements in Fig. 44), were all highly irregular at meiosis (Uhl, 1994
). In all these hybrids the Villadia chromosomes were distinguishable by their larger size.
Villadia sp. aff. recurva. In 1953 R. T. Clausen grew a plant from the Sierra Madre del Sur of northwestern Oaxaca, of which we found a good photograph but no specimen and no precise locality or source. We agree with him that it looks like a new species. It has n = 12 (Fig. 24), as in the rather different V. imbricata, but it more resembles V. recurva, which has n = 14.
Villadia squamulosa (S. Watson) Rose is a slender herb with an open spike or narrow thyrse of small rose-colored flowers, the corolla with spreading segments, and with relatively large flabellate nectaries. It is the northernmost species, growing in the mountains of Chihuahua, Durango, Zacatecas, and the Big Bend region of western Texas, where it is the only species native to the United States. Two collections from Big Bend National Park had n = 17 (Fig. 22), a number apparently unique in Villadia.
Sedum species formerly in Altamiranoa
Sedum andinum Ball is a small tufted Peruvian plant with minute hemispheric leaves and a small cyme of red flowers. Two collections supposedly from near the type locality, east of Lima, had many unpaired chromosomes at metaphase I and a somatic number estimated at very close to 100. It is not clear whether the meiotic irregularity was due to a hybrid origin or to odd-ploidy. A plant, possibly a different species, from more than 800 km to the southeast in far southern Peru about 30 km from the border with Chile, was nearly normal at meiosis except for several large elements at metaphase I. Allowing for these presumed multivalents, n = 40 probably (Fig. 26).
Sedum goldmanii (Rose) Moran is a common plant in the mountains of central Mexico, generally on old lava. It is best known as Villadia batesii (Hemsl.) Baehni and J. F. Macbr., but it is highly variable and also includes Altamiranoa alpina Fröd., A. batesii var. subalpina Fröd., A. ramulosa Fröd., and V. mexicana (Schltdl.) H. Jacobsen (Moran, 1997b
). The plant is often much branched from the base and bears a cyme of white or reddish flowers that are unique in having the corolla tube equaling the lobes. We studied 26 collections from throughout the range, from western Michoacán across the state of México, southern Hidalgo, the Federal District and Tlaxcala to southeastern Puebla, including near topotypes of A. ramulosa Fröd., Cotyledon batesii Hemsl., and Umbilicus mexicanus Schltdl. Though one collection was meiotically irregular and probably triploid, the other 24 all had n = 25, a number so far found only in this species.
Sedum grandisepalum R. T. Clausen has small cymes with greenish-yellow petals that are generally shorter than the sepals. After originally describing it in Sedum, Clausen (1959)
later transferred it to Villadia; we now believe it is best kept in Sedum with the other Altamiranoas. It is known only from the type collection, from the eastern Sierra Madre del Sur of southern Oaxaca. It has n = 44 (Fig. 27), but meiosis is slightly irregular, with occasional univalents and probable multivalents.
Uhl (1994)
crossed Sedum grandisepalum (n = 44) with four species considered to be diploids: Villadia recurva (n = 14), V. guatemalensis (n = 20), Sedum alamosanum (n = 18) (ca. 22 paired + 12 unpaired elements in Fig. 45), and S. obcordatum R. T. Clausen (n = 34). As expected in triploid hybrids, each hybrid looked more like S. grandisepalum than like the other parent. Also, each hybrid formed 17–22 paired elements (bivalents and multivalents) at meiosis, with a large number of unpaired chromosomes. We conclude that S. grandisepalum is tetraploid with a basic number of x = 22 and that its 44 chromosomes formed the paired elements in these hybrids. The chromosomes from the diploid parents had no full homologues, and except for a few that occasionally attached to corresponding parts of a grandisepalum chromosome, most remained unpaired.
Supporting this conclusion, when S. grandisepalum (n = 44) was crossed with another tetraploid, Echeveria chapalensis Moran and C. H. Uhl (n = 45), the hybrid was more intermediate in appearance, and at meiosis it formed 33–40 paired and 2–7 unpaired elements. Here again, we believe that most or all the paired elements were of homologous chromosomes of the two sets from the same tetraploid parent. However, the presence of multivalents suggests some pairing between chromosomes of the different parents, and univalents hint that some homologies between corresponding chromosomes from the same parent may be weak. In their chromosome pairing the parents seem not to be typical autotetraploids.
Sedum grandyi Raym.-Hamet of northern Peru has short stems, broadly ovate to suborbicular leaves ca. 5 mm long, and a lax cyme of small flowers with corolla shorter than the calyx. A collection from cultivation, supposedly of this species, generally had a few univalents and multivalents at meiosis, suggesting hybridity or polyploidy; allowing for these, the gametic number is probably 38 (Fig. 28).
Sedum jurgensenii (Hemsl.) Moran, with weak elongate stems and with cymes of white flowers, is distinguished by its puberulent herbage. It has long been known as Villadia elongata (Rose) R. T. Clausen (Moran, 1997a
). Primarily a plant of the wetter eastern slope of the Sierra Madre Oriental, it occurs from Guanajuato and southeastern San Luis Potosí south through northeastern Querétaro and Hidalgo to northern and eastern Puebla, edging into adjoining Veracruz. We studied plants of the typical subspecies from 13 localities throughout the range, including a topotype of V. elongata, and six others from cultivation. All had n = 23 (Fig. 29), a number known only in this species. V. necaxana (Fröd.) H. Jacobsen appears to be just a less puberulent form, and a plant from its type locality also had n = 23 (Fig. 30). Its hybrid with V. nelsonii (n = 20) is mentioned above. Subspecies attenuata Moran (1997a)
differs in its strongly ascending and imbricated, narrower, and more sharply acute leaves. The type collection and two others also had n = 23.
Sedum latifilamentum R. T. Clausen, though never formally assigned to Altamiranoa, clearly belongs with that group, whatever its status. It has tuberous roots and white, globular corollas. The incurved petal tips, separating 2 mm or less at anthesis, the stamens, and the sulcate carpels without styles suggest that it is self-pollinated. We studied the type collection, from northern Hidalgo, and another plant, from northeastern Querétaro. Both had n = 20 (Fig. 34), a number found also in Villadia guatemalensis and V. nelsonii and in an unnamed plant from Oaxaca. A hybrid with Sedum alamosanum (n = 18) had little or no chromosome pairing at meiosis (38 univalents in Fig. 46), and one with Graptopetalum fruticosum (n = 31) appeared authentic but never flowered in 10 yr (Uhl, 1994
).
Sedum scopulinum (Rose) Moran is a rare plant of southern Puebla, whose creeping stem bears persistent enlarged white leaf bases and a small cyme of white flowers. A single plant had n = 28.
Sedum sp. of central Oaxaca has obtuse leaves, and the pistils are oblong and rather broad, narrowing abruptly to the slender styles. Three collections had n = 20 (Fig. 32), as in S. latifilamentum of Hidalgo and Querétaro.
Sedum sp. from high on the Nevado de Colima in western Mexico somewhat resembles S. goldmanii but differs in its much shorter corolla tube and in other ways, and it has n = 29 (Fig. 33), a number known only from this collection.
Sedum sp. An unidentified collection from extreme northern Peru, like most of our other collections from South America, was irregular at meiosis. It had about 50 elements at metaphase I and many chromosome bridges and laggards at anaphase I; it appears to be a hybrid.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSPECIES AND OBSERVATIONSDISCUSSIONLITERATURE CITED |
|---|
) and Mexican Sedum (Uhl, 1985
) diploid and tetraploid numbers intergrade or overlap, with no sharp boundary. DNA studies are needed to show whether Villadia is monophyletic or whether some species may have evolved independently from different Sedum ancestors. We consider more primitive such species as V. nelsonii (n = 20) and V. guatemalensis (n = 20, 21), which are more branched and often woody at the base, with thyrsoid inflorescences, larger flowers, short corolla tubes and spreading segments, and long styles; and we note that these species generally have more but smaller chromosomes. On the other hand, we take as more specialized such species as V. aristata (n = 10) and V. cucullata (n = 11), which are less branched and often die to the base after flowering, with narrower, more spicate inflorescences, smaller flowers, longer corolla tubes and more erect segments and short and abruptly out-turned styles. These species generally have fewer but larger chromosomes. At meiosis in diploid intergeneric hybrids the chromosomes of the Villadia parent can often be distinguished by their larger size from those of the other parent (Figs. 39–43).
As with Villadia, apparently so also with species of Sedum formerly in Altamiranoa: our few hybrids with diploid species again seem to show that the species with low numbers (n = 20, 23) are effectively diploid. However, Sedum grandisepalum (n = 44) is tetraploid, and in all its hybrids most of the corresponding chromosomes of its two sets apparently pair preferentially with each other.
Hybrids of Villadia with diploid species generally have multivalents at meiosis, showing that parts of a chromosome from one parent are homologous with parts of two or more chromosomes from the other. Also, most hybrids have chromosome bridges, showing they are heterozygous for inversions. In hybrids of species with n = 20 or lower the number of paired elements (bivalents and multivalents) at meiosis approaches or equals the number of chromosomes from the parent having the smaller number, and it never exceeds this. As in Echeveria (Uhl, 1992
) and Mexican Sedum (Uhl, 1985
), we believe this shows that pairing in these hybrids occurs only between chromosomes from the two parents, which means that both parents are effectively diploid.
We believe that over time a series of translocations has rearranged the ancestral genome of Villadia into progressively fewer but larger chromosomes, from n = 21, or probably even more, down to n = 9. Thus Villadia, like Echeveria (Uhl, 1992
), has come to have a long descending series of dysploid chromosome numbers. Both in Villadia and in Echeveria the series of consecutive dysploid numbers is much longer than has been reported elsewhere (Grant, 1975
) and the numbers start much higher; but note that species with all numbers in the series behave cytologically as effective diploids.
In Sedum species formerly in Altamiranoa we have found gametic chromosome numbers of 20, 23, 25, 28, 29, and several higher, evidently polyploids. The species appear most similar, morphologically and cytologically, to Sedum section Fruticisedum Berger (n = 26–31, Uhl, 1980
), from which they are probably derived. As in Villadia, they are dysploid, but their generally higher chromosome numbers suggest they have changed less from their ancestors.
These observations suggest something of how Villadia and relatives may have evolved: (1) Villadia and especially the Altamiranoa group of Sedum appear closest to subshrubby species of Sedum section Fruticisedum. Those species have gametic chromosome numbers of 26 to 31, and hybrids of six of them show that none of their chromosomes can pair with each other and thus that they are effectively diploid (Uhl, 1980
). (Although these high numbers appear to be diploid now, they may have arisen long ago as polyploids from unknown diploid ancestors, but there is no direct evidence for this.) (2) These plants generally grow in pioneer habitats such as cliffs and lava, often as small and ephemeral populations, allowing opportunity for new forms to establish themselves locally. (3) Over time many translocations and inversions have rearranged the basic genome into ever fewer but larger units, as Uhl (1992)
suggested for the long descending dysploid chromosome series in Echeveria. (4) These changes would restrict genetic interchange between plants with the original and those with the new karyotype, helping to isolate them reproductively from each other. (5) More genes become linked together, but into fewer groups; those in new chromosomal neighborhoods are likely to be activated at different times during development, which could affect the appearance of the plant. (6) With generally fewer and larger chromosomes, Villadia seems to have changed more from its supposed Sedum ancestors than have species of Sedum of the Altamiranoa group.
Although we know of no natural hybrids of Sedum species with Villadia, there are three probable natural hybrids with other Sedum species formerly in Altamiranoa. (1) Sedum (XSedadia) amecamecanum is apparently a hybrid between Sedum praealtum DC (n = 34) and Sedum goldmanii (= Villadia = Altamiranoa batesii) (n = 25) (Clausen, 1959
; Uhl, 1978
). It has been found in nature only with its putative parents, and it is meiotically irregular and sterile. (2) For two localities near Tlaxcala Clausen (1959)
reported sterile hybrids intermediate between Sedum quevae Raym.-Hamet (n = 20) and what he called Villadia scopulina Rose (= Sedum scopulinum) but what evidently was S. goldmanii (n = 25). (3) A third probable hybrid (UC54.170) was grown by Eric Walther, who said he collected it at 4250 m on Nevado de Toluca. Distributed by the International Succulent Institute (number 151) as Villadia batesii, it has caused much confusion in Europe regarding that common species (R. Stephenson, in litt.). Clausen (1959)
thought it a possible hybrid between Sedum moranense and V. batesii (= S. goldmanii). This parentage is improbable, because S. goldmanii is unknown in the area. However, hybrid status is supported by its extreme irregularity at meiosis (24–27 paired elements plus 3–14 univalents in eight cells analyzed) and by its abortive pollen (Uhl, 1983
).
Despite the rarity of natural hybrids, Uhl (1994)
has made 15 different hybrids by crossing eight species of Villadia or Sedum (formerly in Altamiranoa) in various combinations with nine species of other Mexican genera (Cremnophila, Echeveria, Graptopetalum, Pachyphytum, and Sedum). Species he crossed were V. albiflora (n = 15), V. aperta (n = 15), V. guatemalensis (n = 20), V. nelsonii (n = 20), and V. recurva (n = 14), and S. jurgensenii (n = 23) and S. grandisepalum (n = 44) as well as S. latifilamentum (n = 20). He also made four hybrids between different species of Villadia and three between species of Villadia and species of Sedum formerly in Altamiranoa. The chromosome number of the second parent ranged from n = 18 to n = 34, and for each hybrid the two parents had different chromosome numbers. We cite the hybrids above in the discussions of their parental species. Doubtless with a little patience and luck, many other species of Villadia could be interconnected.
In most intergeneric hybrids of Villadia fewer chromosomes paired than in most other intergeneric hybrids of the Mexican Crassulaceae (Uhl, 1994
), perhaps suggesting a more distant relationship. In each intergeneric hybrid the Villadia had fewer chromosomes than the other parent, and often these were noticeably larger and could therefore be distinguished at metaphase I in the hybrids studied (e.g., Figs. 39–43).
The South American species are poorly known, and our sample of them is very small. Three that we studied, Sedum ?grandyi (n = 38, probably), S. andinum (n = 40–48) and S. sp. (n = ca. 50), have been placed in Altamiranoa, but we reassign them to Sedum, where they were originally named. Their chromosome numbers suggest that they are polyploid, but they are more irregular at meiosis than polyploids of Mexican genera (Uhl, 1992
), and our plants may have been hybrids or possibly odd-ploids. V. incarum, with a racemose inflorescence, is perhaps a true Villadia. Plants under this name have n = 88–89, with normal meiosis.
Geographically and cytologically, the Crassulaceae of South America (aside from Crassula) fall into two groups, with little or no overlap. We have no records of Villadia or Sedum from Venezuela, Colombia, or Ecuador, but more than 60 collections of Echeveria from these countries had n = 21 or n = 22, with also a triploid and several tetraploids based on these numbers (Uhl, 1993
, and unpublished data). These echeverias from northern South America have an almost continuous range of forms, and delimitation of species is difficult. They give a clear impression of a single group that has recently arrived and is evolving rapidly to occupy a variety of ecological niches.
Our most northerly records of Villadia and Sedum in South America are from northern Peru. Our small sample of these plants appears to be entirely polyploid, with gametic numbers from 38, probably, to 89, and most were irregular at meiosis. Our collections of Echeveria from these countries likewise were all polyploid. Two tetraploids with n = 44 from extreme northern Peru resembled the species from farther north. However, all 25 collections of Echeveria from farther south in Peru and from Bolivia and northern Argentina appeared quite different from the northern plants, and they had n = 50 or more (Uhl, unpublished data). In fact, 15 of these collections had gametic chromosome numbers from ca. 120 to 260. Many also were in some degree irregular at meiosis.
Villadia, Sedum (formerly Altamiranoa), and Echeveria are all clearly centered in Mexico and must have originated there. We suppose that Crassulaceae may have entered South America at least twice from the north. The first immigrants are represented now by the polyploids of Echeveria, Villadia, and Sedum that occur from Peru southward. Much more recently a single species of Echeveria must have entered South America from the north and have evolved rapidly in Venezuela, Colombia, and Ecuador; but apparently no villadias or sedums came with it.
|
|
|
|
|
FOOTNOTES |
|---|
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSSPECIES AND OBSERVATIONSDISCUSSIONLITERATURE CITED |
|---|
Berger, A. 1930 Crassulaceae. In A. Engler: Die Natürlichen Pflanzenfamilien, ed. 2, 18a: 352–485.
Britton, N. L., and J. N. Rose. 1903 New or noteworthy North American Crassulaceae. Bulletin of the New York Botanical Garden 3 (9): 1–45.
———, and ———. 1905 Crassulaceae. In North American Flora 22: 7–74.
Clausen, R. T. 1940 Studies in the Crassulaceae: Villadia, Altamiranoa, and Thompsonella. Bulletin of the Torrey Botanical Club 67: 195–198.
———. 1959 Sedum of the Trans-Mexican Volcanic Belt. Cornell University Press, Ithaca, NY.
Fröderström, H. 1935 The genus Sedum L., part IV. Acta Horti Gotoburgensis 10. Appendix.
Grant, V. 1975 Genetics of flowering plants. Columbia University Press, New York, NY.
Moran, R. 1970 Villadia nelsonii Rose. Cactus and Succulent Journal (U. S.) 43: 84–87.
———. 1990 Villadia aristata (Crassulaceae), a new species from northeast Mexico. Cactus and Succulent Journal (U. S.) 62: 177–182.
———. 1996 Altamiranoa into Sedum. Haseltonia 4: 46.
———. 1997a Sedum jurgensenii (Hemsl.) Moran (Crassulaceae) Haseltonia 5: 72–76.
———. 1997b Sedum goldmanii (Rose) Moran (Crassulaceae). Haseltonia 5: 86–93.
———, M. Kimnach, and C. H. Uhl. 1998 Villadia recurva, a a new species from southern Mexico (Crassulaceae). Cactáceas y Suculentas Mexicanas 43: 3–6.
———, and C. H. Uhl. 1991a Villadia patula (Crassulaceae), una nueva especie del centro de México. Cactáceas y Suculentas Mexicanas 36: 27–30.
———, and ———. 1991b Villadia acuta (Crassulaceae): a new species from San Luis Potosí, Mexico. Cactus and Succulent Journal (U. S.) 63: 200–202.
———, and ———. 1992 Villadia aperta (Crassulaceae), una nueva especie del oeste de Durango. Cactáceas y Suculentas Mexicanas 37: 31–34.
———, and ———. 1995 Villadia laxa, a new species from northwestern Mexico. Cactáceas y Suculentas Mexicanas 40: 31–33.
———, and ———. 1998 Villadia cucullata Rose and the new subsp. apiculata (Crassulaceae). Cactus and Succulent Journal (U. S.) 70: 201–202.
Uhl, C. H. 1978 Chromosomes of Mexican Sedum II. Section Pachysedum. Rhodora 80: 491–512.[ISI]
———. 1980 Chromosomes of Mexican Sedum III. Sections Centripetalia, Fruticisedum, and other woody species Rhodora 82: 377–402.[ISI]
———. 1983 Chromosomes of Mexican Sedum IV. Heteroploidy in Sedum moranense. Rhodora 85: 243–252.
———. 1985 Chromosomes of Mexican Sedum V. Section Sedum and subgenus Sulcus. Rhodora 87: 381–423.[ISI]
———. 1992 Polyploidy, dysploidy, and chromosome pairing in Echeveria (Crassulaceae) and its hybrids. American Journal of Botany 79: 556–566.[CrossRef][ISI]
———. 1993 Notes on Echeveria in Venezuela. IV. Variant forms. Cactus and Succulent Journal (U. S.) 65: 81–87.
———. 1994 Intergeneric hybrids in the Mexican Crassulaceae IV. Villadia. Cactus and Succulent Journal (U. S.) 66: 214–217.
van Ham, R. C. H. J. 1995 Phylogenetic relationships in the Crassulaceae inferred from chloroplast DNA variation. In H. 't Hart and U. Eggli, Evolution and systematics of the Crassulaceae, 16–29. Backhuys, Leiden.
This article has been cited by other articles:
|
|
 |
|
  P. J. De Lange, P. B. Heenan, D. J. Keeling, B. G. Murray, R. Smissen, and W. R. Sykes Biosystematics and Conservation: A Case Study with Two Enigmatic and Uncommon Species of Crassula from New Zealand Ann. Bot., April 1, 2008; 101(6): 881 - 899. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
