3. SEDUM DENDROIDEUM (figs. 15-20)

Bright green, spatulate leaves and paniculate cymes of yellow flowers on axillary branches distinguish Sedum dendroidewn from most other species of Sedum in the Trans-Mexican Volcanic Belt. The plants are shrubby, usually pendent from cliffs or epiphytic. The stems surpass in length those of all other species which I have studied. Sometimes they are 5 in. long. The leaves arise spirally. The leaves of S. obcordatum and S. stahlii are opposite. Larger vege­tative and floral parts separate S. dendroideum from S. confusum, S. luteoviride, and other yellow-flowered species. S. aoikon, with leaves similar in size to S. dendroideum, has shorter sepals.

Several authors have reported that Sedum dendroideum (including S. praealtum) has medicinal properties. Tetzmitl, crudely illustrated in the Badianus manuscript, the Aztec herbal of 1552 (Emmart, 1940), very likely is this species. Its medicinal value probably was known long before the Spanish conquest. In the Badianus manu­script, the stems are mentioned as being used in a lotion for swollen eyes, fevers, swellings about bruised veins, and burns. According to the edition of Hernández, 1651, by Ximénez (1888, p. 135), Tetzmitl de Tonalla possesses similar properties, as well as addi­tional ones, but the description of the leaves as long and narrow is inappropriate for S. dendroideum. The reported distribution in the warm regions of the lower Mixteca needs to be confirmed, though that is a reasonable possibility. Standley (1922) stated that the juice of S. dendroideum is astringent and is used for hardening the gums and for hemorrhoids, chilblains, and dysentery. He also re­ported the belief that if applied to the forehead, it stops nosebleed. Martinez (1944) cited the use of the crushed plant for burns, for washing the gums, and for scurvy. In addition, he mentioned that the juice of the leaves, applied to the eyes, cures irritation. Like­wise, Standley and Steyermark (1946) stated that about Coban, in Guatemala, the juice of the leaves is employed for treating in­flammation of the eyes and mouth. Palarea (1954) has reported the complete recession of cataracts in the eyes of three dogs and of three women after application of the juice from leaves of plants from Guatemala which he called "S. praealtum." According to Thomas and Ranson (1954), metabolism in plants which they listed as "S. praealtum" is of the acid type characteristic of other Crassulaceae, in which respiratory C02 is consumed in acid syn­thesis in the hours of darkness.

Description. Materials: 19 plants—3 from Rio Blanco, Vera Cruz; 1 from Maltrata, Vera Cruz; 8 from Malinche; 4 from Iztaccihuatl; 1 from Popocatepetl; 1 from San Jose Chiltepec, Oaxaca; and 1 of unknown origin in the wild. Plants marked be­low as cult, were grown in the greenhouse at Ithaca, N.Y. Meas­urements marked pr. are from flowers preserved in alcohol. Those marked cf. are of plants cultivated in an experiment in a cold frame at Ithaca.

Plants of Sedum dendroideum are shrubs with stems erect, pendu­lous, or prostrate, .6-5 m. long, with bark of older portions brown or gray-brown. In the greenhouse, the stems are erect. Occasionally stems are cristate. The following heights apply to nonflowering plants after a year of cultivation. Differences among populations are not significant.

Leaves are elliptic-oblanceolate, spatulate. or obovate, sessile or petiolate, rounded, obtuse, or apiculate, lustrous, sometimes either red or dotted with red on margins. Marginal glands, either red or dark green, occur only on the leaves of plants from Maltrata and San Jose Chiltepec.

Differences in length of leaves among populations in the wild are not significant, but in cultivation highly significant differences exist between plants from Rio Blanco and Malinche and between plants from Malinche and Iztaccihuatl. The difference between dried and fresh leaves is significant.

Differences in width of leaves among populations in the wild are insignificant, but in cultivation the differences between plants from Rio Blanco and Malinche and between plants from Malinche and Iztaccihuatl are highly significant. The difference between fresh and dried leaves is highly significant.

Differences in regression of length per millimeter of width of eaves among populations and between fresh and dried leaves are lot significant.

Differences in thickness of leaves among populations in the wild are not significant, but in cultivation the difference is highly sig­nificant between plants from Rio Blanco and Malinche. Also highly significant is the difference between fresh and dried leaves. The plants from Maltrata and San Jose Chiltepec have the thickest leaves.

Inflorescences are paniculate cymes of 7-38 cincinni at apices of axillary branches.

Flowers are sessile or on very short pedicels, rarely 4- or 6-merous; one flower of a plant on Iztaccihuatl had 7 petals and 10 pistils.

Sepals are ovate, lanceolate, or elliptic-oblong, sometimes min­utely spurred, obtuse, green, unequal, usually separate, but con­nate for .2-.3 mm. in preserved flowers of a cultivated plant of unknown origin. The plant with connate sepals is similar vege-tatively to plants from Rio Blanco. Dimensions are of the longest sepals of each flower.

Petals are lanceolate, acute or obtuse, mucronate-appendaged, yellow, green on dorsal keel above middle.

Stamens have yellow anthers.

Nectaries are subquadrate or reniform, truncate, broadly rounded or emarginate, pale yellow, yellowish white, or trans­lucent.

Pistils are erect or divergent and are yellow.

Follicles are brown, widely divergent, with lips of ventral mar­gins . i -.2 mm. wide.

Seeds are brown, elliptic-pyriform, finely verrucose.

Variation. Visits to populations of Sedum dendroideum in 1955 were mostly at the wrong season for flowers. Further, only two plants from the Trans-Mexican Volcanic Belt, propagated at Ithaca, have flowered. As a result, data for an appraisal of varia­tion are sparse. Although the four populations studied in 1955 differ from each other in some respects, testing 15 characters of plants under natural conditions does not reveal significant differ­ences. In the greenhouse, differences among samples from these same populations are highly significant in length, width, and thickness of leaves. Within populations, plants differed in the field in length and width of leaves and length of follicles and in culti­vation in length and width of leaves, diameter of flowers, and length of both sepals and petals.

On the basis of the shrinkage, noted in the preceding tables, of ten leaves collected from a plant on the western slope of Iztaccihuatl, measured when fresh on Sept. 7, 1955, and again in dried condition on April 21, 1957, herbarium specimens are unreliable for information about the dimensions of leaves.

The four populations studied in the field in 1955 differ ecologi­cally. Some of the differences are indicated in the following table:

The plants at Rio Blanco are in the humid, tropical zone. They are shaded for part of the day. Those at Maltrata are exposed to the sun through most of the day. The plants on Malinche and Iztaccihuatl are in the zone of firs, Abies religiosa, where frosts are of frequent occurrence. Since different specific names are available for three of these populations and since the fourth is as different from the others as these are from each other and also since differ­ences in length, width, and thickness of leaves are highly significant when plants from these populations are grown under similar con­ditions, some taxonomic recognition is reasonable. The popula­tions are too much alike to be different species, but too dissimilar to belong to the same subspecies. Accordingly, they are included in four different subspecies which are isolated geographically and ecologically.

Fig. 15. Leaves of four subspecies of Sedum dendroideum from plants cultivated in the greenhouse, Ithaca, N.Y. From left to right: ssp. dendroideum from Maltrata, ssp. praealtum from Rio Blanco, ssp. parvifolium from Malinche, and ssp. monticola from Iztaccihuatl, April 22, 1957 (ca. x 1.4).


A. Leaves spatulate, broadly rounded at apex, cuneate and petiolate basally, averaging 4 mm. thick, 41 mm. long, and 17 mm. wide, with margins having subepidermal glands which appear as reddish

or dark green dots .......... ssp. dendroideum, p. 69

AA. Leaves oblanceolate and obscurely petiolate or oblong-elliptical and sessile, averaging 2-3 mm. thick, 40-91 mm. long, and 16-30 mm. wide, with margins lacking subepidermal glands .......... B

B. Leaves oblong-elliptical and sessile, averaging 46-67 mm long and 16-22 mm. wide; petals averaging 2-2.2 mm. wide .......... ssp. praealtum. p. 70

BB. Leaves oblanceolate and obscurely petiolate, averaging 40 91 mm. long and 18-30 mm. wide; petals averaging 2.5-3.6 mm. wide .......... C

C. Leaves averaging 40-52 mm. long and 18-18 mm. wide; sepals averaging 3 mm. long .......... ssp. parvifolium. p. 72

CC. Leaves averaging 65-91 mm. long and 24-30 mm. wide; sepals averaging 8 mm. long .......... ssp. monticola, p. 73

Sedum dendroideum ssp. dendroideum

Nomenclature. Sedum dendroideum A.P. DC, Mem. Fam. Crass., p. 37, pl. 9 (1828). The brief, original description is based on drawings, reproduced as plate 9, taken from Mociño's un­published Flora of Mexico. The distinctive shape of the leaves, mentioned in the description as obovate-cuneate and shown in the illustration, makes reasonably certain the application of this name. Whether or not the plant was preserved from which the drawings were made is uncertain. Neither is the locality known from which this plant was originally collected. In the absence of a type, de Candolle's plate 9 should serve for typification.

The distinctive features of ssp. dendroideum are the thick, spatulate leaves with the margins having subepidermal glands which may appear as reddish or dark green dots. No other subspecies of Sedum dendroideum has this condition. Plants belonging to ssp. dendroideum, included in the above description, are from Maltrata and San Jose Chiltepec. The plant from San Jose Chiltepec flowered in the greenhouse in Ithaca in December, 1956.

Distribution. Subspecies dendroideum occurs in the Central American Volcanic Upland of Guatemala and in the Sierra Madre del Sur of Mexico. It occurs at least as far north as the southeastern base of Mt. Orizaba at Maltrata, Vera Cruz. The description of Sedum praealtum by Standley and Steyermark (1946, p. 413) indicates that plants which they had seen are S. dendroideum ssp. dendroideum. Further, herbarium specimens collected by them and by E. W. Nelson from five localities in Guatemala are this subspecies. In Mexico, it is known from Maltrata and from Tlaxiaco and San Jose Chiltepec in Oaxaca. the two latter populations discovered by T. MacDougall.

Fig. 16. Known distribution in the Trans-Mexican Volcanic Belt and adjacent regions of the subspecies of Sedum dendroideum: ssp. dendroideum (•), ssp. praealtum (+), ssp. parvifolium (o), and ssp. monticola (x).

At Maltrata, I found only three plants on Nov. 6, 1955. These were in an area of 10 sq. m., at an altitude of 1,780 m., in the midst of populations of Sedum lucidum and S. stahlii. The plants were on a slope facing southeast and, except in the late afternoon, were exposed to the full amount of sunlight. The largest plant was 7.5 dm. high, with a trunk 3 cm. in diameter. The rocks at the site were limestone; drainage was excessive; and pH was 7.2. Be­sides the other two species of Sedum, commonest competitors were composites, mints, a species of grass, and a Tillandsia. At the time of my visit, the S. dendroideum had tiny floral buds. Although stems with buds were sent to Ithaca, transplanting was too hard on the cuttings and no flowers developed. In the garden at Ithaca, a plant from Maltrata was more sensitive to freezing than were plants of the other subspecies, although it did survive the first, light frosts.

S. dendroideum ssp. praealtum, see there.

S. dendroideum ssp. parvifolium, see there.

S. dendroideum ssp. monticola, see there.

Relationships. The species most closely related to Srdum dendroideum are S. aoikon, S. confusum, and S. cremnophila. These all appear to be more specialized than S. dendroideum. Sedum cremnophila is native in the Trans-Mexican Volcanic Belt, S. aoikon may be native, but S. confusum is known only from cultivation.

Sedum aoikon is similar to ssp. praealtum of S. dendroideum in the narrowness of its petals, but differs in dimensions of leaves and sepals. My samples of both are too small for a satisfactory com­parison. On the basis of two plants of S. aoikon and three plains of S. dendroideum ssp. praealtum cultivated in the greenhouse, differ­ences in average width and thickness of leaves are significant, respectively 14 mm. and 1.8 mm. in S. aoikon and 20 mm. and 3 mm. in S. dendroideum ssp. praealtum. Likewise, S. aoikon resembles S. dendroideum ssp. parvijolium in length of sepals and in width and thickness of leaves, but in each of these characteristics is smaller. Sedum confusum appears like a small edition of S. dendroideum. It is easily separated from ssp. parvijolium, the subspecies which it most resembles, by the following differences in dimensions of leaves, which are based on plants cultivated under similar conditions:

In addition, the petals of Sedum confusum are significantly shorter than those of S. dendroideum, and the inflorescences have fewer flowers (x = 24).

Sedum cremnophila obviously differs from S. dendroideum in its thick, dark green leaves, long, narrow inflorescences, and smaller flowers. Of the subspecies of S. dendroideum, ssp. monlicola appears most similar to S. cremnophila. Both have long sepals, large leaves, and large nectaries, but they are easily distinguished by the follow­ing differences:

The data for leaves are from plants cultivated under similar conditions in the greenhouse. The data for petals are from plants in the wild.

Sedum aoikon, S. confusum, and S. cremnophila all differ from S. dendroideum in having smaller flowers. Sedum aoikon and S. confusum also have shorter anthers than do either S. dendroideum or S. cremno­phila. Further, the sepals of S. aoikon are minutely connate as in some flowers of S. dendroideum ssp. praealtum, to which it appears closely related, and its inflorescences appear condensed. The leaves of S. confusum are much smaller than those of S. dendroideum, but the leaves of the other two species are as large or larger. In considering these four species phylogenetically, S. dendroideum, with its large leaves and flowers, appears to be the least specialized of the group, S. aoikon and S. confusum intermediate, and S. cremnophila, with its thick leaves and adjustment to a semiarid environment, plus its seemingly condensed, elongate inflorescences, the most specialized.

The superficial differences between Sedum dendroideum and S. botterii have already been considered in the discussion of the latter species. The decision as to which of these is the least specialized is difficult to make. Both species have wide, disjunct ranges. Both are variable, but S. dendroideum has become adjusted to a larger variety of environmental niches. Some of its populations regularly endure freezing; others have become adjusted to semiarid condi­tions. Possibly S. dendroideum is the more specialized of the two species, but comparative anatomical studies still are necessary and, of course, fossils, if they could be found, would be very helpful.

An unexpected relationship of Sedum dendroideum is with Villadia batesii. Though my evidence is all circumstantial, it supports the idea that these two species have hybridized in nature and produced plants of the sort which Praeger named S. amecamecanum, based on a collection of Carl Purpus in 1906, his no. 108, from Mt. Iztaccihuatl. I have studied an isotype (US), and I have also had in cultivation plants, distributed in the horticultural trade, which could be parts of the original collection. The three specimens of the isotype are without flowers or fruits. They are labeled as from "rocks near and above timberline." Praeger listed the type locality as Amecameca. Possibly he had this information from Dr. Rose, who sent the specimen to him. In any event, the conclusion from the data on the herbarium label and the published informal ion is that the type locality is on the side of Iztaccihuatl toward Ame­cameca. Praeger described the type as having pale yellow (lowers with nectaries whitish basally and deep orange above.

On Sept. 4, 1955, in walking from La Joya, on the southwestern side of Iztaccihuatl, to Amecameca, I saw about 20 plants of Sedum amecamecanum on the northern bank and at the base of cliffs in a large ravine tributary to the Milpulco Valley, at an elevational range of 3,430 to 3,450 m. above sea level. The plants were in bud. One appeared very similar to the S. amecamecanum which I had cultivated in Ithaca. Other Crassulaceae nearby were S. dendroideum and Villadia batesii. On Dec. 18, 1955, some plants of S. amecamecanum were at the height of flowering, but most were with­out flowers. Sedum dendroideum and Villadia batesii also were in flower. Distances between plants of S. amecamecanum and S. dendroideum were >5 cm., and between S. amecamecanum and Villadia batesii >4 dm. Plants of S. amecamecanum were both in semishade and in the open exposed to the south, south-southeast, and southwest. Altogether I counted 50 plants in an area of 2,000 sq. m. The site is on the north side of the brook, just northwest of a spectacular rock known to natives as the Finger. The pH of the soil about the roots of S. amecamecanum ranged from 6.6 to 6.8. Drainage was good. Although small plants were present, all could have devel­oped vegetatively. I found no plants which positively were seed­lings. Some plants of S. amecamecanum appeared so similar to Villadia batesii that close examination was necessary to see the differences. Because time on Iztaccihuatl was precious, I devoted primary attention to S. amecamecanum. I made detailed measure­ments and notes on 10 plants of that, 2 plants of S. dendroideum, and 2 plants of Villadia batesii. The results, based on plants in nature, are summarized in the following table. The intermediate status of S. amecamecanum is obvious.

Sedum amecamecanum differs highly significantly from both S. dendroideum and Villadia batesii in six of the characters studied. In each of these characters it is intermediate between the other two species. In addition, a highly significant difference between S. amnamecanum and V. batesii exists in width of petals and a significant difference in length of sepals, but the differences in these characters (figures underlined in foregoing table) between S. amecamecanum and S. dendroideum are not significant.

At Ithaca, I have cultivated Sedum amecamecanum for at least seventeen years and regarded it as a valid species, though plants rarely flowered and, when they did, the flowers appeared abnormal. When I first saw plants in the field on Sept. 4, 1955, I still believed that S. amecamecanum was an undoubted species. On Dec. 18. 1955, however, when the plants were in flower, the relationship with Villadia batesii was apparent. In the greenhouse, I have grown all plants studied in the field. Seven have flowered. Since the S. den­droideum has not flowered, a comparison of floral parts of plants cultivated under similar conditions is not possible, but the follow­ing comparison of leaves shows that some of the differences be­tween the two parents and the hybrids are maintained in cultivation. Although differences among plants of .9. dendroideum appear to be less in cultivation than in nature, differences in dimensions of leaves among plants of S. amecamecanum are greater in cultivation.

Fig. 18. Plant of Sedum dendroideum ssp. monticola x Villadia batesii from Mt. Iztaccihuatl, cultivated at Ithaca, N.Y. A. Habit sketch (x .4). B. Flower from above (1.6). C. Flower from side (x 1.6). D. Two petals and stamens (x 2.4). E. Pistils (x 2.4). F. Single pistil (x 3.2). G. Nectary (x 4).

Fig. 19. Plants cultivated in the greenhouse, Ithaca, X.Y., from left to right: Sedum dendroideum ssp. monticola from Iztaccihuatl, Sedum dendroideum ssp. monticola x Villadia batesii from Iztaccihuatl, and Villadia batesii from the Sierra de las Cruces, May 31, 1956 (x .2).

Fig. 20. Flowers of Sedum dendroideum ssp. monticola x Villadia batesii (lclij and Villadia batesii (right) from plants cultivated in the greenhouse, Ithaca, N.Y., May 31, 1956 (x 4).

Three of the floral characters of greatest importance in distin­guishing Sedum amecamecanum from S. dendroideum and Villadia batesii were studied on plants in the greenhouse. The plants in cultivation showed less variability than those in nature. The expressions of two of the characters in the greenhouse were highly significantly different from the expressions in nature. The data follow:

In nature, most anthers of flowers of Sedum amecamecanum ap­peared undeveloped; likewise the ovules appeared abortive. In flowers of plants in cultivation, many pollen grains are small and undeveloped, but the ovules appear well filled out, though they do not develop into seeds. Evidently S. amecamecanum is sterile or nearly so. Most of the plants on Iztaccihuatl may be the products of vegetative reproduction. In fact, on the basis of the small amount of variability among plants in cultivation, perhaps only two crosses would be necessary to account for the observed varia­tion in S. amecamecanum. One plant, my no. i, clearly is different from the others. It has longer, broader leaves, but was not in flower on Iztaccihuatl on Dec. 18, 1955, and has not flowered in cultiva­tion. The differences in floral characters of the other plants, ob­served in the field, disappear in cultivation. Except for plant no. 1, differences in width of leaves among plants are slight. Likewise, except for plants nos. 1 and 4, differences in length of leaves among the cultivated plants are negligible. Since the culture of no. 4 was in poor condition, possibly as a result of infection with nematodes, the shortness of the leaves of that probably is of no significance. In foliage, plant no. 1 is most nearly intermediate between S. dendroideum and Villadia batesii. Apparently it is what Praeger culti­vated and described as S. amecamecanum, and also it is the same as plants which I have cultivated. The other plants appear inter­mediate between that and V. batesii. A possibility is that no. 1 is an F1 hybrid and that the other plants are the result of a single backcross to V. batesii.

Sedum amecamecanum, on the basis of the evidence just cited, seems to be the result of hybridization of Sedum dendroideum and Villadia batesii. Plants intermediate between these two species may be either F1 hybrids, as perhaps Praeger's type, Jour. Bot. 55: 43 (1917), and my no. 1, or backcrosses to either parental species. The existence of such bigeneric hybrids and others to be described under S. quevae need not invalidate the generic status of Villadia. Any group of species which have evolved together and developed one or more important and distinctive characters may constitute a genus, as long as there are no fertile hybrids or species which are both reproductively fertile and morphologically intermediate. By an important distinctive characteristic is meant one with a complex genetic basis which in the group concerned is relatively stable. The present evidence does not require combining Villadia and Sedum, but it does indicate their close genetic relationship. Detailed genetic studies, involving study of chromosomes, of many plants both in nature and in cultivation, are necessary for an understanding of the mechanics of the hybridization just reported.

© Sedum of the Trans-Mexican Volcanic Belt, 1959