a' 3,? 1.113i2135‘?{ Michigan 32:25.3 Uzivasi’z ;‘ ‘ This is to certify that the thesis entitled A REVISION OF MATELEA SUBGENUS DICTYANTHUS (APOCYNACEAE, SENSU LATO) presented by Warren Douglas Stevens has been accepted towards fulfillment of the requirements for Ph.D. dPgTPP in Botany Major professor Date 26 Februar 1976 ABSTRACT A REVISION OF MATELEA SUBGENUS DICTYANTHUS (APOCYNACEAE, SENSU LATO) By Warren Douglas Stevens A revision of Matelea subgenus Dictyanthus (Gonolobeae) is provided. . Dictyanthus can be distinguished from the rest of Matelea by having a : digitate corona with the axes of the lobes entirely adnate to the corolla, simple inflorescences, a mixed indumentum with at least some of the trichomes glandular and at least some of the long trichomes uncinate, and follicles with thickish projections. Dictyanthus is here considered to be comprised of 10 species; two species, Matelea mac- vaughiana and fl. hamata, are proposed as new. Another species, previ— ously included in synonymy, is recognized and a new combination, Matelea aenea, proposed for it. Three species closely related to sub— genus Dictyanthus are also treated. One of these, Matelea altatensis, was previously included in Dictyanthus. Another, Matelea aspera, is a new combination published as a result of this study and is the type Species of subgenus Pachystelma, a taxon of uncertain status. The third, Matelea sepicola, is a new species published as a result of this study and is apparently most closely related to M, aspera. As a group, the three species can be distinguished from Dictyanthus especially by having corona lobes which are partially or entirely free from the corolla. A data—matrix is provided for subgenus Dictyanthus. Warren Douglas Stevens Also provided is a brief literature—based descriptive survey of ASclepiadoideae, especially Gonolobeae, and a summary of my own prelim— inary morphological and anatomical studies of the subfamily, especially with respect to the treated species. The stems of the treated species have inter—xylary phloem, a prominent pericyclic region defined by bundles of bast fibers, an endodermis represented by a starch sheath in young stems but not evident in older stems, and asymmetrical secondary xylem. Druses and branched, nonarticulated latex tubes occur in most tissues. The leaves have an unspecialized dorsiventral anatomy with a single palisade layer and anomocytic stomata. Only uniseriate trich— omes occur and these can be glandular or nonglandular, straight or un- cinate, and of varying lengths. Specialized glands occur on the stip— ular region, the base of the adaxial surface of the leaf blade, and the inside of the calyx tube below each sinus; because of their complex structure and specific sites, apparently homologous glands can be iden— tified in related families. The inflorescences are interpetiolar heli- coid cymes. The general morphology of the flower is that of Asclepiad— oideae. Each corona lobe of Dictyanthus is produced by an enation from near the base of a filament. Chromosome number data of Apocynaceae show a basic number of §_= 11; about 97% of the genera and 86% of the species counted can have this number. Polyploidy occurs mainly in a few groups, but about 36% of the genera and 22% of the species counted can have polyploid numbers. The relationships and distributions of the subfamilies of Apocyn— aceae and tribes of Asclepiadoideae are briefly discussed. Relation— ships within Matelea and within Dictyanthus are also discussed. A REVISION OF MATELEA SUBGENUS DICTYANTHUS (APOCYNACEAE, SENSU LATO) By Warren Douglas Stevens A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Botany and Plant Pathology 1976 —— 4——_“ -_ .____... ACKNOWLEDGMENTS First and foremost I would like to thank Dr. John H. Beaman for his patience and help throughout this project. Field work for my studies has been aided by a grant from the Latin American Studies Center, Michigan State University, and by the Univer— sity of Michigan Herbarium. The ten line drawings were prepared by Judy Appenzeller, supported by NSF Grant GJ—573. Drs. Beaman, Rollin Baker, Aureal Cross, and S. N. Stephenson have reviewed various parts of this manuscript and have offered valuable suggestions. Many other individuals have helped in various aspects of my work. Although not a complete list, I thank at least the following: Dr. Garrett Crow for examining a type specimen at the British Museum, Dr. Robert Cruden for two of the photographs used in this thesis, Dr. Melinda Denton for seeking out specific collection localities in Mexico and for briefly serving as my major professor, Dr. Richard Harris for examining my Latin diagnoses, Dr. Daniel Janzen for determining a col- lection of weevils, Dr. Rogers McVaugh for various kinds of help but especially for the use of his notes on the Sessé and Mocifio collec— tions, Dr. Lorin I. Nevling, Jr. for making available to me the Field Museum photographs of the Sessé and Mocifio herbarium, and Dr. Willaim Tai for advice on microscopy as well as the use of his equipment. And finally the one indispensable part of this study was the access to collections and I am indebted to the curators of the cited herbaria for allowing me to examine their material. ii . .---.__-____...._' __..____.‘ .—.~. TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . LIST OF FIGURES . . . . . . . . . INTRODUCTION . . . . . . . . . . GENERAL DESCRIPTION . . . . . . . . Morphology and anatomy . . . . . Roots . . . . . . . . Stems . . . . . . . . Leaves . . . . . . . . . Indumentum . . . . . . Glands (colleters) . . . . . Inflorescence . . . . . Flowers . . . . . . . calyx . . . . . . corolla . . . . . . . gynostegium . . . . . pollinia . . . . . . ovaries Fruits and seeds . . . . . Cyt010gy . . . . . . . . . . Distribution . . . . . . . . Ecology . . . . . . . - RELATIONSHIPS . . . . . . Family—subfamily . . . Tribe—genus . . . . . Subgenus—species . . TAXONOMIC TREATMENT . . . . Matelea subgenus Dictyanthus Matelea subgenus unassigned Notes on characters used in Taxonomic Treatment Artificial key to species 0 Species treatments (subgenus Dictyanthus) 1. Matelea hemsleyana 2. Matelea tuberosa 3. Matelea hamata . 4. Matelea pavonii . 5. Matelea macvaughiana 6. Matelea standleyana 7. Matelea ceratopetala 8. Matelea dictyantha 9. Matelea agnga . 10. Matelea yucatanensis Species treatments (subgenus unassigned) . . 11. Matelea altatensis 12. Matelea sepicola l3. Matelea aspera Cited collections . . Cited scientific names . Cited common names iv 66 66 71 75 75 76 76 82 82 91 97 103 117 126 130 138 148 154 159 159 169 176 188 193 195 APPENDIX (TAXONOMIC DATA-MATRIX) LITERATURE CITED . . o 196 205 Table LIST OF TABLES Page Chromosome numbers of Apocynaceae (see text for explanation) . . . . . . . . . . . . . . 53 Sessé and Mocifio collections pertinent to Matelea pavonii . . . . . . . . . . . . . . . 109 vi Figure 10 11 12 13 14 15 LIST OF FIGURES Page Representative basal parts of Matelea subgenus Dictyanthus (x ca 0.5) . . . . . . . . . . . 10 Representative stem thin-sections of Matelea sepicola and M} dictyantha . . . . . . . . . 14 Representative stem thin—sections of Matelea and HOEa O O O O O O O O O O O O O O O O O 16 Leaf surfaces of Matelea subgenus Dictyanthus . . . 22 Leaf features of Matelea and Hoya . . . . . . . 24 Representative glands of Matelea and Hoya . . . . 31 Outline of floral structures of Matelea dictyantha as seen in serial transverse sections . . . . . . 39 Representative flower thin—sections of Matelea sepicola, M, hemsleyana, and M, dictyantha . . . . 41 Pollinia of Matelea subgenus Dictyanthus, epi- illumination photographs, x 37 . . . . . . . . 48 Pollinia of four species of Matelea related to Matelea subgenus Dictyanthus, epi-illumination photographs, x 37 . . . . . . . . . . . . 49 Matelea hemsleyana (drawn from Stevens C-l62, a cultivated specimen of Stevens 1399) . . . . . . 85 Distribution of Matelea hemsleyana and M, tuberosa . 9O Matelea tuberosa (drawn from Stevens C—163 and C-l64, cultivated specimens of Stevens 1458 and 1473, respectively and Stevens 1473) . . . . . . 94 Holotype of Matelea hamata . . . . . . . . . 100 Inflorescence of holotype of Matelea hamata . . . . 101 Figure Page 16 Distribution of Matelea hamata and M, pavonii . . . 102 17 Matelea pavonii (ArD drawn from Stevens C—160, a cultivated specimen of Stevens 1375, and E from Stevens 1427) . . . . . . . . . . . . . . 107 18 Isotype of Matelea macvaughiana (VT) . . . . . . 121 19 Representative features of Matelea macvaughiana . . 123 20 Distribution of Matelea macvaughiana and M} standlgyana . . . . . . . . . . . . . . 125 21 Matelea standleyana (drawn from Stevens C—l61, a cultivated specimen of Stevens 1392) . . . . . . 129 22 Matelea ceratgpetala (drawn from Stevens 1245) . . . 133 23 Distribution of Matelea ceratopetala and M} dictyantha . . . . . . . . . . . . . . . 137 24 Matelea dictyantha (A-D drawn from Stevens C-105, cultivated specimens of Graham 1231, and E from Stevens 1311) . . . . . . . . . . . . . . 141 25 Regional variation of Matelea dictyantha . . . . . 143 26 Matelea aenea (A-D drawn from Stevens C-157, cultivated specimens of Stevens 1145, and E from Stevens 1145) . . . . . . . . . . . . . . 150 27 Distribution of Matelea aenea and M, ygcatanensis . . 153 28 Mgtelea_yucatanensis (drawn from Stevens C—158, a l cultivated specimen of Stevens 1168) . . . . . 157 ‘ 29 Representative specimen of Matelea altatensis (Wiggins §_Rollins 140, A) . . . . . . . . . 163 30 Representative features of Matelea altatensis . . . 165 31 Distribution of Matelea altatensis and M, sepicola . 168 32 Matelea sepicola (drawn from Stevens 1436) . . . . 171 33 Representative features of Matelea sepicola . . . . 172 34 Matelea aspera (drawn from Stevens 1296) . . . . . 179 35 Distribution of Matelea aspera . . . . . . . 131 viii Figure Page 36 Character list for DICTIM (revised 14 Dec 1975) . . 197 37 DICTlM, a general format taxonomic data—matrix for Matelea subgenus Dictyanthus . . . . . . . . . 201 38 Samples of interactive specimen identification using IDENT4 and DICTlM . . . . . . . . . . . . 203 ix INTRODUCTION The genus Dictyanthus was described by Decaisne in de Candolle's Prodromus in 1844. The description was based on a Sessé and Mocifio collection which had been distributed by Pav6n. Judging from Decaisne's annotations, he had also seen Galeotti specimens of the type species, Dictyanthus pgvonii, but he did not cite them in the proto- logue. A few years later, sometime in the late 1840's, Dictyanthus pavonii was introduced into European botanical gardens and became a relatively well-known plant. During this period the species was well illustrated in horticultural journals and was provided with several new names. Other than the addition of new names, the next treatment of the genus is that of Bentham and Hooker in their Genera Plantarum (1876). They considered the genus to be comprised of three to four Mexican species. Six years later, Hemsley (1882) treated the genus in Biologia Centrali-Americana, Botany and recognized four species, one of which he described as new. In Engler and Prantl's Die natfirlichen Pflanzenfamilien, Schumann (1895) again considered Dictyanthus to be a Mexican genus of three to four species. The next treatment, that of Standley in his Trees and Shrubs g£_Mexico (1924), includes six Species, one of which was described as new. Woodson (1941), in provid— ing a generic revision of the North and Central American Asclepiadace— ae, reduced Dictyanthus to a subgenus of Matelea and made new combina— tions for the ten species he recognized. These were simply listed, 2 with partial synonymy, to document the subgenus; there was no intent to provide a species-level revision. I am recognizing seven of the species Woodson listed, considering two of them as synonyms of other species, and removing One from the subgenus. Standley and Williams (1969) included the two Guatemalan species of this subgenus in their treatment of Asclepiadaceae in the E1253.2£ Guatemala. This summarizes essentially the entire taxonomic history of Dictyanthus except for the descriptions of individual species. As can be seen, Dictyanthus has not been previously studied except as part of a consideration of the whole subfamily Asclepiadoideae. To date, no species have been added to or removed from Woodson's subgenus and no new species belonging to the group have been described since 1930. Taxonomic neglect of one kind or another has been the fate of most asclepiads in tropical and subtropical North and Central America. A major problem is that the species tend to be poorly represented in herbaria. This is partially a reflection of the rarity of many species and partially the result of collecting biases. Many collectors tend to avoid the disturbed lowland forests which are relatively rich in asclepiads, to avoid collecting vines, and to avoid collecting groups, such as the asclepiads, which are not easily determinable. Except for the sorts of whole—subfamily studies mentioned above, and somewhat more detailed studies for species occurring within the United States, only three of the nine genera Woodson (1941) recognized in North and Central America have ever been revised for this region. The New World species of Marsdenia were revised by Rothe in 1915. Although the revision was well—done for the time, it is now almost totally out-of—date. Woodson (1954) revised Asclepias, probably the largest and certainly one of the 3 most difficult asclepiad genera in North and Central America. Asclepi- gg, being primarily temperate and primarily in temperate habitats even in the tropics and not being a vine, is much better represented in herbaria than the other genera. Unfortunately, Woodson provided only a natural key to the species. Even with an illustration for each spe- cies, identifications of Asclepias are often tedious. Sarcostemma, as revised by Holm (1950), is the only asclepiad genus in which species can be readily identified. Although his key is again largely natural, the much smaller number of taxa, the use of more conspicuous and dis- tinctive characters, and the generally better marked species make most determinations relatively easy. It was in part because of this general taxonomic neglect that I originally began working with Asclepiadoideae. Also attractive were the possibilities for ecological study of the highly specialized insect ‘pollination and insect predation of this group. The specific choice of Matelea subgenus Dictyanthus as my first project was influenced by sev- eral factors. Probably most important was the fact that the group was of such a size that the species could be considered in some detail, yet distinctive enough that one could be relatively certain of considering all the closely related species at once. Also, the ranges of the spe— cies were such that I could have a reasonable chance of examining all of them in the field. Also significant was the fact that the species tend to have large and attractive flowers, probably the largest of the New World Asclepiadoideae. This last fact has probably also been the cause for this group's being relatively well—collected, relative at least to the other viney asclepiads. Using Woodson's ten species of the subgenus as a basis, I began the T I l ' l 4 study by gathering the pertinent literature and by examining the her— barium specimens at Harvard Herbaria, the Smithsonian Institution, the University of Michigan, and Michigan State University. In this way, the available material was sorted essentially into what I now con- sider to be 14 species (there was a little lumping and splitting in this original sorting). These sorted elements were considered to in- clude all of the subgenus Dictyanthus plus the most closely related species. 0n closer examination, one of the species, Matelea congesta (Decaisne ig de Candolle) Woodson, proved to be only superficially sim— ilar and not actually closely related to the others. Therefore, it is not included in this revision. The remaining thirteen species are treated, ten as belonging to subgenus Dictyanthus and three not assign— ed to a subgenus; the status of the three unassigned species is dis— cussed in the section on Relationships. With these reasonably good concepts of the taxa and specific col- lection localities, about three months were spent in Mexico and Guate— mala collecting Matelea. Before Dictyanthus was chosen as a specific project, I spent three months in Mexico with a University of Michigan expedition; during this trip there was ample opportunity to collect asclepiads and several collections which proved to be important to the Study were made. After the taxonomic treatment was largely finished, an additional one—month trip was made to western Mexico to fill certain gaps in my material. As has been my practice since the beginning of my field work, considerably more than herbarium specimens are collected when an asclepiad is found (and members of certain other groups as well). Of major importance has been the collection of spirit~preserved flowers. Asclepiad flowers, especially when large or fleshy, are _.... . dras boil side tonn visit been Ear), I field been nany or or Specn about Huh The revisj the IE 3111103: In thi Signif that 1 % Plants 801119 C; mega E examins 5 drastically deformed by pressing and drying and no amount of soaking or boiling will restore their original shape (cf. Drapalik, 1970). Be— sides spirit—preserved flowers, also collected, as feasible, were ana- tomically preserved materials, insects either feeding on the plants or visiting the flowers, information on local names and uses (little has been contributed in this regard), and living plants or seeds. In sum— mary, of the 14 species originally considered, 13 have been seen in the field, spirit—preserved flowers have been collected of 12, and 10 have been successfully cultivated. In addition to my own collections, I felt it necessary to examine as many herbarium specimens as reasonably possible. In this regard, loans or other information were solicited from a total of 46 herbaria and specimens were actually received from 35. For the 13 treated species, about 880 specimens, representing 341 separate collections, have been examined. The major result of this work is the essentially classical taxonomic revision presented in the Taxonomic Treatment. The descriptions, thus the resulting differences and similarities among the taxa, were derived almost entirely from the population of herbarium specimens examined. In this sense, the resulting taxonomy would probably not have differed significantly if no field work had been done. The major difference is that I probably would not have been inclined to distinguish between atelea aenea and M, yucatanensis without having examined living lants. It should also be mentioned that additional field work may in ome cases provide justification for further taxonomic distinctions; hese are discussed under the appropriate headings. However, having Xamined a variety of material, especially spirit-preserved flowers and 111111 Relal lite] in 1); studj aCCUJm cerely family 6 living plants, I believe that I have a much better understanding of the taxa, their inter—relationships, and their relationship to the rest of Matelea than would have been possible from herbarium specimens alone. This understanding is at least partially expressed in the section on Relationships. The section on General Description is, in part, a brief literature—based survey of Asclepiadoideae, especially Gonolobeae, and, in part, a summary of my own preliminary morphological and anatomical studies of the subfamily, especially with respect to those species in— cluded in the Taxonomic Treatment. During the course of this study a number of specific questions re— garding the ecology and taxonomy of this group of species have become apparent. My contribution to the understanding of Apocynaceae will be relatively small if my future work is not directed toward these prob- lems as well as toward additional revisionary studies within the fami— ly. A considerable body of material, information, and inspiration has accumulated from my field, herbarium, and literature studies and I sin— cerely hope I will be able to continue studying this most interesting family. Roots diffL genus tens, store 3533! tiona @3332 two 3 0ften littl Subge Seton Woody and 3 Dim SPEci Subfa GENERAL DESCRIPTION Morphology and anatomy Beats The roots of most species of Matelea with perennial woody stems are diffuse and fibrous. The more herbaceous species, especially of sub— genus Chthamalia, typically have fleshier and less diffuse root sys— tems, probably associated with the increased importance of the roots as storage structures. The strictly erect species of subgenus 322597 trichis, which are probably ill—placed in the genus Matelea, are excep— tional in having a few thick, fusiform roots. The roots of subgenus Dictyanthus are diffuse and fibrous. The roots are least woody in the two species, Matelea ceratopetala (Figure 1D) and M, standleyana, which often have thin, nearly herbaceous, horizontal rhizomes which are little more than stems with adventitious roots. The other species of subgenus Dictyanthus appear to develop prominently woody taproots and secondary roots (Figure lA-C). When exposed on the surface, these woody roots develop fissured corky bark similar to that of the stem. The stem anatomy of Asclepiadoideae has seldom been studied. Scott and Sargent (1893) and Groom (1893) studied the aerial roots of Dischidia rafflesiana. Francke (1927) studied the exodermis of 12 Species, including examples from the various habit types within the subfamily. In their classic study on internal phloem, Scott and Brebner (1891) included discussions of about ten species of Apocynaceae; they r specir three three demi: cylinr specil third there Specir a Spel @ Der north. esting Occurs St& The herba( the p] cult t Perem but ft (010 C Stems) Repres 8 they discussed in some detail the root and root—stem transition of a species of Asclepias. Mayberry (1938) briefly described the anatomy of three North American species of Asclepias. He found the roots of all three species to have a suberized hypodermis (=exodermis) and an endo— dermis, this forming a continuous cylinder in two species and a broken cylinder in the third. The xylem was in a triarch arrangement in one species, tetrarch in the second, and forming a solid strand in the third. The cortex was filled with starch grains and oil globules and there were conspicuous calcium oxalate crystals in the cortex of two species. As far as I am aware, the roots have never been examined for a species of Gonolobeae. I have not examined the root anatomy of Matelea. Demeter (1923) studied the mycorrhizal associations of several north—temperate, herbaceous species of Apocynaceae. It would be inter— esting to know how widely the phenomenon of mycorrhizal associations occurs in the family, especially among the epiphytic species. M The perennial stems of Matelea vary from distinctly woody to nearly herbaceous and from aerial to ground—level or below. This aspect of the plants is often ignored by collectors and it is thus often diffi- cult to evaluate the habit of a given species. The most common type of P6rennial stem for the species of this revision is aerial and woody, but four species perennate from ground—level woody or fleshy caudices (two other species can have woody caudices in addition to aerial woody stems) and two species can have thin, nearly herbaceous, rhizomes. Representative examples of the basal parts of Matelea subgenus QEEEXf anthus are shown in Figure 1. These summarize the types of perennial Figure 1. Representative basal parts of Matelea subgenus Dictyanthus (x ca 0.5). A. erect woody stem, Matelea aenea, Stevens 1145 (MSC); B. woody caudex, M. dictyantha, Stevens 1311 (MSC); C. atypical form with elongate rhizome, M, dictyantha, Stevens 1311 (MSC); D. thin, nearly herbaceous rhizome, M} ceratopetala, Harmon §_Dyyer 3230 (M0). srems, Other e the Tax apparen piadoid ops on the nor perenni The be erec Of subg the cap the gm of Ma£ genus g A run of ASCL ll stems, and thus the habits, occurring in the species of this revision. Other examples are included in the figures of individual species within the Taxonomic Treatment. The species included in this revision, and apparently most of the rest of Matelea and many other genera of Ascle- piadoideae, are notable for the thick, fissured corky bark which devel— ops on the woody stems. This feature is often useful, especially for the more robust species, in distinguishing between the annual and perennial stems. The annual stems of Matelea are predominantly twining but can also be erect or procumbent. Although in natural conditions certain species of subgenus Dictyanthus are almost always erect, they all seem to have the capability to twine when under favorable conditions, especially in the greenhouse. There are, however, species within Woodson‘s concept of Matelea, e.g. M, balbisii (Decaisne 13 de Candolle) Woodson (sub- genus Pherotrichis) and M, caudata (A. Gray) Woodson (subgenus Mgli— ostemma), which seem to totally lack the ability to twine. A number of contributions have been made concerning the stem anatomy of Asclepiadoideae, but again Gonolobeae have been almost entirely ig— nored. The most comprehensive study was that of Treiber (1891). Treiber considered 59 species of Asclepiadoideae, of which one, 92227 lghgg hirsutus Michaux (=Matelea carolinensis [Jacquin] Woodson) be— longed to Gonolobeae. Most early references to Gonolobus, e.g. most of those in Treiber (1891) and Solereder (1908), apply to_§. condurango Triana, which was later found to be a species of Marsdenia, M, 22292— Egggg Reichenbach f., which is in the tribe Marsdenieae. Other notable references include the studies by Puech (1912a, 1912b) on the leafless species of Madagascar, by Zemke (1939) on several succulent African speciesf (1938) ( phloem ( by Chau‘ the Sta provide. The‘ are: 1 bicolla phloem C0ntinu (intra~ branche Parentl tissues a Prom Either dermis; of sube detail, SimPle The deScri; Figure: T All fixed j Paraffj Safram‘ of thit 12 species, by Scott and Brebner (1891) on internal phloem, by Mayberry (1938) on Asclepias, and by Singh (1943) on the inter— and extra—xylary phloem of Leptadenia. The laticifers have been studied in some detail by Chauveaud (1891) and Schaffstein (1932). The following summary of the stem anatomy is based on these studies as well as the information provided in Metcalfe and Chalk (1950) and Solereder (1908). The most significant features of the stem anatomy of Asclepiadoideae are: l) the presence of both inter- and extra-xylary phloem, both as bicollateral bundles and in the vascular cylinder; the inter—xylary phloem of the vascular cylinder may be in discrete bundles or in a continuous ring and may or may not be produced by a separate cambium (intra—xylary phloem also occurs in some species); 2) the presence of branched, nonarticulated or occasionally articulated latex tubes, ap— parently throughout most tissues; 3) the presence of druses in most tissues, these probably of calcium oxalate (cf. Esau, 1965, p. 29); 4) a prominent pericyclic region defined by a ring of bast fibers, these either separate or organized into bundles; 5) a well-defined endo— dermis; and 6) the superficial origin of the bark, either epidermally or subepidermally. Although I have not considered wood anatomy in any detail, the vessels of the viney species can be summarized as having simple perforations, bordered pits, and very wide lumina. The species considered in this revision conform to the general description above. Representative thin—sections are illustrated in Figures 2 and 31. The vascular tissue of the stem hardly passes All thin—sections used in this study are from material killed and fixed in 50% FAA, dehydrated through a standard TBA series, embedded in paraffin, sectioned at ten microns on a rotary microtome, and stained in safranin—fast green, all procedures after Sass (1958). All figures 0f thin-sections were photographed on a Zeiss photomicroscope. 13 Figure 2. Representative stem thin—sections of Matelea sepicola and M, dictyantha. A. transverse section of pedicel of small flower bud, Matelea dictyantha, x 150; B. transverse section of young stem, M. sepicola, x 35; C. transverse section of young stem, note inter—xylary (iph) and extra—xylary (eph) phloem, M. sepicola, x 120; D. longitudinal section through primary and early secondary xylem of very young stem, M, sepicola, x 250. l4 ....". .... \. . ...l . albino was... aEEsEfiEinitial_. is n... . minimo. g. .Iwnm I c. “£4.11qu .5 jaw/3%? 0‘.”- , .. , . . . . l P... . . . . : , 4A l l I! 2": ’ . ._'1.'. -. a2! O I. 25- _ t fl . I " q \_.:'_ with .I . . "Lav v.* 0.0a: . . .... .3“ flew-WW»... 15 Figure 3. Representative stem thin—sections of Matelea and Hoya. A. transverse section of very young stem, note subepidermal initiation of bark (b), Matelea sepicola, x 225; B. trans— verse section of young stem, note bundles of bast fibers and proliferation of bark, M, sepicola, x 70; C. transverse sec— tion of a single bundle of bast fibers, E, sepicola, x 305; D. transverse section of pith of young stem showing a druse, fl, sepicola, x 530; E. longitudinal section of young stem, note endodermis (e) outside of fibers (f), Hoya obovata Decaisne in de Candolle, x 55; F. transverse section of receptacle of small flower bud, with branched, nonarticulat— ed latex tubes, fl. dictyantha, x 395. :. .. .. .. 3.812}... . ... . . s .2 , 5....4 n.....qu.p..ww'r ~\.4r\. .. . . . . . . 'l t 0 $17.41.). _ .,. a J A . ..,,.~.~. ..L..0: .i y . .u t . 9. 1. . .. ‘4: n d \. . A. ... r-.. I... 19.; through cylinder young st the inte :o incre ent camb Branched and the dominant pith (F1 lied and Sanized starch 5 Contrast ASclepia Parenchy $th is i“itiate Perennia riCally, does Mt Same Phe‘ elements to hello. Fitting 1111111113 01 17 through a stage with discrete vascular bundles; a continuous vascular cylinder is formed even in young stems. Figure 2A illustrates a very young stem (the pedicel of a small flower bud). Even in young stems, the inter—xylary phloem is in the form of a continuous ring and appears to increase little in quantity as the stem enlarges; there is no appar— ent cambium specifically for the inter~xylary phloem (Figure 2C). Branched, nonarticulated latex tubes are conspicuous in both the cortex and the pith (Figure 3F). Within the stem, the latex tubes branch pre— dominantly or exclusively at the nodes. Druses are also common in the pith (Figure 3E) and cortex. The bast fibers are completely unligni- fied and aggregated into bundles (Figure 3C) and these bundles are or— ganized into a ring (Figure 2B). The endodermis is represented by a starch sheath in young stems but is not evident in older stems. In contrast, the conspicuous endodermis said to be characteristic of Asclepiadoideae is illustrated by a species of Egy§_in Figure 3E. The parenchyma of the cortex, and to a certain extent of the pith, of older stems is filled with starch grains (e.g. Figures 3C and 3E). Bark is initiated subepidermally (Figure 3A) and is produced in quantity on perennial stems (Figure SB). The secondary xylem is produced asymmet- rically, giving the vascular tissue an oval shape (Figure 2B), but this does not appear to alter the external shape of even old stems; this Same phenomenon was described by Handa (1936) for a species of garg— 23313 and is possibly a common feature. The ontogeny of the tracheary elements (cf. Esau, 1965, p. 232, fig. 11.4) from annular thickenings to helical thickenings to scalariform pitting and finally to circular Pitting (bordered pits) can be seen in Figure 2D. Vessels with wide lumina occur in the secondary wood and, although no very old stems have been 5 rent ( hggxgg The Asclep consul quist, never ( stipul; tricho: oideae "Membex Without ent. 1 ten foL Vestigj Oped p5 m0Stly descrip has the and gen Hickey. Petiole more pr ter the the art meaSUre 18 been sectioned, appear to have a weakly—developed ring—porous arrange— ment (Figure ZB). Leaves The leaves of Matelea, as well as other Gonolobeae and almost all Asclepiadoideae, are opposite in arrangement. Depending on the source consulted, the group is considered either to be exstipulate (e.g. Cron— quist, 1968) or to have minute stipules (e.g. Lawrence, 1951). I have never observed typical laminar stipules in Asclepiadoideae, but the stipular region usually has a fringe of glands (colleters) and/or trichomes (Figure SE). A similar situation also occurs in Plumeri— oideae and Apocynoideae, of which Standley and Williams (1969) state: "Members of this family [Apocynaceae, sssx] are usually said to be without stipules, however stipules or stipular vestiges are often pres— ent. Interpetiolar stipules much like those in some Rubiaceae are of— ten found . . . ." Boke (1947) interprets these stipular glands as vestigial stipules. All the species of this revision have well—devel— oped petioles and basically ovate leaf blades with lobate bases and mostly acuminate to attenuate apices. The terminology used for the description of the leaves is that of Hickey (1973). This terminology has the advantages of being comprehensive, similar to current usage, and generally with precise definitions. Although not specified by Hickey, I measured the length of the blade from the attachment of the petiole to the apex, that is, the length of the midrib. This seems more precise than including the prominent basal lobes and does not al— ter the basic leaf shape (the widest part of the leaf is often below the attachment of the petiole). It should also be noted that the leaf measurements were always based on the largest leaf of a specimen. Althoug there 1‘ stem, ( A are ty; tend tr apices rower s | leaves, Species naturi1 d0 not 0f Asci cluding ticelh (coller Hickey‘ acropet rib (a1 blade, tion, , % midrib, The Hickeyr 08y Sec Someom 19 Although the leaf development is probably not truly heteroblastic, there is often a marked difference in leaf size and shape along the stem, especially associated with the erect habit. The largest leaves are typically near the middle of the annual growth; the lower leaves tend to be smaller, broader, and with broader sinuses and more abrupt spices and the upper leaves tend to be smaller, narrower, and with nar— rower sinuses and longer apices. Even when measuring only the largest leaves, the resulting size and shape ranges are still quite large. This is probably in part a reflection of the natural variability of the species but is certainly also influenced by differences in habitat, maturity of the specimens, and the fact that herbarium specimens often do not have the section of stem bearing the largest leaves. The leaves of Asclepiadoideae always have entire margins, but certain species, in— cluding some treated here, can have a ragged appearance due to the mul— ticellular bases of the long trichomes. The placement of the glands (colleters) on the leaf presents something of a problem with respect to Hickey's terminology. I have tentatively referred to the position as f acropetiolar, but they are actually on the adaxial surface of the mid— 1 rib (and sometimes also on the first lateral veins) at the base of the blade. Hickey provides no term which specifically describes this posi— 3 tion, which occurs on most Asclepiadoideae and some Apocynoideae (e.g. Mandevilla), nor for the similar position of being scattered along the midrib, which occurs in some Apocynoideae. The leaf venation has not been carefully described according to Hickey's terminology. Although this is to be desired and the terminol— ogy seems clear, I am inclined to believe that it should be done by someone familiar with a broad range of venation types. For instance, it is n ed. Th or rwo lwfsu sions, arrange K actinod paring examini The deg raised Venatic are ill with ha sured 1 EHCY tc tend tc reduce that rt [“83st preSSed The only si Smooth determj that th 20 it is not clear to me how the basic venation pattern should be describ- ed. The basic pattern is probably brochidodromous, but the first pair or two of lateral veins are often fully as large, both in the amount of leaf surface they supply and in their thickness and number of subdivi— sions, as the midrib and are often congested into virtually a palmate arrangement (Figure 5F). This pattern could just as well be called actinodromous (basal, perfect, reticulate). Someone interested in com— paring angiosperm venation patterns could most likely profit more from examining the leaves I have studied than from my descriptions of them. The degree to which the veins, including the higher—order veins, are raised from the lower leaf surface is one conspicuous feature of the venation that does vary within the species I examined. The extremes are illustrated in Figures 4E and AF. This feature varies somewhat with habitat but is still often of diagnostic value. I have not mea— sured leaf thickness, but two species, Matelea altatensis and M, aspera, appear to have thinner than average leaves and a strong tend— ency to wilt. Matelea altatensis, and to a certain extent fl: aspera, tend to grow in arid environments and this may be one adaptation to reduce transpirational water loss. A practical consequence of this is that the herbarium specimens of these two species often have poorly— pressed leaves. The herbarium specimen in Figure 29 is about the best— pressed example of Matelea altatensis. The terminology of leaf surfaces is according to Stearn (1966). The only Significant variation is that in some species the surface is smooth while in others it is pusticulate (Figure 4A—D). I have not determined the anatomical basis for the pusticulations, but I suspect that they mark the sites of large druses. The pusticulations have some Figure 4. 21 Leaf surfaces of Matelea subgenus Dictyanthus. A. adaxial leaf surface of Matelea standleyana, Stevens 1392 (MSC), smooth, x ca 65; B. abaxial surface of M. Standley— ana, Stevens 1392 (MSC), smooth, x ca 65; C. adaxial surface of M. pavonii, Stevens 1435 (MSC), pusticulate, x ca 65; D. abaxial surface of M. pavonii, Stevens 1435 (MSC), pusticu- late, x ca 65; E. abaxial surface of M, dictyantha, Stevens 1311 (MSC), veins raised, x 4; F. abaxial surface of M, standleyana, Stevens 1392 (MSC), veins not raised, x 4. 22 Figure 5. 23 Leaf features of Matelea and Hoya. A. thin—section of blade, note large complex druse, Matelea sepicola, x 30; B. thin—section of blade, M, hemsleyana, x 30; C. thin—section of blade, base of long trichome in lower right corner, M, obovata (H.B.K.) Woodson, x 30; D. trans- verse section of major vein, M. sepicola, x 9; E. young node of Hoya obovata Decaisne in de Candolle, note stipular (sg) and foliar (fg) glands, structures immediately below leaves are adventituous roots, x ca 3; F. adaxial leaf surface showing major veins, M. tuberosa, Stevens 1458 (MSC), x ca 2. 24 O '70 56-? ‘,4 LT . ._ I H' I. tendenr the di: (ranun tingui stomat Sig ae inc and No Studie are th the 15 Multi- f0Und ered } ed be. Sensor M leave; walls 131 SI is Sh 25 tendency to be more prominent along veins and have no relationship to the distribution of trichomes. The stomata are of the anomocytic (ranunculaceous) type; the cells surrounding the guard cells are indis— tinguishable from the other epidermal cells. Paracytic (rubiaceous) stomata are more common in Asclepiadoideae, but the anomocytic type is also recorded as occurring in Sarcostemma, Solenostemma, and Vincetoxi— ggm [Cynanchum?] (Metcalfe & Chalk, 1950) and Heterostemma and Tyloph— g£a_(Krishnamurthy & Kannabiran, 1970). This is apparently the first recorded observation of stomatal type for Gonolobeae. Significant studies of the leaf anatomy of members of Asclepiadoide— ae include those by Vesque (1885), Trochain (1932), Mayberry (1938), and Nolan (1966). The primary specializations considered by these studies are related to epiphytic and xerophytic habits. Among these are thick epidermal walls or cuticles, stomata equally distributed on the leaf surfaces, centric arrangement of the leaf mesophyll, and multi—layered palisades. Certain species of Matelea will likely be found to exhibit some of these features, but the species I have consid— ered have relatively unspecialized leaf anatomy. This is to be expect— ed because the leaves are deciduous and produced only during the wet season. Figure 5A—D illustrates some features of the leaf anatomy of Matelea. As with the stems, latex tubes and druses are present in the leaves. Consistent with their unspecialized nature, the epidermal cell walls and cuticles are thin, the stomata are predominantly on the abax— ial surface, the mesophyll is distinctly dorsiventral, and the palisade is single—layered and loosely packed. Mum 0n. unicel glandr are n. trich. have I hav. serla‘ atic : in a 1 class adequ. To I hav. w Shl they .1 nied l haVe r0113 0ften Slll’fa. diatir larget ifiner 26 Indumentum Only two types of trichomes are known to occur in Asclepiadoideae, unicellular and uniseriate. The complex structures variously known as glandular shaggy hairs, glands, squamellae, denticles, and colleters are not true trichomes and are considered separately. The uniseriate trichomes of Asclepiadoideae can have several modifications and these have systematic significance in at least some cases. Among the species I have considered and perhaps in the whole tribe Gonolobeae, only uni— seriate trichomes occur and their various forms definitely have system— atic significance. I have not critically examined these modifications in a broad enough selection of Gonolobeae to establish a satisfactory classification of the types, but this will certainly be necessary to adequately understand and describe the relationships within the tribe. To describe the indumentum of the species treated in this revision, I have used the convention of referring to all trichomes as short, glandular, or long, and these terms are modified as appropriate. Short trichomes are considered to be those less than 0.1 mm long; they are mostly about 0.05 mm long. They are few—celled, not accompa- nied by specialized epidermal cells, and mostly straight (but sometimes have a hooked terminal cell). Except on the inner surface of the co— rolla, short trichomes are almost always mixed with long trichomes and often also with glandular trichomes. The short trichomes of the inner surface of the corolla, when present, are always unmixed and have a distinctive form; they have from one to a few small basal cells and one larger fusiform terminal cell. When dried, the short trichomes of the inner surface of the corolla typically have a glassy appearance. Glandular trichomes are about the same length as short trichomes but are ir epider collar but wt lum. could no evi caller dried Lor they a What 5 trichc Crete trichc 0f the The e; raisec cTall) Woc M ed int havinE Mate16 \ omes, 27 are inflated in the middle. They are not surrounded by specialized epidermal cells. In dried specimens, the inflated part almost always collapses and presents the appearance of a capitate glandular trichome, but when fresh, these trichomes always have a straight or hooked apicu- lum. The inflated part is thin—walled and presumably the contents could be released by diffusion or by mechanical damage, but I have seen no evidence of anything being actively secreted from them. I have called these trichomes glandular because they have that appearance on dried specimens and they have almost always been referred to as such. Long trichomes are considered to be those more than 0.1 mm long; they are mostly one to three or four millimeters long but can be some— what shorter, especially when without accompanying short and glandular trichomes. When long trichomes occur alone, they are often of two dis- crete lengths, giving a mixed indumentum of long trichomes. Long trichomes are several—celled and, depending on the species and the part of the plant, the terminal cell can be hooked (uncinate) or straight. The epidermal cells surrounding the long trichomes are modified into a raised ring or collar; this basal collar can be quite prominent, espe— cially on the leaf blade. Woodson (1941) partially justified the separation of Gonolobus from Matelea on the basis of the indumentum, Matelea typically having a mix— ed indumentum of long plus glandular trichomes and Gonolobus typically having an unmixed indumentum lacking the glandular trichomes. Within Matelea, the nature of the trichomes, especially the glandular trich— omes, varies considerably. Uncinate long trichomes apparently occur on only a few species of Matelea other than those treated in this revision. The systematic importance of trichomes is further discussed in the sectic indume on Ecc glgndg The cenfus epider which quist single emerge this i emerge tures trichor that r} of Rfli Merge] Cells 1 right 1 trichor. of ASCj Diadau SeTints if the (1965) Qes SUC Hm :r— —-—-— 28 section on Relationships. The possible adaptive significance of the indumentum of the species of this revision is discussed in the section on Ecology and Distribution. Glands (colleters) The identity and occurrence of these structures has been very much confused in the literature. The term "trichome" is normally applied to epidermal appendages and the term "emergence" is used for appendages which involve subepidermal tissues (Uphof, 1962; Esau, 1965). Carl— quist (1959) further implies that a trichome should be the product of a single protodermal initial. The distinction between trichomes and emergences is, however, not always clear (Uphof, 1962; Esau, 1965); this is especially true when a normal trichome is raised upon an emergence. It is perhaps for this reason that various types of struc- tures fitting the definition of emergences are often considered to be trichomes. Uphof (1962, pp. 25’26) states, "It does not seem right that the stinging hairs of Urtica are, in the same way as the prickles of‘figgg and those on the fruits of Aesculus and Datura, regarded as emergences, simply because in these stinging hairs too the subepidermal cells take part in the development." But neither does it seem right that the glandular structures of Apocynaceae be regarded as trichomes. Metcalfe and Chalk (1950) refer to the glandular structures of Asclepiadaceae as "glandular shaggy hairs" and do not include Ascle— piadaceae in their list of "Families in which only unicellular or uni- seriate hairs have been found," in which it would certainly be included if the glandular structures were not considered to be hairs. Esau (1965, p. 309), in discussing secretory structures, refers to "emergen~ ces such as the shaggy hairs of Nerium [Apocynoideae]" in the text, but rah.— in tht ones. calle son & lar g when ' Woodsr ahom ed to! referr desire The tive u geneti Figure they a finger fully and ob ally s tiSSue axis 0 than t‘ Rutlei tend s. Stalks glands 29 in the text (Fig. 13.1) labels them as "multicellular glandular trich— omes." Besides the terms already mentioned, these structures have been called, in Apocynaceae, squamellae when they occur on the calyx (Wood- son & Moore, 1938; Holm, 1950; Woodson, 1954; Safwat, 1962) and stipu— lar glands, glandular emergences, foliar glands, denticles, and glands when they occur on the leaves or on the stipular area (Rothe, 1915; Woodson & Moore, 1938; Holm, 1950; Woodson, 1954; Leach, 1970). The almost certainly homologous structures of the Rubiaceae have been term- ed colleters (cf. Anderson, 1972). In this treatment I have simply referred to the structures as glands, but a more specific term would be desirable. Whatever they may be called, these structures have a very distinc- tive morphology, anatomy, and distribution on the plant. Their phylo— genetic distribution and function appear to be consistent as well. Figure 7 illustrates the glands of Matelea and Hoya. Developmentally, they appear to be initiated in the subepidermal tissues and form l finger—shaped projections covered by a single epidermal layer. When fully developed, the glands are generally one to two millimeters long ” and obliquely conical; there is a solid core of cells which are gener— i ally somewhat smaller and more tightly packed than the subepidermal tissues with which they are continuous and somewhat elongated with the axis of the gland. The epidermal cells are palisade—like, much larger than the adjacent epidermal cells, densely staining, and with the nuclei situated near their bases. Normal epidermal cells typically ex— tend somewhat up the base of the gland, making the gland shortly— stalked. In all the examples I have examined from Asclepiadoideae, the glands are especially peculiar in that the tip is somehow chemically Figure 6. 30 Representative glands of Matelea and Hoya. A. young leaf showing glands (g) at base of adaxial surface of blade, Hoya obovata, x l; B. longitudinal section of calyx gland, Matelea hemsleyana, x 200; C. longitudinal section of shoot apex, note how leaf (fg) and stipular (sg) glands enclose apical meristem (see also Figure 5E), H. 92: ovata, x 35; D. longitudinal section of leaf gland, note secretion and differentiated tip, M, obovata, x 85; E. transverse section of calyx lobe (cl) and glands (g) of a young flower bud, M, dictyantha, x 90; F. longitudinal sec- tion of leaf gland, note differentiated tip, M, hemsleyana, x 140. 31 0. .. é}? h.‘&, s_‘ y [‘1‘ diff calf when rani tiss nenb 1972 and} CUSSI 1:: (Q unger lungs liege note leave tions gland in pr as th fOlia EnClo Seen early also denCe 32 differentiated from the body; the tip is hardly, if at all, morphologi— cally or anatomically distinct, but has a slightly different color when living and distinctly different staining characteristics in saf— ranin—fast green. In Asclepiadoideae I have not observed vascular tissue entering or even approaching these glands, but in some other members of Apocynaceae (Woodson & Moore, 1938) and Rubiaceae (Anderson, 1972) the glands are sometimes vascularized. The glands apparently function in secreting substances which coat and protect the meristematic tissues. Rothe (1915, p. 359), in dis— cussing the foliar glands of Marsdenia notes, "An lebendem Material von M, cundurango Rchb. fil. fand ich bei meinen anatomischen Untersuch— ungen, dass diese Drfisen schon in der Knospe im Verhaltnis zu den jungen Blattanlagen ausserordentlich gross und schon an den Knospe an— liegenden Blfittchen voll entwickelt sind." Metcalfe and Chalk (1950) note that the stipular glands (glandular shaggy hairs) coat the young leaves with mucilaginous and resinous material. Both of these observa— tions apply to the material I have examined in Asclepiadoideae. The glands are fully developed (and thus relatively large) and functional in proximity to the meristematic tissues and appear to become inactive as the surrounding tissues mature. As can be seen in Figure 6C, the foliar and stipular glands, in their opposite positions, essentially enclose the apical meristem. Traces of the glandular secretion can be seen in Figure 6D. The calyx glands are also relatively large in the early developmental stages of the flower buds (Figure 6E) and probably also coat the meristematic tissues. One piece of circumstantial evi— dence for this function in flower buds is that in several species of Secamone with the calyx lobes arranged so that two are outer and three N.— are 1' 1962) tarie enlig to be the z asclr the r the ; threc stipr cept; aPier Sargr leaw redur of m.- stipy sPee: ilTe: ae’ ] are S in A] Short impl: 33 are inner, only the three inner lobes are supplied with glands (Safwat, 1962). Rothe's (1915) suggestion that the glands were extrafloral nec— taries which attracted ants as a defense against herbivors was quite enlightened for his era but is unlikely because the glands only appear to be active on very young tissues where they would be inaccessible to the ants. I have never observed significant numbers of ants on an asclepiad, but Rothe's suggestion might be considered with respect to the myrmecophytic species of Dischidia (Asclepiadoideae). In Asclepiadoideae glands can be found in three specific sites on the plants and they are nearly universal in their occurrence. The three sites are the base of the adaxial surface of the leaf blade, the stipular area, and the adaxial surface of the calyx tube. The only ex— ceptions of which I am aware are that the glands are also found on the apices of the leaves of the myrmecophytic species of Dischidia (Scott & Sargent, 1893), they are missing from the leaf blades of a few broad- leaved species, are often missing from the leaf blades of species with l l I . s 1 reduced and caduceus leaves, and are missing from the stipular region of many of the stem—succulent species. The presence or absence of stipular glands is sometimes of taxonomic significance in the succulent species of Ceropegieae (Leach, 1970; Dyer, 1971). In the other subfam— ilies of Apocynaceae, glands are not as universal as in Asclepiadoide— ae, but are found on the same places on the plants (except that they are sometimes also scattered along the petiole and midrib of the leaf in Apocynoideae). The peculiar structure and positions of these glands should make it possible to identify truly homologous glands in other families, and they could therefore have some interesting phylogenetic implications. In this regard, it would probably be useful to have a 1" .1 speci to pr colle ther secre perha nolog havir 34 specific descriptive term for this type of gland, but I would not like to propose one until the matter has been further studied. The term colleter, a "name given by HANSTEIN to those glands which secrete ei- ther mucilage, resin, or a mixture of the two, and which start their secreting activity at a very early moment" (Uphof, 1962, p. 174) is perhaps the best available term, but as currently used includes nonho— mologous structures. Metcalfe and Chalk (1950) list 31 families as having glandular shaggy hairs, but again these are almost certainly not all homologous. In the list, Asclepiadaceae are incorrectly noted as having these glands "infrequent or rare" when, in fact, they are much more common in Asclepiadoideae than in Apocynaceae (§,§,), for which they are listed as "especially common." From Metcalfe and Chalk's list and family descriptions, there are several other possible examples of homologous glands. As mentioned above, the glands of Rubiaceae are almost certainly homologous with those of Apocynaceae (§.I.). Other notable possibilities include Bartonia and Obolaria of Gentianaceae, Fagraea and Strychnos of Loganiaceae, and Calonyction and Ipomoea of ConVOIVulaceae (referred to as extrafloral nectaries). The distribu— tion of these glands is an especially interesting subject with definite Phylogenetic implications and warrants further study. Inflorescence The asclepiad inflorescence has been a source of interest since the earliest botanical writings. Students, both metaphysical and scientif— ic, have speculated on its pecularities to the point where it is hard to imagine a novel hypothesis being proposed. The earlier literature is well summarized by Nolan (1967) and more recent pertinent c0ntribu— tions have been made by Brunaud (1968a, 1968b, 1968c, 1968d, 1969a, l l r I . 19691 (1961 ints to t] "strl lore cenc Nola Cle .‘ pendr acts Nolar aSClr rate not ; mom] 11’ dc inter temm but I Nolar reSec Show The I bran agair 35 1969b, 1970a, 1970b, 1971a, 1971b), Jonsson (1968a, 1968b), and Nolan (1969). The major peculiarity of the inflorescence is that it is interpetiolar and not subtended by any structure. Apparently related to this is the fact that at the flowering nodes one leaf subtends a "strong" bud while the opposite leaf subtends a "weak" bud. The phyl— lotaxy of the leaves, the "weak" and "strong" buds, and the inflores— cences (which side and which axil they are closest to) are related. Nolan, through extensive anatomical studies, concludes that the pedun— cle and shoot apex are produced by a true dichotomy and thus are inde— pendent of subtending structures. Tomlinson et a1. (1970) apparently accept this as the only known example of a true dichotomy among dicots. Nolan's suggestion that this is a primitive character and that the asclepiads have evolved separately from the rest of the dicots, sepa— rate even from Apocynaceae (§t§.), warrants no consideration. I will not attempt to evaluate the competing theories. The easily observed morphological features of the species treated in this thesis are brief— ly described below. All the species treated, and apparently all Asclepiadeae, have interpetiolar helicoid cymes. Nolan (1967) is probably the first con— temporary botanist to refer to the inflorescence as a helicoid cyme, but that is a perfectly accurate descriptive term in this case. Nolan's (1967, figure 122) diagram of a flowering shoot with inflo— rescences ("inflorescence-units," Nolan considers the whole flowering shoot to be the inflorescence proper) essentially fits my observations. The peduncle produces a single terminal pedicel subtended by a single ract. The bud in the axil of this bract continues to develop and gain terminates with a pedicel ("major pedicel"), this time subtended _...___. ___—-_-——-_____...- by tn is pr cence subte the c the i right intri sive dire: calyx flowe tiate "nine Wise 5PeCi tyne thesi €9.an ed wi ti8hr here , Ceflee t10m hapPe that 36 by two bracts. From the axil of one bract a pedicel ("minor pedicel") is produced and from the axil of the other bract a shoot (= inflores— cence axis) is produced which again terminates in a "major pedicel" subtended by two bracts, one axil giving rise to a "minor pedicel" and the other the inflorescence axis, and so on. In a given inflorescence, the inflorescence axis is consistently produced from the axis of the right or left bract, thus producing a right— or left—handed helix. An intriguing observation of Nolan's is that the inflorescences of succes— sive nodes are enantiomorphic. They are enantiomorphic not only in the direction of the helix, but also in the sequence of initiation of the calyx lobes; in the right—handed (clockwise) helices, the terminal flowers (produced by the "major pedicels") have their calyx lobes ini— tiated in a clockwise order, and the lateral flowers (produced by the "minor pedicels") have their calyx lobes initiated in a counterclock— ‘wise order. In the left—handed helices, the reverse is true. In some species, e.g. Asclepias syriaca, Nolan found the (apparent) helicoid cyme condensed into a virtual umbel. In the species treated in this thesis which produce many flowers, e.g. Matelea sepicola, the inflores— cence axis becomes considerably thicker than the peduncle and is cover— ed with the scars of the pedicels and bracts, but still conforms to a tight helix. One exception to Nolan's scheme occurs in the three species treated here which can produce compound inflorescences. The compound inflores— cence occurs when one of the "minor pedicels" is replaced by an addi— tional inflorescence axis. See Figure 30C for an example. This can happen from one to several times in an inflorescence. The observation that subgenus Dictyanthus produces only simple inflorescences is of some rescr (see they dunc I of viou CEBU m sympr adna cept the base Prod in f eVen lobe scri Scri CUSs Para cial Shou [finch 37 some diagnostic value. A more troublesome exception is when an inflo— rescence is reduced to a single flower which lacks the subtending bract (see Figure 19D for an example) which at least topologically separates the peduncle from the pedicel. Another exception occurs when the pe- duncle is branched below the level of the first bract (see Figure 33C). I offer no explanations for these exceptions but the phenomena are ob— viously significant in evaluating the nature of the asclepiad inflores— cence. Flowers The asclepiad flower can be described as actinomorphic, S—merous, sympetalous, with epipetalous stamens completely connate laterally and adnate to the style apex, and with two carpels which are separate ex— cept at the apex and have essentially superior ovaries. The gynostegi— um can be loosely referred to as the androecium plus the style apex but the corolla, below the insertion of the stamens, forms at least the base of the structure. A "corona" is often present and is typically produced by enations from the filaments (no longer readily identifiable in the gynostegium), but sometimes other parts of the gynostegium or even the corolla are involved. At least in Dictyanthus, the corona lobes are distinct enough from the rest of the gynostegium to be de- scribed as separate structures. Despite this apparently simple de— scription, the flowers are actually much too complicated to be dis— cussed in any significant detail here. I will restrict the following Paragraphs to a brief discussion of the most important features, espe— cially as they apply to Matelea subgenus Dictyanthus. The description Should be prefaced, however, by noting that Gonolobeae have been as much ignored with respect to their floral morphology and anatomy as with volv Corr (19% (196. Figu spec are] the 1 shape the: tYpit insic 38 with the previously discussed topics; there is no study directly in— ' volved with Gonolobeae, but the most useful references include those by Corry (1884), Demeter (1922), Deshpande and Joneja (1962), Drapalik (1970), Frye (1902), Mulay et a1. (1965), Payer (1857), Rao and Ganguli (1963), Safwat (1962), Woodson (1933), and Woodson and Moore (1938). Figures 7—10 demonstrate various floral features of the treated species. Calyx. The calyx of Dictyanthus is 5—1obed nearly to the base and is always green. During development of the flower bud, the calyx lobes are relatively large and erect (Figure 7A). The calyx thus encloses the rest of the bud without actually clasping around the corolla. The shape of the lobes has been described in the same way as the shape of the leaves, that is, according to Hickey (1973). The calyx lobes are typically provided with an indumentum on the outside and are glabrous _. ;__-_.__‘_._._— inside. The calyx vasculature varies considerably in Apocynaceae (Woodson & Moore, 1938; Safwat, 1962); Dictyanthus is of the l-trace, T l—gap type. Each trace trifurcates almost immediately and the adjacent lateral branches typically join to form a ring of vascular tissue which h gives rise to sevaral major veins to each calyx lobe. Within the E calyx, slightly below each sinus, is one or occasionally two glands. As described above, these glands apparently function in producing a protective secretion for the developing bud. Corolla. The corolla of Dictyanthus is most often relatively large nd distinctly campanulate. In nearly mature flower buds the corolla 's plane or slightly concave on top and strikingly drum—shaped, hence he name of one of the synonymous genera, Tympananthe. I have des— ribed the shape of the corolla tube as "convoluted" in the literal igure 7.. Outline of floral structures of Matelea dictyantha as seen in serial transverse sections. Outlines of sections ca 1 mm apart, a nearly mature flower bud, ca x 6, from A near top of flower to I near base. Figure 8. 40 Representative flower thin-sections of Matelea sepicola, M, hemsleyana and M, dictyantha. A. longitudinal section of young flower bud, note epipetal- ous stamens with initials of corona lobes (ci) and terminal appendages (ta) overlapping style apex (sa), one style (3) and one ovary (o) visible in this plane, Matelea sepicola, x 40; B. longitudinal section of nearly mature flower bud, corona lobe (cl) fully formed, both styles and ovaries visi- ble in this plane, M, sepicola, x 40; C. transverse section of nearly mature flower bud, note corona lobes (c1) and ad- jacent wings of corolla (co), style apex (sa), corpuscula (cp), essentially horizontal pollen sacs (pc), anther wings (aw), stigmatic chamber (sc), and stigmatic surface (ss), refer to Figure 7, especially B and C, for orientation, M, hemsleyana, x 40; D. transverse section of corona lobe of nearly mature flower at level of nectary, corona lobe axis (cl) adnate to corolla (co) and adnate by a wall to gynoste— gium proper (gyn), secretory epidermis (se) of nectary on sides of lobe, refer to Figure 7E for orientation, M. dicty— antha x 45. _ ..,,. av"— .m 4'- ___‘.h-f'_ If»! sen aem sem whh vhil bulg drav fear coro have coro such fans 0the I3 [:3‘ b6cor from the 5 "Shar Often ten S 3180 the c COust. fella “Se—13 tunaj 42 sense of the word rather than in the conventional botanical sense (the aestivation is actually best described as imbricate). In transverse section (Figure 7), the corolla tube is convoluted because the parts which are adnate to the corona are drawn in toward the gynostegium while the alternating parts are bulged out (more or less saccate). The bulged out parts in some cases actually meet behind (outside of) the drawn in parts. Woodson (1941) states that, "The only really unique feature of Dictyanthus is that the faucal callus, or annulus, of the corolla is digitate, as are the segments of the corona." Woodson may have been referring to the overall shape of the corolla tube, but the corolla is not specialized, either anatomically or morphologically, in such a way that there is a structure which could be identified as a faucal annulus or callus. There are, however, good faucal annuli in other Gonolobeae. In two species of Dictyanthus, Matelea tuberosa and M. hemsleyana, the base of the corolla, lateral to the corona lobes, ecomes somewhat winged, separating the bases of the sacs (or bulges) rom the gynostegium (Figure 8C). Woodson also states, in his key to he subgenera of Matelea, that Dictyanthus has corolla lobes which are ' The former is 'sharply revolute” and "essentially glabrous within.‘ ften but not always true and the latter is true for only one of the en species. The margin of the corolla limb, as well as the lobes, is 1so often sharply revolute. The indumentum of the inner surface of he corolla is always of short trichomes but its distribution is of onsiderable diagnostic value. The pattern of markings within the co— 011a is also of considerable diagnostic value. In order to more pre— isely describe the pattern of indumentum and markings within the co— lla, I have divided the corolla into the tube, the limb, and the lob leru vidt whet plus ure ous leas baSe Push (Fig with Part Steg baSe SPee Stru cies are with ePid. not 1 % theE whit} 43 lobes, rather than the more conventional tube and limb. The two corol— la measurements used are the length, from the base to a sinus, and the length of the lobes, from the tip to a sinus to sinus line. The indi— vidual lobes are l—trace, l—gap structures and the pattern of markings, when present, follows the pattern of the major veins. Gynostegium. The gynostegium is composed of the connate stamens plus the adnate style apex and, at the base, the corolla tube (see Fig— ure 7F-G). In anatomically prepared material the identity of the Vari— ous structures is relatively easy to determine. In Dictyanthus, at least, the corona is always the product of one enation from near the base of each filament; developmentally, these enations bulge out and push their way against and up the corolla and become completely adnate (Figure 8A-B). The vascular bundle of each filament is carried along with the bulge and forms a prominent loop (Figure 8B). A thin wall or partition typically connects the corona lobe to the body of the gyno— stegium (Figures 7D—F and 8B and D), which is generally stipitate. The i bases of the corona lobes themselves become connate in eight of the ten species. The corona is the most variable and difficult asclepiad structure, especially so in Matelea. The seven most specialized spe- . cies of Dictyanthus have nectaries on the corona lobes. The nectaries re located on the sides of the lobes near the base, are often ringed ith distinctive colors, and are composed of palisade—like secretory pidermal cells (Figure 8D). Matelea tuberosa and M, hemsleyana appear Ct to have nectaries and the poor condition of the flowers of M, amata make it impossible to determine whether or not it has them. On he gynostegium (excluding the corona lobes) there are wings or alae hich correspond, at least in position, to the lateral margins of adj fun the one pem are (Fig rah but Seri are Path mags glam 44 adjacent anthers (Figure 8C). The wings perform the dual, but related, functions of guiding the legs or hairs of the pollinating insects to the corpuscula, which attach the pollen sacs to the insect, and subse— quently in guiding the pollen sacs to the stigmatic, or alar, chambers (Figure 8C). The wings and stigmatic chambers are not as well formed as, for instance, with Asclepias, but at any rate the style apex is stigmatic only in the five spots corresponding to the stigmatic cham— bers (Figure 8C). The structures referred to as "terminal anther ap— pendages" correspond to apical prolongations of the connectives; they are laminar and partly to completely overlie the top of the style apex (Figure 8A). The tip of the style apex can become variously elabo— rated. In Dictyanthus the tip varies only from concave to apiculate, but is of some diagnostic value. Pollinia. A pollinium is, in one form or another, a coherent mass of pollen which is transferred by a pollination vector as a unit. Pol- linia have evolved independently at least three times. Some species of Acacia (Fabaceae) have evolved a sort of pollinium, but pollinia only become highly developed in Apocynaceae and Orchidaceae. In Apocynaceae rudimentary pollinia are found in a few species of Plumerioideae and Apocynoideae, more specialized pollinia are found in Periplocoideae and Secamonoideae, and highly specialized pollinia are found in Asclepia— doideae. Orchidaceae exhibit a similar phylogenetically correlated series of pollinium specializations. Only two loculi of each anther are fertile in Asclepiadoideae. Secretions from the style apex, and perhaps from the anthers as well, aid in fusing the pollen into a solid mass and form the translator arms, or caudicles, and the corpuscula, or glands. One corpusculum is formed at the apex of each pair of anther vim nea pom rvo some wing slit cans sect sacs len stig sery Stud Brow Cuig (194 of p is,‘ fert: haViI Unifr 20nd hYah‘ 1.. Cc disty 45 wings. A translator arm connects each side of a corpusculum to the nearest pollen mass, or pollen sac. A pollinium proper, then, is com— posed of a corpusculum, two translator arms, and half the pollen of the two adjacent anthers, or two pollen sacs. A pollinium is removed when some part of an insect, often a hair, is caught between the anther wings and pulled up to the corpusculum. The corpusculum has a narrow slit which is confluent with the slit between the anther wings and, be— cause of the decreasing width of the slit, fastens itself to the in— sect. When the corpusculum is removed by an insect, the pair of pollen sacs is carried along. When the insect visits another flower the pol— len sac can be caught between the anther wings and deposited in the stigmatic chamber. In Asclepias the broken translator arm has been ob— served to remove another corpusculum, forming chains of pollinia. Studies pertinent to the pollinia of Asclepiadoideae include those by Brown (1833), Corry (1883), Dop (1902, 1903), Galil and Zeroni (1969), Guignard (1903), Liskens and Suren (1969), Richharia (1934), and Volk (1949). The tribes of Asclepiadoideae are essentially defined on the basis of pollinium characters: Asclepiadeae have pendulous pollinia (that is, with the corpuscula borne above the pollen sacs) with uniformly fertile pollen sacs, Ceropegieae have erect pollinia with pollen sacs having a sterile, hyaline margin, Marsdenieae have erect pollinia with niformly fertile pollen sacs, and Gonolobeae have more or less hori— ontal pollinia (see Figure 8C) with pollen sacs having a sterile, yaline section associated with the attachment of the translator arm. n Gonolobeae the sterile, hyaline part of the pollen sac is usually istinctly concave or excavated on one side and the translator arms are ofU crh orh vid rem van desc lini spec icl illL orie in 0 mati fact COnt adl °Var APoe: gina 46 often broad and flat. .The pollinia of Matelea fit this general des— cription but there is still considerable variation in size, shape, and orientation. In order to adequately organize the pollinium variation within the genus it will be essential to have a set of well-defined and readily comparable descriptive terms, but I have not seen enough of the variation to attempt this. Pollen sacs of Dictyanthus can be generally described as obliquely obovate and the corpuscula sagittate. The pol— linia of Dictyanthus are shown in Figure 9 and those of four related species in Figure 10. The pollinium measurements given in the Taxonom- ic Treatment were determined with an ocular micrometer on an epi— illumination apparatus; the measurements were of wet material in the orientation they assume when removed. Ovaries. In all Asclepiadoideae, the gynoecium is composed of two pistils which are separate below but fused at the style apices (Figure 8B); this condition is probably unique to Apocynaceae. As mentioned above, the style apex is stigmatic only in five lateral sites alternate with the anthers. Hypothetically, one pollen sac placed in one stig— matic chamber should result in the complete fertilization of the ovules in one ovary. A second pollen sac, if placed in an appropriate stig— matic chamber, should result in the second ovary being fertilized. In fact, the number of ovaries actually maturing into fruits appears to be controlled by factors in addition to pollination (see Moore, 1946, for discussion relative to Asclepias). Some genera regularly have both varies maturing into fruits but most have only one. I have never seen W0 mature follicles produced by a single flower of Dictyanthus. Pocynaceae are generally considered to have superior ovaries with mar— inal placentation. Baum (1949) and Safwat (1962), among others, have Figure 9. 47 Pollinia of Matelea subgenus Dictyanthus, epi—illumination photographs, x 37. A. Matelea standleyana, Stevens 1392 (MSC); B. M, cerato— petala, Stevens 1245 (MSC); C. M— pavonii, Stevens 1462 (MSC); D. M. hamata, Langlassé 257 (US); E. M. yucatanensis, Stevens C— —158, a cultivated plant of Stevens _ll68 (MSC); .macvaughiana, Faberge Exfl- (TEX); G. M. aenea, StevensF C— 157A, a cultivated plant of Stevens 1145 (MSC); H. M. dictyantha, Stevens 1343 (MSC); I. M. hemsleyana, Stevens C— 162, a cultivated plant of Stevens 1399 (MSC); J. M. tuberosa, Stevens 1468 (MSC). 48 FigL 49 Figure 10. Pollinia of four species of Matelea related to Matelea subgenus Dictxanthus, epi—illumination photographs, x 37. A. Matelea aspera, Stevens 1296 (MSC); B. M, sepicola, Stevens 2038 (MSC); C. M. congesta, Stevens 1462 (MSC); D. M, altatensis, Stevens 2062 (MSC). 1 l I l \ not the 11k ten pom reL for whi( can inf] mant 86mm isti them M tYpe beea Spec theS1 50 ._. ._.__.. noted that the placentation in their material was actually submarginal; the placentation in Dictyanthus also appears to be submarginal. It has likewise repeatedly been repeatedly shown that various members of the family have so—called subinferior ovaries (e.g. Woodson & Moore, 1938; Safwat, 1962). The ovaries of Dictyanthus also bear ovules somewhat dowu into the receptacular tissue and on that account could likely be termed subinferior. This may seem to be of minor consequence, but the potential for producing inferior ovaries is important in evaluating the relationships within Gentianales and related orders. There have been numerous embryological investigations of asclepiads; see Davis (1966) for a nearly complete bibliography. Fruits 32d seeds With a few minor exceptions, Asclepiadoideae always produce fruits which are follicles and seeds which are apically comose. The follicles can become quite woody, as with some species of Marsdenia, or greatly inflated, as with Calotropis, and the surface can be variously orna— i mented. Matelea, excluding Macroscepis, which should stand as a good E genus, and Matelea viridiflora (Meyer) Woodson, which is better placed h in Gonolobus, has follicles which are either smooth or equipped with one or another sort of projection. It appears that follicle character— istics are taxonomically useful but little attention has been given them. I suspect that Woodson's (1941) subgeneric classification of Matelea would have been different had he known more about the different types of follicles. Woodson can hardly be faulted in this case though because the follicles are poorly known in general. In many asclepiad sPecies the fruits mature long after the plant has finished flowering; these plants generally flower near the middle of the wet season, esp dle res not Nit ait hav cie lic ind fru‘ The} aha} wel.‘ See< atté tiv; 51 especially July-August, but the follicles are not mature until the mid— dle of the dry season, mostly February—March. Two related problems result from this. The first is that fruiting collections often cannot be correlated with flowering collections of the same species and sit undetermined in herbaria. The related problem is that collectors tend to avoid fruiting specimens when they realize that they are probably not determinable. The fruits of many species are still unknown. Matelea pavonii, for instance, is the most common species of 222517 anthus, yet mature fruits and seeds are unknown for this species. I have examined about 70 collections and nearly 200 sheets of this spe- cies and have seen one immature follicle and three old, dehisced fol- licles from the previous season. This is also, at least in part, an indication that species such as Matelea pavonii produce relatively few fruits. Seeds are, of course, even more poorly known than fruits. There is little basis for judging the systematic importance of seed characters but they have been described in the Taxonomic Treatment as well as possible on the basis of the available material. Fruits and seeds were collected whenever possible and when the seeds were viable ttempts were made to grow the plants to flowering. Through this cul— tivatiOn, I have been able to make certain flower—fruit correlations, s for instance with Matelea prosthecidiscus Woodson (Stevens, 1975). s fruiting specimens become more determinable, follicle characteris— ics will probably take on greater importance in the taxonomy of atelea. chr dat Moo sou ges dea cor fau dat not pec mim VidI but likr and thar the of t been Q0m; and SPur been View 52 Cytology Although several cytological features could be considered, only chromosome number data are presented here. Table 1 summarizes these data. The table has been compiled primarily from Federov (1969) and Moore (1973, 1974), but a few counts have been derived from other sources. It is important to note that these data are presented in di- gested, not literal, form. Chromosome data are especially difficult to deal with because a certain, probably significant, number of the re— corded counts are inaccurate, either because of faulty taxonomy or faulty counting, or are insignificant for other reasons. Since the data are summarized basically at the tribal level, taxonomy is probably not a significant problem here. Many of the counts, however, are sus- pect, especially those from the older literature. I have attempted to minimize this problem by evaluating the probable credibility of indi— vidual counts. The decisions were admittedly arbitrary in some cases, but I believe this has clarified rather than altered the results. Most likely to be eliminated were old counts which differed from more recent and more reasonable counts. Still, some species appear to have more than one possible number and these have been included; the totals of the individual counts, therefore, are somewhat higher than the totals 0f the genera and species. The few infraspecific taxa counted have been treated as species. For convenience, meiotic counts have been converted to the mitotic equivalent. Counts of artificial polyploids and hybrids have been mostly eliminated. There are undoubtedly some Spurious counts still included and possibly some legitimate counts have been eliminated but the resulting table gives at least a general over— view of the chromosome number data of Apocynaceae. .l.. ~a ‘4w1llyu ((((( (. Ulrlrr | xvhrEfi$£3W . COHUWC—w QVnm MO. Uvnmyu va MWUUWCKUOQ< 0 QINWQFDJC WEOWOEOHCU H H 49. IN. MVHnumrfi 53 v—l H H m mH ON H mmHowmm H mumnwo mH OH wmeuoz wmmvHockooa< mMHowam H H H mnmcwo o m wmouwnuwo r—i I—ll—i Hr—i H N H Ln meome H muwcww m H wmwbcmEmHH< meooam dynamo OH N wmeMHo3smm r-lln Flt-l Hx‘f H q m H m H H q 33me m mumawo mm m ommHHwESHm H 0H mmHoomm H m mumswo 9H m mmocmucoeomcnman «H meowmm wuoooo mH m mmwmeumu mmovHOHuoade r—Ir—l Ch +OOH mm mm mm co oq qq oq om lhm mm «N NN om wH 0H m AGNV muasoo mo kumEEDm pmwmwwoz onommm mnwnoo mnHuH I kHHEmwnsm .AGOHumCmmew how uxwu mmmv wmwumukooa< mo mhmnasc wEoonousJ .4 04:04 r~\<..( rll(t\ r (rdle 54 +OOH Nm mm N m N we H H H N NH H H N Nuoa< moHummm wuwamo wmoncomumm prue . aHHamansm AmenHuaooV H «HAMH 55 moHommm N H a m we H o H m o Ham mm HN OH H H H H m H NH H m m m a we a a H H museum 0mm aw mHmuoH H mmHooam H muonmu H H wmoAOHocoo H m mm o oNH mmHoomm H N n N HH muwaow 05H mH mmonmmopoU +OOH mm mm NH we we «e oq om Imm Nm «N NN oN wH 0H m AGNV mucoou mo mumaadm wwHmeoz moHuwmm muoaou opHnH . sHHammanm AmosoHucoov H oHan 56 The subfamilies and tribes used are according to Wagenitz (1964) except Cerberoideae has been treated as a tribe of Plumerioideae and Secamoneae has been treated as a subfamily. Refer to the section on Relationships for a discussion of the subfamilies. No counts have been reported for Chilocarpeae, Ambelanieae, and Skytantheae, all in sub— family Plumerioideae. Two obvious conclusions can be made from the data in Table l. The more significant of these is that the primary basic number for the fam— ily is most probably §_= 11. About 97% of the genera and 86% of the species have this number (Zn_= 22, 33, 44, 66, 88) on the basis of re- corded counts (but, as mentioned above, some species have more than one recorded number). Six of the tribes or subfamilies have only §_= 11 reported and only one (Allamandeae, with the single genus Allamanda) lacks a report with §.= 11. The apparently secondary basic numbers (35 = 8, 9, 10, and possibly 12 and 23) could be derived from the prima- ry basic number. This is the same conclusion reached by Roy Tapadar (1964) in a study of Apocynaceae in the strict sense. It can also be seen that polyploidy has had some influence on chromosomal evolution in this family. About 36% of the genera and 22% of the species can be as— sumed to be of polyploid origin. Polyploidy occurs especially in Plumerieae, Rauwolfieae, Cerbereae, and Ceropegieae. Polyploids are unknown in nine of the tribes or subfamilies with recorded counts. The most anomalous count is 22_= 8 for Cryptolepis buchanani Roemer & Schultes (Mulay et a1., 1965) in Periplocoideae. This count was made both meiotically and mitotically and there is no good reason to doubt it, but the number is unusual enough that confirmation is much to be desired. Most of the related families have basic numbers similar to 57 those of Apocynaceae and to my knowledge only Gelsemium rankinii Small (Zn_= 8) of Loganiaceae has a number this low. Since it is unlikely that either the number or the species is especially primitive within the family, the number must have been derived through reduction. If this is true, it should be expected that some intermediate numbers will be found in Periplocoideae. This family, as well as most other primarily trOpical families, is poorly known cytologically. Although nearly 10% of the species have recorded chromosome counts, the majority of the counts are from temper— ate representatives and from the horticulturally important (and conse— quently taxonomically inflated) stapeliads (Ceropegieae, in part). There is, for instance, but a single recorded count for the nearly 300 species of Gonolobeae. This count is recorded for a species treated in this thesis, Matelea ceratopetala. The count was made in 1934 and it is very likely that the species was actually Matelea dictyantha. Distribution Apocynaceae are essentially tropical in distribution but there are a few well—known temperate genera, especially Apocynum and Asclepias. Of the five subfamilies recognized here, three (Plumerioideae, Apocynoide— ae, and Asclepiadoideae) are widespread and two (Periplocoideae and Secamonoideae) are restricted to the Old World. Of the four tribes of Asclepiadoideae, two (Asclepiadeae and Marsdenieae) are widespread, one (Ceropegieae) is restricted to the Old World, and one (Gonolobeae) is restricted to the New World. Asclepiadoideae are the most specialized and can be assumed to be derived from Apocynoideae which can in turn be derived from Plumerioideae; Periplocoideae and Secamonoideae are at n m a e a s 1 In. a f 1 r L“. 1 f t 1 t Cl 3 m a 10 LL C LL 00 LL p 58 usually considered to be intermediate stages of specialization between Apocynoideae and Asclepiadoideae (cf. Good, 1956). Raven and Axelrod (1974) suggest that the subfamilies and some of the tribes of Apocyna- ceae were probably differentiated when South America and Africa were closer together. Assuming that the continents have moved in the manner summarized by Raven and Axelrod, this must have been true for the three widespread subfamilies and even for some of their tribes and more broadly circumscribed genera, especially when it is considered that there are very few "weedy" species in the family and Asclepiadoideae, in particular and despite their comose seeds, are generally unsuccess— ful at migrating to islands (cf. Good, 1952). Since the essentials of the distributional patterns of the subfamilies and tribes are relative- ly well—known and there are not likely to be any great revelations in the understanding of their phylogenetic relationships, any explanations for their distributions must be based on existing information. Some aspects of the distributions can undoubtedly be explained by overland migrations across the existing continents and by long distance dispers- al, but in the face of current theories and accumulating data it would be a mistake not to attempt to correlate the modern distributions with the historical arrangements of the landmasses. Although the arguments can become circular, the viability of including continental movement as a factor in distributions can be tested by examining the distributions of many groups of organisms and taking into account such factors as time of origin, dispersibility, and ecological requirements. Biogeo— graphical models including continental movement will be acceptable in the degree to which they fit the geological facts and successfully ex- Plain the distributions of a variety of organisms. ( i .. K s e 1V .1 .L 1 1 e o r f r r 1 r In. 1 m m u L“. 0 l t a A c ml. 6 0 b C a n n a 0 0 0 D. f w d S n A t 59 Periplocoideae are widespread in the warmer parts of the Old World, occupying nearly all the landmasses, except isolated islands, between latitudes 40° N and 40° S (Good, 1952). On the basis of the distribu— tions of the least specialized genera and the total numbers of genera and species, it would appear that Periplocoideae are basically an African group. Furthermore, many of the distinctive vegetative spe— cializations found in the subfamily are adaptive to arid environments. The periplocoid pollinia, at least theoretically, can be thought of as evolutionary intermediates between those of Apocynoideae and Plumeri- oideae on one hand and Secamonoideae and Asclepiadoideae on the other, but they are in their own way highly, sometimes even bizarrely, spe— cialized and not at all functional intermediates, but functional equiv— alents of the asclepiadoid and secamonoid pollinia. Although I have never seen it suggested, my basic reaction is that Periplocoideae do not represent evolutionary intermediates but are a separate evolution- ary line. The line could have had its origin in Apocynoideae near that of Secamonoideae-Asclepiadoideae or it could have had quite a distinct Origin within either Apocynoideae or Plumerioideae. If this separate origin is, in fact, the case, it is easy to speculate that Peri— plocoideae differentiated in Africa after it had become well—separated from South America, possibly associated with the increasing aridity Which commenced in the Miocene (cf. Raven & Axelrod, 1975). The Asian distribution of Periplocoideae is largely the result of a few wide— spread genera and could be accounted for by dispersal from Africa northward and eastward into Asia and subsequently southward into Australasia. The Australasian range south of about 10° S latitude is the result of a single genus, Gymnanthera (Good, 1952). This is all 60 extremely speculative, of course, but would, among other things, ex— plain the absence of Periplocoideae from the New World. Secamonoideae, on the other hand, almost certainly represent an in- termediate evolutionary step toward the more highly specialized Ascle—» piadoideae and must have differentiated at least concurrently with the tribes Asclepiadeae and Marsdenieae, or while migration between South America and Africa was still possible. Although not directly explain— ing its absence from South America, it should be noted that the subfam- ily is small and not particularly diverse. It is normally divided into at least three genera with a total of about 150 species, which are best represented in Africa and Madagascar. The genera are separated mainly on the basis of differences in the style apex and there is reason to believe that the variation is best represented as comprising a single genus (Stevens, 1971). Whether represented by one genus or by three or more, the group currently appears to exhibit little evolutionary poten— tial. A group of this type could be supposed to have either been elim- inated from South America or to have had a limited distribution while Africa and South America were closer. It is not unreasonable to assume that Ceropegieae and Gonolobeae differentiated after South America and Africa were isolated from each other. Ceropegieae, with their striking xeromorphic adaptations, very likely differentiated in Africa in reSponse to increasing aridity. Gonolobeae, with which I am most concerned here, are not easily as- sessed. The major prOblem is that they are poorly known and under— stood. It is possible to identify certain species or groups of species as being derived on the basis of highly specialized features; the sub— genera Labidostelma and Dictyanthus of Matelea, for instance, could be 61 considered rather advanced in the tribe because of their highly spe— cialized corona lobes. Other than with such specific examples I have little idea about what is primitive or advanced in the tribe. Likewise, I have no idea from which group the tribe may have evolved or where it may have evolved. Essentially by default, the tribe can be guessed to have differentiated in South America from something like Marsdenieae or Asclepiadeae. Raven and Axelrod (1974) indicate that although there are a few temperate North American genera, most genera of North and Central American Apocynaceae are South American in origin. Matelea is too poorly known, both as to the relationships within the genus and the numbers and distributions of species, especially in South America, to provide evidence either for or against a South American origin. The major center of diversity of Matelea in North and Central America is apparently in the area of Chiapas and Guatemala (cf. Williams, 1968) and the diversity decreases significantly southeastward in Central America, but remains relatively high northward through most of Mexico. Alternatively, Gonolobeae could have had a tropical North American origin and, either because of its time of origin or its trop— ical nature, did not disperse more widely in Laurasia. The South Amer— ican and perhaps even the temperate North American parts of the range would then be considered relatively recent. Subgenus Dictyanthus ranges from Nicaragua northward to Veracruz on the east side of Mexico and southern Sonora on the west side. The three other related species treated here fall within this range except One extends a little farther north in Sonora. I have seen no SOuth American species which appear to be closely related to this group; the closest affinities I can detect are with Mexican species. Of the two 62 most generalized species of Dictyanthus, one forms the northernmost ex- tension of the range and the other nearly equals the southernmost ex‘ tension. Ecology Matelea subgenus Dictyanthus, as with the majority of tropical as— clepiads, tends to occupy low, open, seasonally dry forests and the ad— jacent grasslands. Since most are vines which do not grow particularly tall, they are not often found in taller forests except in openings. The species, as a group, occur from sea level up to about 2250 m; only Matelea pavonii (mostly 650—1750 m) and M, dictyantha (1250-2250 m) normally occur above 1500 m. Most of the species are apparently toler— ant of a variety of substrates; Matelea dictyantha is at least commonly associated with limestone and the two species on the Yucatan Peninsula occur only on limestone—derived soils. Flowering normally starts in June or July and is completed by September or October. The Yucatan species are the earliest—flowering and Matelea tuberosa, from north— western Mexico, is the latest. Despite the amount of time spent in the field, the pollination biol— ogy of Dictyanthus remains essentially a mystery. I have seen many ex— amples of pollinator activity on tropical species of Asclepias, £22227 ghgm, and Sarcostemma (Asclepiadeae) and Marsdenia (Marsdenieae), but almost never on Gonolobeae. Admittedly I have been more concerned with Collecting good material of as many species as possible in various parts of their ranges than observing individual populations in detail. A realistic study of gonoloboid pollination biology will require a different type of effort. This is one case where taxonomic and 63 ecological investigations are not particularly compatible, at least not when the level of taxonomy at which I am working and the logistic prob— lems of field work in tropical America are taken into consideration. This is unfortunate because the selective pressures giving rise to the elaborate floral features of Dictyanthus are, of course, related to pollination, both in attracting insects to the flowers and in manipu- lating their behavior in such a way that pollination can take place. My best guess is that Dictyanthus is pollinated by dipterans. This suggestion is based on the flower colors and color patterns, on the fact that at least one of the species, Matelea standleyana, produces a faint foetid odor, and on the small amount of nectar which is openly presented. The flower colors are mostly dull shades of red, brown, purple, and yellowish-green and are organized into intricate patterns, features often associated with dipteran pollination, as are foetid odors. The open placement of the nectaries, on the sides of the corona lobes, is compatible with the mostly small feeding organs of dipterans. Nectar does not normally accumulate in the flower, probably because of evaporation, but a small drop can be seen to form if the flower is placed in a moist chamber for a few hours. Drapalik (1970) found that the southeastern United States species of Matelea were pollinated by small flies. Because of the occurrence of certain kinds of compounds, especially cardenolides (cardiac glycosides), asclepiads are often toxic to, and thus protected from, general predators. At the same time, however, certain groups of insects have become specialized predators of ascle— piads and assimilate the poisonous compounds as part of their own de- fense mechanisms. This is a fascinating aspect of asclepiad biology, 64 but again one to which I have contributed little and which is not dis— cussed here. As well as possible, I have noted in the Taxonomic Treat- ment the occurrence and type of predation observable on collected spec- imens and from field observations. I hope eventually to be able to identify the groups of insects which utilize Matelea as a host. Trichomes appear to have special ecological significance in the spe- cies treated here. Valuable reviews on the ecological aspects of trichomes are provided by Uphof (1962) and Levin (1973). These authors provide numerous examples of both speculation and experimentation on the roles of plant trichomes. Although I am not providing experimental evidence, the indumentum is such a prominent aspect of the species that a short discussion seems appropriate. The type and density of trich— omes varies somewhat both between and within the taxa, but they are al- ways dense on the young parts of the shoot tips. Uphof (1962) lists the following as possible functions of dense trichomes on vulnerable plant parts: protection against excessive water loss by transpiration, insulation against marked temperature changes, protection against the blocking of stomata by rain or dew, protection against strong irradia— tion, and mechanical protection. On mature parts, the density of trichomes has been repeatedly demonstrated to influence the type and degree of insect predation. In the case of the species of this revi— sion, the hooked, or uncinate, long trichomes are especially signifi— cant. Hooked hairs of vines often aid in climbing (Uphof, 1962), but this is not the case with these species because, with the exception of Mgtelea aenea and M, yucatanensis, hooked hairs are not found on the ,internodes (the long straight hairs of the stems are often somewhat re— flexed and could be of some benefit in this regard), but are found on 65 various other parts. Levin (1973) gives several examples where hooked trichomes have been shown to be a specific defense against herbivorous insects, especially soft—bodied insects which are likely to be punc— tured as well as snared by the legs. The example by Gilbert (1971) where the hooked trichomes of Passiflora adengpoda snared and punctured the larvae of heliconiine butterflies, specific predators of Passiflora (Ehrlich & Raven, 1964), is especially interesting because certain lep— idopteran groups are among the most successful asclepiad predators. It is more difficult to suggest a function for the glandular trichomes of Matelea. If they actually have a defensive function, they must contain some chemical deterrent to insects. Since the glandular trichomes are not secreting visible substances from their cells, they could be se— creting volatile deterrents or they could contain nonvolatile sub— stances which serve as gustatory repellents (Levin, 1973). In this re- gard, it should be remembered, as Levin notes, that insects often have chemoreceptors on their legs as well as on their mouthparts and do not necessarily need to commence feeding to be chemically repelled. It should be possible to test some of these possible defense funtions by experimentation with natural asclepiad predators. RELATIONSHIPS Family—subfamily Although the phylogeny of angiosperms is still largely speculative, and current students differ somewhat in their interpretations, there are many examples where the relationships seem undeniable. It seems clear that the phyletically large and diverse Loganiaceae (in the broad sense but excluding Buddlejaceae) are the remains of an evolutionary line that has produced several modern groups. Loganiaceae can and have been divided into a number of smaller families, and these probably rep— resent lines which have maintained more of the primitive characters and have been generally less successful. Lines which have diverged farther and been more successful are represented by Gentianaceae, Apocynaceae (in the broad sense), and Rubiaceae. Menyanthaceae, Buddlejaceae, Ole- aceae, and perhaps Caprifoliaceae (excluding Sambucus and Viburnum) may also be derived directly from proto—loganiaceous ancestors, but the relationships seem less clear. To substantiate my own speculations on this subject would require a discussion of characters which is hardly appropriate here. The discussion would be hindered at any rate by the lack of data on the nature and distribution of some of the most impor- tant characters. Suffice it to say that I believe Loganiaceae (s313), APocynaceae (3.1.), Gentianaceae, and Rubiaceae are intimately related and have their common origin best represented in modern Loganiaceae and 66 67 that there are very likely other families directly but somewhat more distantly related. A question more pertinent to this thesis is whether or not Asclepi- adaceae should be separated from Apocynaceae. If we ignore the fact that some of the more primitive members of Apocynaceae (§,§,) are hard— ly distinct from Loganiaceae <§2l°)’ almost everyone, with the possible exception of Nolan (1967), would agree that what is represented by Apocynaceae and Asclepiadaceae (including Periplocaceae) is a single taxon at some classificatory level; the cumulative similarities within this group far outweigh the differences. The major subdivisions within the group are also generally agreed upon and these can be easily ar— ranged into a linear sequence based on the degree of specialization for entomophily, though this does not necessarily represent a phylogenetic sequence. The question resolves to whether or not this obviously natu- ral group, which can be divided into several well-defined and probably natural subgroups, should be divided into two or more families. Robert Brown (1810) was the first to separate Asclepiadaceae from Apocynaceae. Since that time it has become traditional to recognize at least these two families; only Hallier (1912) and more recently Thorne (1968), among leading phylogenists, have recombined the two. Even some of the phylogenists who accept the distinction, however, admit that it is ar— bitrary. Cronquist (1968), for instance, states ". . . there is a step by Step gradation of characters, so that the line between the two fami— lies is only arbitrary established." I have no basic objection, how- ever, in arbitrarily delimiting families and hence I cannot agree with Thorne's (1973) justifying the recognition of a single family on the basis of the width of the gap between the two classically recognized 68 families. One need only compare current phylogenetic systems to real— ize that it is still difficult to put families into their proper order; it is even more difficult to compare families on the basis of the phy- letic gaps between them. My contention that this group is best repre- sented as a single family is, nonetheless, partially based on phyletic gaps, but only on gaps within the group, and ignoring how these gaps may compare with those separating other families. As I evaluate the group, five distinct subgroups can be recognized. The largest of the gaps among them is that which separates the two classically recognized subfamilies of Apocynaceae (Plumerioideae and Apocynoideae). If I have correctly evaluated the subgroups and their distinctions, then the classical separation of the two families inaccurately describes the relationships within the whole group. There are three more or less interrelated choices in remedying the situation. The first would be to continue to recognize two families, but to redefine them by raising Plumerioideae to family rank and adding Asclepiadaceae to Apocynaceae for the second family. This distinction between the two families would still be arbitrary, but would more accurately represent their relation— ship. The second choice would raise Plumerioideae to family rank and would also raise some of the other subgroups to families as well, up to a total of five families. This tendency has already been partially in- dicated by the often recognized family Periplocaceae. The last choice, and the one I obviously prefer, is to recognize but a single family, this with five subfamilies. The difference between recognizing one family with five subfamilies and recognizing five separate families cannot be objectively evaluated; this is primarily a matter of individ~ ual preference and convention, and I am more comfortable with a single 69 family. The family Apocynaceae as I conceive it then, is comprised of the subfamilies Plumerioideae, Apocynoideae, Periplocoideae, Secamon— oideae, and Asclepiadoideae. A second objection I have to the classical distinction between Apocynaceae and Asclepiadaceae is the implication it carries that the latter is monophyletically derived from the former. While the subfam— ilies recognized here are probably natural and may actually represent a single line of evolution, it would be difficult on the basis of cur— rent knowledge to argue that they are derived from one another in a linear sequence. The most advanced members of Apocynoideae and Peri— plocoideae, for instance, are much too specialized to give rise to the succeeding subfamilies. It would not be surprising, as mentioned above, to find that Periplocoideae had a separate origin, within either Plumerioideae or Apocynoideae, from that of Secamonoideae and Asclepi- adoideae. The recognition of a single family with five subfamilies implies only that the group, as a whole, had a single origin. It is significant to add that the two most comprehensive studies of the floral anatomy of the group, those of Demeter (1922) and Safwat (1962), both concluded that the group is best represented as a single family. Safwat's conclusion is essentially identical with my own. The closing paragraph of his summary is as follows: In View of the present study, as well as from investigations of others, I am inclined to believe that phylogeny is better portrayed by combining Asclepiadaceae and Apocynaceae into a single family and re—subdividing the group into five subfamilies according to the de- gree and kind of specialization of the translator apparatus and the relationship between the pollen tetrads to one another at the time of maturity, thus: Plumerioideae, Echitoideae (or Apocynoideae), Periplocoideae, Secamonoideae and Asclepiadoideae. 70 Thorne's (1968) treatment differs only in the recognition of an ad— ditional subfamily, Cerberoideae. This follows Wagenitz (1964) and I cannot evaluate this further subdivision of Plumerioideae at this time. In summary, my recognition of Apocynaceae in the broad sense, while contrary to most current interpretations, is neither without precedent nor without justification. I plan in the future and in a more appro— priate place to provide additional justification on the basis of a careful evaluation of a broad range of characters. Tribe—genus The subfamily with which this thesis is most concerned is Asclepia— doideae. It is most often, and I tentatively think most accurately, divided into four tribes: Asclepiadeae, Marsdenieae (=Tylophoreae), Ceropegieae (including Stapelieae), and Gonolobeae. Each of these tribes is relatively specialized in one way or another and I can only generally speculate on their interrelationships. Asclepiadeae and Marsdenieae are the larger (ignoring the extreme splitting of the stapeliads) and the more widespread of the tribes (both tribes as well as several of their constituent genera occur in both the Old and New Worlds). Ceropegieae tend to occupy arid areas and exhibit a strong tendency towards succulence, culminating in the remarkable stem—succu- lent stapeliads. Ceropegieae are apparently derived from Marsdenieae, with which they are sometimes combined (e.g. Good, 1952), and are res— tricted to the Old World. Gonolobeae have the smallest number of gen— era and species and are restricted to the New World. The affinities of Gonolobeae to the other tribes are still unclear to me. Woodson (1941) recognized three genera of Gonolobeae in America: —i—"Fm"”m H" -.a- --_.__..-- t . e i _ 71 Matelea, Gonolobus, and Fischeria. Although Woodson did not at the time consider the South American species, it is clear from the broad circumscription of the genera, especially Matelea, and from his later work on South American asclepiads, that he would have recognized few, if any, additional genera of Gonolobeae. I would recognize a few ad- ditional genera, but Woodson's concepts of Matelea and Gonolobus still clearly form the core of the tribe. Subgenus—species Even if the most discordant elements are removed, Matelea probably still contains nearly 200 species, or roughly two—thirds of the species of Gonolobeae. Woodson (1941) divided the genus into a number of not entirely satisfactory subgenera and sections, many of which had been previously recognized as genera. While I do not pretend to understand all the complexities of the genus, I can vaguely recognize a large group composed of species of the following of Woodson's subgenera: Dictyanthus, Pachystelma, at least part of Chthamalia, and parts of Macroscepis, Heliostemma, and Matelea. The tendencies I see represent— ed in this group are to have a corona, lobed or unlobed, which is com- posed of a single series of enations which are connate laterally and adnate to the corolla, to have a distinctly mixed indumentum which often has at least some of the trichomes uncinate, and to have folli— cles which are tuberculate. The species of Dictyanthus, as well as the other three species treated here, have strongly lobed corollas and Otherwise fit the above tendencies quite well. As may be surmised from reading this thesis, Woodson was prone to making errors of detail in his taxonomic work. The value of his 72 revolutionarily conservative approach to generic limits of the ascle— piads has, nevertheless, been repeatedly confirmed (at least in large part). Woodson was "very reluctant to merge Dictyanthus with Matelea" but chose to do so in large part because M, altatensis provided "a very suggestive link with either Pachystelma or Eumatelea, according to one's viewPoint at a particular time." Although I do not agree with the circumscription of some of his subgenera, including Pachystelma and "Eumatelea," and have, on technical grounds, removed Matelea altatensis from subgenus Dicgyanthus, I still believe Dictyanthus is inextricably linked to the larger concept of Matelea and is best kept in the genus. The virtue of Woodson's conservative concept is thus reaffirmed. I have defined Dicgyanthus, as nearly as possible, in the same manner as Woodson, which was the same as the generic concept. In a natural key to the current subgenera of Matelea, Dictyanthus could be identified as having digitate corona lobes with their axes entirely adnate to the corolla, simple inflorescences, a mixed indumentum with at least some of the trichomes glandular and at least some of the long trichomes un- cinate, and narrowly fusiform follicles with thickish projections. The "faucal annulus" Woodson refers to in describing Dictyanthus simply does not exist. Pollinium characters are apparently important, though not as overwhelmingly important as Woodson supposed, but they have not been uniformly enough described to be of much use at this time. Dictyanthus could be redefined, either to exclude Matelea hemsleyana and M, tuberosa or to include additional related species, especially the three treated here, but I think it best not to consider doing this until more of the related species are studied. Dictyanthus, or any of the other subgenera, is best circumscribed in the context of all the 73 related subgenera. Most of the genus is not yet well enough studied to provide this kind of perspective. This is essentially my basis for not assigning the three non-Dictyanthus species treated here to any exist- ing subgenus. With respect to a broader study of the genus, it should be noted that there will need to be a better set of descriptive terms for the coronas, trichomes, and pollinia to make the descriptions more comparable. Within Matelea subgenus Dictyanthus, the relationships, on a sub- jective basis, seem relatively clear. Evolutionary changes in Matelea are most strikingly exhibited in the shape of the floral structures, particularly the coronas. The changes that take place in the shape of these structures, while of paramount importance in reproductive biol- ogy, are not particularly suited to objective measurement and, in fact, almost defy description. In order to employ an objective technique, these characters would need to be described and properly weighted (non— floral characters can in no sense have the same evolutionary signifi- cance as floral characters in the asclepiads). Rather than attempting to employ an objective technique and then comparing the results with my subjective evaluations, I will simply summarize the relationships as I see them. Subgenus Dictyanthus, if defined in the strictest reasonable sense, would include seven species (species number four through ten in the Taxonomic Treatment). Among these, Matelea aenea and M, yucatanensis are the most closely related and obviously derived from something like the other five; these two are unique enough in several respects that they may have been isolated for some time and show no special affini— ties with any particular one of the other five. On a character by 74 character basis, however, they would be most similar to Matelea giggyf EEEEE’ which is, incidentally, the only other species which commonly occurs on limestone-derived soils. The remaining five species are all closely related, but among them Matelea pavonii and M. ceratopetala form the closest pair, M, macvaughiana is only a little more distantly related, and M. standleyana and M. dictyantha slightly more distant. Although Matelea pavonii should not be considered directly ancestral to the other six species, they may have been derived from something simi— lar to it. This species appear to be the most generalized and variable member of the seven core species of Dictyanthus and is also the most common and widespread. When circumscribed somewhat more broadly, as done here, Dictyanthus includes three additional species (species number one through three in the Taxonomic Treatment). Matelea hemsleyana and M, tuberosa are rela— tively distant from the seven species described above and definitely more generalized. These two, especially Matelea hemsleyana, appear to be intermediate between Dictyanthus in the strictest sense and the rest of Matelea. Matelea hamata apparently bridges the gap between M, £3; berosa and M. hemsleyana and the seven species noted above. Unfortun— ately it is very poorly known and additional collections are much to be desired. As hinted above, these subgroups could be taken into account, both in defining the subgenus and introducing additional infrageneric categories, but this is pointless outside a broader study of Matelea. TAXONOMIC TREATMENT Matelea subgenus Dictyanthus Matelea subgenus Dictyanthus (Decaisne 2E de Candolle) Woodson, Ann. Missouri Bot. Gard. 28: 236—237. 1941. Dictyanthus Decaisne lg de Candolle, Prodr. 8: 605. 1844. Type species: Dictyanthus pavonii Decaisne i3 de Candolle. Tympananthe Hasskarl, Flora 47: 258—259. 1847. Type species: Tympananthe suberosa Hasskarl. Rytidoloma Turczaninow, Bull. Soc. Imp. Naturalistes Moscou 25(2): 319—320. 1852. Type species: Rytidoloma reticulatum Turczaninow. Plants erect, trailing, or twining, herbaceous or woody, with or without a woody or fleshy caudex. Woody parts typically with thick, fissured, corky bark. Indumentum variable and often mixed; trichomes multicellular, uniseriate, simple, straight or uncinate, of three gen- eral types: short nonglandular, short glandular, and long nonglandu— lar. Leaves ovate in general outline, apices mostly acuminate to at- tenuate, bases lobate, with acropetiolar glands; exstipulate but with an interpetiolar finge of long trichomes and glands. Inflorescence extra-axillary, a condensed, simple, helicoid cyme or reduced to a Single flower with or without an apparent peduncle. Calyx 5—lobed nearly to the base, with one or two glands below each sinus within. Corolla deeply to shallowly campanulate; tube convoluted, with raised 75 _7—mxm—_-~ 76 parts opposite corona lobes and sacs formed between them. Corona dig— itately 5-lobed, lobes connate below or not, adnate to gynostegium and adnate for their entire length to corolla. Gynostegium stipitate, apex pentagonal and concave to apiculate, terminal anther appendages cover— ing margin of apex. Corpusculum sagittate; translators winged, hardly distinct from pollen sacs; pollen sacs flattened, excavated and hyaline along upper margin, obliquely obovate. Follicles fusiform, with few to numerous, thick to thin, straight to arcuate projections. Seeds obov- ate, flattened, with a raised, smooth or radially grooved, entire or toothed margin, surface otherwise verrucate to rugose, light to dark brown; with a white apical coma. Includes species number l-lO. Matelea subgenus unassigned For a discussion of the status of these species, consult the section on Relationships. Includes species number 11—13. As a group, these species differ from subgenus Dictyanthus as follows. Plants always twining. Inflorescence simple or more often compound. Corolla urceolate or shallowly campanulate; tube convoluted or not. Corona lobes partially or entirely free from corolla. Gynostegium apex slightly or not at all convex, terminal anther appendages covering from nearly half to the entire apex. Pollen sacs tending to be smaller, broader, and more angular than those of subgenus Dictyanthus. Notes 22_characters used iE_Taxonomic Treatment For more detailed discussions, consult the appropriate sections under Morphology and anatomy. The most critical points are listed below. 77 (1) The description of the indumentum has been simplified and, to a certain extent, generalized by the convention of referring to all trichomes as either shggg) glandular, or 123g and modifying these terms as appropriate. These trichomes are all uniseriate and multicellular and can have straight or uncinate tips. Short trichomes are less than 0.1 mm long, typically about 0.05 mm. Short trichomes on the inner surface of the corolla, when present, are somewhat different in form and have a glassy appearance when dried. Glandular trichomes are the same length to slightly shorter than the short trichomes, with which they are almost always mixed, and have a short stalk, an inflated mid— dle, and a short apiculum. The glandular trichomes are probably not actually secretory, but the inflated part typically collapses on dry— ing, giving these the appearance of normal capitate glandular trich— omes. Long trichomes are more than 0.1 mm long, typically much more. The maximum length of long trichomes is given only for the stem; they tend to be somewhat shorter on other structures. When only long trich— omes are present on a structure, as is often the case with the leaf blade, they often occur in two discrete lengths, giving much the same appearance as mixed long and short trichomes. (2) The terminology used for describing the surfaces of the leaves and seeds is according to Stearn (1966). (3) The leaves are described essentially according to Hickey (1973). The same terminology is employed to described the shape of the bracts, calyx lobes, and corolla lobe apices. The leaf length has been considered to be the length of the midrib. In all cases the leaves are described on the basis of the largest leaf of each specimen examined. The largest leaves, especially on specimens of the erect species, tend 78 to be near the middle of the stems; the lower leaves tend to be broader and the upper leaves tend to be narrower. Even using this method, the leaves are markedly variable in size and shape. (4) The inflorescence and floral characters are described only on the basis of examples in anthesis. The bracts are described on the basis of the largest bract of each inflorescence. The first bract (opposite the first flower) tends to be the largest and the subsequent bracts somewhat smaller. (5) For descriptive purposes, I have considered the corolla lobes to be distinct from the limb. The corolla, then, is composed of the tube, the limb, and the lobes. The descriptions of flower colors have been much simplified. In general, only the basic color pattern of the corolla has been described. This color pattern applies only to the inner surface of the corolla and considerable care should be exercised in attempting to discern the pattern by examining the outside of pressed flowers. (6) Measurements of pollinia are taken in lateral View, that is, the depth or thickness is ignored, and in the normal orientation they assume when removed. The length of the pollen sac is taken from the point of attachment of the corpusculum to the tip, including, there— fore, the translator arms or caudicles. This has been done because in most species of Gonolobeae there is no sharp demarcation of the trans— lator arms from the pollen sacs. 79 Artificial key 59 species 1. Corolla tube with parallel vertical lines, these only occasion— ally with a few cross—connections (sometimes difficult to see in the small urceolate corolla of M, sepicola) 2. Corolla lobes less than 8 mm long; corona lobes less than 4 mm long, not basally connate 3. 3. Corolla deeply campanulate or tubular, with dense short trichomes on limb and lobes within, corolla tube much exceeding corona lobes . . . . . . . . . . . 2. M. tuberosa Corolla urceolate, entirely glabrous within, corona lobes equalling corolla tube . . . . . . . 12. M, sepicola 2. Corolla lobes more than 8 mm long; corona lobes more than 4 mm long, basally connate 4. Corolla base—sinus length more than 12.5 mm, with a nar— row band of short trichomes around corona lobes within; long trichomes of peduncles and pedicels mostly uncinate; twining woody vines without thickened caudices 4. M. pavonii Corolla base—sinus length less than 12.5 mm, glabrous around corona lobes within; long trichomes of peduncles and pedicels straight; erect or weakly twining herbaceous vines with thickened caudices . . . . . . 5. M. macvaughiana 1. Corolla tube with circular lines, distinct reticulations, or without a distinct pattern 5. Corolla entirely glabrous within; corona lobes basally con— nate and forming a distinct cup; inflorescence bracts more than 1.5 mm long, elliptic in general shape . . . 3. M. hamata 80 . Corolla with dense short trichomes within, at least on limb; corona lobes basally connate or not but not forming a distinct cup; inflorescence bracts less than 1.5 mm wide, linear or ovate in general shape 6. Corona lobes spathulate, with prominent, glistening, pur— plish—black, deeply rugose tips; long trichomes of inter— nodes uncinate; plants of Yucatan Peninsula 7. Corolla lobes 7—12 mm long, length to width (sinus— sinus) ratio greater than 0.80, margins revolute, limb and lobes plane or slightly reflexed; corolla densely grayish—purple—reticulated . . . . . . 10. M. yucatanensis Corolla lobes 5—9 mm long, length to width (sinus—sinus) ratio less than 0.80, margins not revolute, limb and lobes ascending; corolla yellowish—green when fresh, sometimes drying darker and somewhat reticulated 9. M. aenea Corona lobes of various shapes but never modified as above; long trichomes of internodes rarely uncinate; plants not of the Yucatén Peninsula 8. Corolla campanulate, base to sinus length 7 mm or greater, margins strongly revolute; corona lobes more than 4.5 mm long, linear to linear—spathulate, adnate to corolla; inflorescence a simple cyme 9. Corolla tube with circular lines . . . 6. M. standleyana 9. Corolla tube with a reticulate pattern or without a distinct pattern 81 10. Gynostegium apex apiculate; corona lobes 8 mm long or longer; twining vines without thickened caudices; plants from southeast of the Isthmus of Tehuantepec . . . . . . . . . . . . . . . . . . . . 7. M, ceratopetala 10. Gynostegium apex shallowly concave; corona lobes 8 mm long or shorter; erect or weakly twining vines, mostly from thickened caudices; plants from northwest of the Isthmus of Tehuantepec . . . . . . . . . 8. M, dictyantha 8. Corolla shallowly campanulate to nearly rotate, base to sinus length 11 mm or less, margins slightly or not at all revolute; corona lobes less than 4.5 mm long, shape various but not linear or linear-spathulate, adnate to corolla or not; inflo— rescence a simple or compound cyme ll. Peduncles 4 mm long or shorter; pedicels 5 mm long or shorter; inflorescence a simple cyme; leaf blades 34 mm long or shorter; erect or weakly twining vines with thick- ened caudices . . . . . . . . . . . . . . . . l. M, hemsleyana ll. Peduncles 3 mm long or longer; pedicels 5 mm long or longer; inflorescence a simple or more often a compound cyme; leaf blades 31 mm long or longer; twining vines mostly without thickened caudices 12. Corona lobes triangular in outline, not inflated, ad- herent to corolla to tip; corolla distinctly convoluted; plants of Sinaloa and Sonora . . . . . 11. M, altatensis 12. Corona lobes ovate in outline, inflated, free from corolla above; corolla only very slightly convoluted; plants from Jalisco and Veracruz to Nicaragua 13. M, aspera Species treatments (subgenus Dictyanthusl l. Matelea hemsleyana Woodson, Ann. Missouri Bot. Card. 28: 237. 1941, based on Dictyanthus parviflorus Hemsley. Dictyanthus parviflorus Hemsley, Biol. Centr. Am. Bot. 2: 329. 1882, 222 Matelea parviflora (Torrey) Woodson. Lectotype: Chiesbreght 663 (K? not seen, lectotype; GH! MO! NY! iso— lectotypes). Syntype: Ghiesbreght fimfl- (K! syntype; GH! L! P! probable isosyntypes). Dictyanthus prostratus Brandegee, Univ. Calif. Publ. Bot. 7: 329. 1920, 222 Matelea prostrata (Willdenow) Woodson. Type: Purpus §411.2,2,1 (UC! holotype; GH! MO! NY! US, 2 specimens! VT! isotypes). Matelea diffusa Woodson, Ann. Missouri Bot. Card. 28: 236. 1941, based on Dictyanthus prostratus Brandegee. Plants erect to trailing or rarely weakly twining. Stems 20—60 (-90) cm long, with a woody caudex to 4 cm long and 2 cm wide, this with thin to thick corky bark, also often with short woody stems above caudex, these with or without corky bark, otherwise herbaceous and lacking bark, with dense short and glandular trichomes and sparse to dense, mostly straight long trichomes to 3 mm long. Leaf blade ovate to very—wide—ovate, 13—34 mm long, 13-36 mm wide, with mostly uncinate long trichomes and also often with scattered glandular trichomes below, 1The use of 2,2. (2:9 parte) indicates cases where different taxa are represented on different sheets with the same collection number or where sheets with the same collection number are given different col— lection data, i.e., those cases where normal collection practices would indicate different collection numbers. Cases of more than one taxon on one sheet, i.e., mixed collections, are mentioned parenthetically where judged significant. 83 surface smooth, smaller veins sharply raised below, apex acuminate to attenuate or rarely obtuse, base lobate, lobes overlapping to diver— gent, with 2—6 (-8) acropetiolar glands, margin often somewhat thicken— ed and revolute; petiole 7-18 (—26) mm long, with dense short and glandular trichomes and sparse to dense, mostly uncinate long trich— omes. Inflorescence a simple cyme; peduncle 1—4 mm long, with dense short and glandular trichomes and sparse to dense, straight or uncinate long trichomes; bracts linear or lorate to lanceolate, 2—4 mm long, with indumentum of leaf or nearly glabrous; pedicel 3-5 mm long, with indumentum of peduncle. Calyx lobes narrow—ovate or occasionally lan- ceolate, 4—6 mm long, 1.5—2.5 mm wide, apex acute to attenuate, with one gland below each sinus, abaxial surface with scattered glandular trichomes and scattered to dense, straight or uncinate long trichomes, adaxial surface glabrous. Corolla shallowly campanulate, base to sinus length 3—6 mm, limb not distinct, margin slightly revolute; lobes (3—) 4—6 (-7) mm long, apex acute or occasionally rounded, plane or slightly reflexed at tip, margin slightly revolute; glabrous within except with dense short trichomes on limb and lobes and these sometimes extending down raised ridges within tube, indumentum on outside of straight long trichomes or sometimes limb and lobes nearly glabrous; tube convoluted, with raised parts opposite corona lobes, forming shallow pockets be- tween them, with corona lobes in distinct pockets in bases of raised parts; moderately to densely brownish—purple—reticulated, becoming pale purple on and around corona lobes. Corona lobes 1.0—1.5 mm long, basi— cally short—spathulate with an acute apex, main axis adnate to corolla and adaxially adnate to gynostegium, upper surface with a narrow ridge which extends as a short spur to edge of gynostegium. Gynostegium 84 1.0—1.5 mm high and 1.5-2.0 mm wide at apex, short—stipitate, apex con- vex and slightly bilobed, terminal anther appendages covering margin of apex. Corpusculum 0.18—0.22 mm long, 0.08-0.10 mm wide, pollensacs 0.58-0.91 mm long, 0.26-0.34 mm wide. Follicles fusiform, 48-70 mm long, 10—18 mm wide, green with white markings, glabrous or with sparse short and glandular trichomes, with 28—54 projections, these to 2 mm long, arcuate and somewhat reflexed proximally, straight and leaning forward distally. Seeds obovate, ca 4 mm long, ca 3 mm wide, with a raised margin, this irregularly toothed distally, inside this margin slightly convex on one side and slightly concave on opposite side, both sides verrucate, concave side with a narrow ridge from apex to near center, apparently light brown; coma ca 25 mm long. Figure 11. DISTRIBUTION AND ECOLOGY. Found in Michoacan, state of Mexico, Morelos, Veracruz, Chiapas, Guatemala, and El Salvador. Figure 12. The gap in the range at the Isthmus of Tehuantepec is as expected on the basis of elevation, but the abSence of collections from Puebla and northern Oaxaca is more difficult to explain. Although parts of this region are too arid and parts are on limestone or dolomite, substrates not yet known for this species, there should be suitable habitats. Perhaps additional collecting in this area will show a more continuous distribution. Collected at elevations of from somewhat below 800 m to nearly 2600 m, but mostly 1000-1500 m. From the meager data on edaphic conditions, Matelea hemsleyana can grow at least on volcanic cinder and rocky clay soils. Found growing on slopes and hills, mostly in grass— lands, but sometimes in open pine-oak forests. Flowering mostly June- September, but flowering specimens also collected once each in April and November. Specimens with mature—sized fruits collected 85 Figure 11. Matelea hemsleyana (draWn from Stevens C—l62, a cultivated specimen of Stevens 1399). A. habit, x 0.6; B—C. flowers, x 4.6; D. pollinium, x 35. 86 August—December. The only grazing insect damage observed was a few chewed leaves on Dressler §_Jones 252. Hemsley 4471 and some specimens of Pringle 13112 have a number of their flower buds modified into galls, these probably being caused by parasitic dipterans. COMMON NAMES AND LOCAL USES. Steyermark gagzg gives the local name chinuna and Steyermark 5923; gives the local name pegapega; both these collections are from Guatemala. Calder6n 1911, from El Salvador, gives the local name yulpate. No local uses were found. DISCUSSION. In combining what Woodson considered to be two species, I was left with two names which were equivalent with respect to priori— ty. My choice is simply arbitrary, but does agree with the priority of the original epithets under Dictyanthus. The lectotypification of Matelea hemsleyana, however, was more of a problem. The part of the protologue referring to specimens reads, "Mexico, Cuernavaca and Chiapas (Ghiesbreght, 663;) Hb. Kew." At first glance, it would seem that the type, and thus the holotype, is designated as Ghiesbreght §63_at K, but by indicating both Cuernavaca and Chiapas, he was implying that he had seen specimens from both localities. Further, I have seen Ghiesbreght specimens from both localities. Specimens of Ghiesbreght 6§3_(GH, M0, NY) bear a printed label with, "Chiapas, etc.: coll. Dr. Ghiesbreght ann. 1864—70." and the specimen at CH bears an additional hand—written label with more detailed collection data, including a locality appar— ently in Chiapas. The other Ghiesbreght specimens of this species (GH, K, L, P) are unnumbered and bear the printed label, "MEXIQUE, Province d‘Oaxaca. M. Ghiesbreght, 1842." The K specimen of Ghiesbreght E32: has, as is often the case, the, "province d70axaca" crossed out and 87 replaced with, "Cuernavaca," apparently in Decaisne's hand. I can only conclude that specimens of both collections were seen by Hemsley and that he implicitly cited both in the protologue, thus requiring lecto— typification. Since the rest of the protologue gives no details which would indicate Hemsley's intent, I believe Ghiesbreght fig; is less amr biguous as a lectotype than Ghiesbreght £32: Unfortunately, I have seen only the latter represented by a specimen at K; whether the former exists there or not, I still believe it is a preferable lectotype. Ad— ditionally, though it does not affect the choice of a lectotype, there are two elements within the species which may eventually warrant recog— nition, as discussed below, and the Ghiesbreght collections represent both elements. It is therefore especially important to the stability of the nomenclature that the type of this name be fixed. There is also a potential source of confusion as to the type of Dictyanthus prostratus. There are at least two collections of this species bearing the type number, Purpus §4;l, One is from Acax6nica (Aug 1919) and the other is from Barranca de Panoaya (Sep 1920). One additional specimen with this number bears the date 1920 and no locali— ty data. This problem apparently results from the fact that the Purpus collections were not numbered chronologically as collected, but rather much later (Sousa S., 1969, p. 15). The protologue, however, specifi— cally gives the type locality as Acasonica [Acax6nica] and the acces- sion number at UC which corresponds to that specimen. The Barranca de Panoaya collection was not mentioned in the protologue and therefore has no status as a type. The plants of this species collected from the part of the range cen— tered around the state of Morelos differ somewhat from the plants in 88 the rest of the range. Woodson, according to his annotations, consid— ered the Morelos plants to be Matelea hemsleyana and the others M, 611— Eggg. Standley and Williams (1969) also considered the plants from southern Mexico to El Salvador to be Matelea diffusa, but perhaps with— out seriously considering the plants from Morelos. Standley (1924) considered the two species to be synonymous. The plants from around Morelos tend to be shorter and more erect, to have thicker caudices, larger, more distinctly veined leaves, larger flowers (to nearly twice as large), and proportionately longer corona lobes. In describing Dictyanthus prostratus (=Matelea diffusa), Brandegee considered it to be different from 2. parviflorus (=M, hemsleyana) in having, "five mi— nute scales attached to the middle of the gynostegium representing an inner corona.” I have found no such character. The "scales" to which he referred were most likely the remains of the attachments of the co— ona lobes to the gynostegium, which are typically torn free when the flower is flattened. Despite the differences described above, I have found no character or set of characters which will faithfully differen— tiate the two elements. It may well be that further collection of ade— quately preserved material will demonstrate that some level of taxonom- ic recognition is preferable, but on the basis of the currently avail— able material, I do not believe that it is warranted. SPECIMENS EXAMINED. MEXICO. MICHOACAN: Morelia, Cerro Azfil, 2300 m, Nov 1911 (fl), Arsene 6,6, (G); ca 20 mi S of Uruapan on Hwy 37, 3300 ft, 7 Sep, year not given (fl), Oliver E£.§l: 666 (MO). MEXICO: Dist. Temascaltepec, Cajones, 2580 m, 15 Sep 1932 (fl), Hinton 1699 (C, K, MO, NY, US); Dist. Temascaltepec, Vigas, 1080 m, 15 July 1933 (sterile), Hinton 4328 (CH, MICH, NY, UC, US); Dist. Temascaltepec, 89 Ixtapan, 1000 m, 2 Aug 1933 (fl), Hinton 6611 (G, M0, NY, US); Dist. Temascaltepec, Carboneras, 20 Aug 1935 (f1), Hinton 6668 (CH, MO, NY, 2 specimens, PH, US, WTU); 5 km al SW de Temascaltepec, sobre la carre- tera a Tejupilco, 1900 m, 4 Sep 1965 (fl), Rzedowski 16616 (ENCB). MORELOS: Cuernavaca [written above "Province d‘Oaxaca" of printed label of K specimen], without date (fl), Ghiesbreght §:2-: syntype of Dictyanthus parviflorus (GH, K, L, P; not necessarily all of the same collection); Cuernavaca, hillsides, 5000 ft, 28 July 1896 (apparently fl) and 18 Sep 1896 (apparently fl & fr), Pringle 6616 (BKL, CAS, ENCB, F, G, 3 specimens, GH, ISC, K, MASS, MEXU, MINN, M0, MSC, ND, NY, P, PH, UC, US, 2 specimens, VT); plains near Cuernavaca, 5000 ft, 10 Sep 1903 (fl & fr), Pringle 11611 (F, L); hillsides near Cuernavaca, 5000 ft, 20 July 1904 (fl), Pringle 16111_(ARIZ, CAS, F, GE, L, MICH, MO, MSG, PH, SMU, US, VT); Cuernavaca, 14 Aug 1906 (fl), Pringle 6,6. (VT). VERACRUZ: region of San Andrés Tuxtla, Cerro Mono Blanco, NW of Cate— maco, 2 Sep 1953 (f1), Dressler 6_666§§_161_(MICH, M0, NY, UC, US); Acanoxica or Acasonica [Acax6nica], Aug 1919 (fl), Purpus §ill.EnB', type of Dictyanthus prostratus (GH, M0, NY, UC, US, 2 specimens, VT); Barranca de Panoaya, Sep 1920 (fl), Purpus 8611_p,p, (ARIZ, DS); with— out precise locality, 1920 (f1), Purpus 6611_p,p, (UC); Cerro de las Animas, near Playa Azfil on Lago de Catemaco, 8 Aug 1971 (sterile), Stevens 1399 (MSC). CHIAPAS: Mpio. of Tenejapa, paraje of Mahben Chauk, slopes along Ala Shashib River below Habenal, 3300 ft, 15 July 1964 (fl), Breedlove 6416 (DS); Mpio. of Tenejapa, paraje of Mahben Chauk, slopes near Habenal, 3500 ft, 26 Nov 1964 (fr), Breedlove 1622 (D8, F); Mpio. of Ixtapa, 26 mi S of Bochil along rd to Tuxtla Gutie— rrez, 3800 ft, 7 Aug 1967 (fl), Clarke 107 (DS); "les montagnes pres de 90 . M. hemsleyana - M. tuberosa .1.-- -1 400 KM Figure 12. Distribution of Matelea hemsleyana and M. tuberosa 91 village indien de Cancunc" [?Cancuc, Mpio. Chilon], June, year not giv- en (fl), Ghiesbreght 661, lectotype of Dictyanthus parviflorus (GH, MO, NY); rocky mountain slopes, Monserrate, June 1925 (f1), Purpus 166 (US); Hacienda Monserrate, Sep 1923 (f1), Purpus 6616 (F, GH, MEXU, MO, NY, US); near Hacienda Monserrate, Sep 1923 (f1), Purpus 6611 (UC); without precise locality, Sep 1923 (fl), Purpus 6,6, (UC); mountains E of Monserrate, June, year not given (fl), Purpus 16661 (UC, US); Mpio. of Tenejapa, paraje of Oshewits, slopes, 3500 ft, 10 Apr 1968 (fl), 166 6666 (DS). STATE UNKNOWN: without date (fl), §E§§§s Mocifio, E£.§l° 661 (F, fragment of MA specimen, MA, not seen, photo from F neg. 41468 of MA specimen at MSC). GUATEMALA. BAJA VERAPAZ: 12 km SW of Grana— dos, slopes above Rio Montagua, 2 Sep 1970 (f1), Harmon 6_66yg£_616§ (MO). JALAPA: Montana Durazno, 2 mi E of San Pedro Pifiula, open slopes of cuesta, 1400—1900 m, 10 Dec 1939 (fr), Steyermark 61611 (F). HUEHUETENANGO: slopes between San Ildefonso Ixtahuacén and Cuilco, 1350-1600 m, 16 Aug 1942 (fl), Steyermark 66161_(F, M0); between Nenton and Las Palmas, via Yalisjao, Rinc6n Chiquite, Chiaquial, Guaxacana, in Sierra de los Cuchumatanes, 800—1200 m, 30 Aug 1942 (fl & fr), Steyer— EEEE.§1§§£ (F, US). E1_SALVADOR. SANTA ANA: near Chalchuapa, 1922 (fl), Calder6n 1017 (US). 2. Matelea tuberosa (Robinson) Woodson, Ann. Missouri Bot. Card. 28: 237. 1941. Dictyanthus tuberosus Robinson, Daedalus 27: 180—181. 1891/1892 [1893]. Lectotype: Pringle 6666_(GH! lectotype; F! VT! iso— lectotypes). Syntype: Palmer 161 (CH! syntype; ENCB! G! K! MO! ND! NY, 3 specimens! P! PH! US! isosyntypes). 92 Plants erect to trailing or sometimes weakly twining. Stems 10—70 (-100) cm long, with a woody caudex to 5 cm long and 3 cm wide, this with thick corky bark, otherwise typically herbaceous and lacking bark (rarely subshrubs with erect, branched woody stems), with dense short trichomes, very sparse glandular trichomes, and sparse to dense, mostly straight long trichomes to 2 mm long. Leaf blade ovate to very—wide- ovate, 17—45 mm long, 17—40 mm wide, with mostly uncinate long trich— omes, surface smooth, smaller veins sharply raised below, apex acumi— nate to attenuate, base lobate, lobes mostly convergent to descending, with 3—6 (-9) acropetiolar glands, margin often somewhat thickened and revolute; petiole 7—31 mm long, with dense short trichomes, very sparse glandular trichomes, and sparse to dense, mostly uncinate long trich- omes. Inflorescence a simple cyme; peduncle 0.5—9.0 mm long, with dense short trichomes, very sparse glandular trichomes, and sparse to dense, straight or uncinate long trichomes; bracts linear or lorate to lanceolate, 2—8 mm long, with mostly uncinate long trichomes; pedicel 4—5 mm long, with indumentum of peduncle. Calyx lobes lanceolate to narrow—ovate or elliptic, 5—9 mm long, 1.5—3.5 mm wide, apex acute to attenuate, with one gland below each sinus, abaxial surface with sparse to dense, straight or uncinate long trichomes, adaxial surface gla— brous. Corolla deeply campanulate, base to sinus length 6—10 mm, limb revolute; lobes 2.5—6.0 mm long, apex acute, slightly to strongly re- flexed, margin revolute; glabrous within except with dense short trich— omes on limb and lobes, indumentum outside of short trichomes on tube and straight or uncinate long trichomes on limb and lobes, occasionally a few long trichomes scattered along tube and occasionally distal third of lobes glabrous; with a pair of ridges within tube opposite each 93 corona lobe, ridges of adjacent pairs almost coming together at base and forming pockets at base of corolla, with corona lobes in distinct pockets in bases of furrows between the paired ridges; within tube with fine grayish—brown vertical lines, limb densely grayish-brown—reticu— lated. Corona lobes ca 2 mm long (but borne distinctly above base of corolla), shape elaborate but basically sagittate in outline, main axis adnate to corolla and adnate to gynostegium by a thin wall. Gynostegi— um ca 2 mm high and ca 2 mm wide at apex, stipitate, apex broadly and shallowly concave, with corpuscula as high points, and slightly convex and bilobed in center, terminal anther appendages covering margin of apex. Corpusculum 0.14—0.22 mm long, 0.08—0.13 mm wide, pollen sacs 0.63-0.86 mm long, 0.29—0.37 mm wide. Follicles fusiform, 55—65 mm long, ll—l9 mm wide, mottled light and dark green, with scattered short and long trichomes, with 50—110 arcuate projections to 2 mm long. Seeds obovate, nearly circular, 5.5—6.0 mm long, 4.5-5.0 mm wide, with a raised, radially grooved margin, this entire to shallowly toothed distally, inside this margin slightly convex and verrucate on both sides, one side with a narrow ridge from apex to near center, light brown; coma ca 25 mm long. Figure 13. DISTRIBUTION AND ECOLOGY. Collected from southern Sonora to south- ern Jalisco at elevations of 500—1600 m. Figure 12. Growing in open oak and pine-oak forests and adjacent grasslands, usually in shallow, red clay soil. Flowering specimens have been collected from late July to early October and the one specimen with mature seeds was collected in March. Three collections, Pennell 16666, 66§6_1666, and Stevens 1666, have Specimens with from a few to most of the flower buds developing into 94 . 4’", \:§‘ :‘v .\“.‘:.'\*:.\:,, A . "(flfifzzf " Rev? 523’ 5;: ".£\“,‘:\"'l§q‘ 'fil/ w‘.‘ '1 ’ "K .1 "1’! EV”, !’ 1: I I ll”. \ omE.s. . wage—venue .2 o as V I 9 \M '0 0| Distribution of Matelea macvaughiana and M, standleyana. Figure 20. 126 SPECIMENS EXAMINED. MEXICO. JALISCO: Huejotitén, July 1912 (fl), Diguet gig. (MICH, 2 specimens, 1 mixed with Matelea pavbnii, P, US); ranch near Coyula (near Tonalé) ca 12 mi E of Guadalajara, July 1963 (fl), Faberge §,E. (TEX); wet seepage area 23 mi S of Guadalajara on Hwy 15, 5300 ft, 13 July 1963 (f1), Molseed §_31£g_119_(ARIZ, MEXU, MICH, M0, NY, UC); wet meadows near Guadalajara [E1 Castillo, according to Davis, 1936, p. 118], 22 Aug 1893 (f1), Pringle 2411_(GH, VT); moist slopes near Guadalajara [between E1 Castillo and Juanacatlan, op. cit., p. 199], 5 Aug 1902 (fl), Pringle 8919, type of Matelea macvaughiana (ENCB, F, G, 4 specimens, GH, L, 2 specimens, MEXU, M0, MSC, NY, P, PH, 2 specimens, POM, UC, US, 2 specimens, VT). MICHOACAN: cultivated fields 6-7 km N of Jaripo, rdside thickets, 1600 m, 1 Dec 1970 (fr), McVaugh 24934 (MICH, MSC). 6. Matelea standleyana Woodson, Ann. Missouri Bot. Card. 28: 237. 1941, based on Dictyanthus tigrinus Conzatti & Standley. Dictyanthus tigrinus Conzatti & Standley 1M_Standley, Contr. U.S. Natl. Herb. 23: 1183—1184. 1924, 22E Matelea tigrina (Grisebach) Woodson. Type: Conzatti 11§9_(US! holotype; GH! isotype). Plants twining vines. Stems herbaceous and lacking bark or some— times rhizomes slightly woody and with thin corky bark, rhizomes thin, horizontal, stem indumentum of sparse to dense short and glandular trichomes and sparse to very sparse straight long trichomes to 1.5 mm long, these very brittle and mostly missing from specimens. Leaf blade wide—ovate to very—wide—ovate or occasionally ovate, 48—104 mm long, 127 36-102 mm wide, indumentum of sparse uncinate long trichomes above and dense uncinate short trichomes below, surface smooth, apex acuminate to attenuate, base lobate, lobes mostly convergent to descending, with 1—7 acropetiolar glands; petiole 35—112 mm long, with sparse to dense short and glandular trichomes and sparse to very sparse, straight or uncinate long trichomes. Inflorescence a simple cyme; peduncle 5—18 (—25) mm long, with indumentum of stem or often with long trichomes absent; bracts linear to lanceolate, 2—6 mm long, with dense short and sparse straight long trichomes; pedicel 7—16 mm long, with indumentum of stem. Calyx lobes narrow—ovate or occasionally lanceolate or ovate, (8.5—) 12—18 mm long, 4.0—6.5 mm wide, apex attenuate, with one gland below each sinus, abaxial surface with dense short trichomes, margin also with sparse straight or uncinate long trichomes, adaxial surface gla— brous. Corolla deeply campanulate, base to sinus length (14—) 17-31 mm, limb revolute; lobes 17—28 mm long, apex acute, plane to reflexed, margin revolute; glabrous within except with moderately dense to dense short trichomes on lobes, limb, and around corona lobes, indumentum on outside of sparse to dense short trichomes; tube convoluted with raised parts opposite corona lobes and sacs formed between them; with thick, brownish—red, circular lines within tube, these becoming thinner and reticulated on distal part of limb and on lobes. Corona lobes 9—13 mm long, linear in outline, connate at base, adnate to corolla and adnate by a thin wall to gynostegium. Gynostegium (3—) 4~5 mm high and 3.0— 4.5 mm wide at apex, stipitate, apex with a blunt projection (formed from apices of anther wings) below each corpusculum and exceeding them laterally, apex convex with tip flattened and slightly bilobed and slightly exceeding corpuscula, terminal anther appendages covering ca -l I I ‘! 128 one third of apex. Corpusculum 0.48-0.55 mm long, 0.23-0.28 mm wide, pollen sacs 1.54—1.88 mm long, 0.45—0.63 mm wide. Mature follicles un— known, immature follicles fusiform, to 85 mm long, to 28 mm wide, ap— parently green, with dense short trichomes, with ca 50 thick, straight projections to 7 mm long. Seeds unknown. Figure 21. DISTRIBUTION AND ECOLOGY. Apparently restricted to northern Oaxaca and adjacent Veracruz at elevations up to ca 900 m. Figure 20. Appar— ently a plant of moist thickets. Flowering specimens collected mainly in August, but also once each in June and late October. The one imma— ture fruit was collected in August. Santos 1118 and §M1EM.§9§_show ap— parent chewing insect damage to the leaves. No pollination activity was observed but the flowers of a greenhouse—grown specimen were noted to produce a faint foetid odor in late afternoon, suggesting possible dipteran pollination. COMMON NAMES AND LOCAL USES. None known. DISCUSSION. Despite the number of collections, this is a poorly knowu species. Most of the collections are inadequately labelled and several are difficult to locate on maps with certainty. My collection, from near the type locality, is the only one obtained since 1943. This species is readily identifiable because of the large, deeply campanu— late corolla with circular markings inside the tube. This is likely the largest—flowered species of New World Asclepiadoideae. Well—formed leaves of this species are, along with those of Matelea pavonii, the largest of the subgenus and have uniquely angled sinuses. For a discussion of the Sessé and Mocifio collections of this spe— cies, see the discussion under Matelea pavonii. i r . | 129 Figure 21. Matelea standleyana (drawn from Stevens C—161, a cultivated specimen of Stevens 1392). A. section of flowering stem, x 0.6; B—C. flowers, x 1.2; D. pollinium, x 12; E. fertile shoot apex, x 6; F. section of old stem, x 0.6. 130 SPECIMENS EXAMINED. MEXICO. VERACRUZ: C6rdoba, 16 Aug 1907 (fl), Arséne 4199 (L); Mirador, June 1856 (fl), Botteri §,M. (US); "In remetis Miradoris Sarton," 1857 (f1), MEEM 9 (US); C6rdoba, 16 Aug 1882 (fl), Kerber 99_(G, 2 specimens, MICH, P); Ejido de Manzanares, 1-4 km NW of Campo Experimental de Hule, El Palmar, Zongolica, 1—5 Aug 1943 (fl), Santos 1919_(MICH). OAXACA: Dist. Tuxtepec, de La Laguna de Ojitlan, 350 m, 31 Oct 1919 (f1), Conzatti 9199, type of Dictyanthus tigrinus (GH, US); Petlapa, 3000 ft, 184? (f1), Galeotti 1999 (P, 3 specimens); Dist. Tuxtepec, en la montafia de Jacatepec, 22 m, 16 Aug 1940 (E1), Martinez—Calder6n 111 (A, UC, US); Ixcatlén, 850 ft, 19 Aug 1895 (fl & fr), 9911M 999 (CH); ca 29.7 mi W of Tuxtepec along rd to lxcatlan, near where it branches from rd to Jalapa de Diaz, 7 Aug 1971 (fl), Stevens 1991_(MSC). STATE UNKNOWN: Trapiche, Aug 1842 (fl), Liebmann 11999 (F, 3 specimens, US); without locality and date (f1), §g§§§, Mocifio, g£_§1, 1199_(MA, not seen, photo from F neg. 41473 at MSG), 9999_p,p, (F, MA, not seen, photo from F neg. 41474 of MA speci— men at MSC; MA specimen partly or completely Matelea pavonii), 9991 (F, MA, not seen, photo from F neg. 41475 at MSG; both specimens mixed with Matelea pavonii); without locality and date (fl), Herb. M.—E. Moricand (G). 7. Matelea ceratopetala (J. D. Smith) Woodson, Ann. Missouri Bot. Card. 28: 236. 1941. Dictyanthus ceratopetalus J. D. Smith, Bot. Caz. (Crafordsville) 18: 208. 1893. Type: nggg_§_19§.§§_1, 9, §El£h.§222.(US! holotype; C! CH! K! NY! US! isotypes). 131 Plants twining or occasionally trailing or erect. Stems 25—55 cm long when erect, woody and with thin to thick corky bark below or occa— sionally entirely herbaceous, otherwise herbaceous and lacking bark, with dense short trichomes, sparse to dense glandular trichomes, and very sparse to moderately dense, mostly straight long trichomes to 2 mm long. Leaf blade ovate to very—wide—ovate, 25—63 mm long, 15-55 mm wide, indumentum above of sparse or occasionally dense, straight or un- cinate long trichomes and occasionally also sparse short trichomes, or rarely glabrous, indumentum below of dense short and sparse straight or uncinate long trichomes, surface smooth, smaller veins often moderately to sharply raised below, apex acute to attenuate, base lobate, lobes mostly convergent, with 1—4 (-7) acropetiolar glands, margin often somewhat thickened and revolute; petiole 13-60 (—70) mm long, with dense short trichomes, sparse to dense glandular trichomes, and very sparse to moderately dense, mostly uncinate long trichomes. In- florescence a simple cyme or sometimes reduced to a single flower but always with a distinct peduncle; peduncle 1—15 (-27) mm long, with dense short trichomes, sparse to dense glandular trichomes, and very r sparse to moderately dense, straight or uncinate long trichomes; bracts 1 linear to lanceolate, 3—5 (—7) mm long, abaxial surface with short, glandular, and long trichomes, adaxial surface with short and occasion- ally also scattered long trichomes; pedicel 3—12 (—15) mm long, some— times accrescent in fruit, with indumentum of peduncle. Calyx lobes narrow—ovate or occasionally lanceolate, 9—13 mm long, 3.5—6.0 mm wide, apex acute or attenuate, with one gland below each sinus, abaxial sur— face with scattered straight or uncinate long trichomes, adaxial sur— face glabrous. Corolla campanulate, base to sinus length 12—20 mm, 132 limb revolute; lobes 11—18 mm long, apex acute, plane to reflexed, mar— gin revolute; glabrous within except with scattered short trichomes 0n lobes, limb, and around corona lobes, indumentum on outside of sparse to dense short and scattered long trichomes; tube convoluted with raised parts opposite corona lobes and deep sacs formed between them; brownish— or reddish—purple-reticulated, reticulations wider within tube. Corona lobes 8—11 mm long, linear—spathulate in outline, connate at base, adnate to corolla and adnate by a thin wall to gynostegium, this wall with a distinct tooth near center of upper margin. Gynoste— gium 3.5—5.5 mm high and 3.0—4.5 mm wide at apex, stipitate, apex apic— ulate, apiculum 1-2 mm high and exceeding corpuscula, terminal anther appendages covering margin of apex. Corpusculum 0.41—0.46 mm long, 0.18—0.22 mm wide, pollen sacs 1.52—1.63 mm long, 0.40—0.49 mm wide. Follicles fusiform, 75—90 (—103) mm long, 20—27 mm wide, striped green and light green or white, with short and glandular trichomes, with (36—) 48-60 straight or occasionally arcuate projections to 7 mm long. Seeds obovate, ca 4.5 mm long and 2.5-3.5 mm wide, with a raised, radi- ally grooved margin, this weakly toothed distally, inside this margin convex and verrucate on one side, concave and verrucate to rugose on opposite side, concave side with a slight ridge extending ca 1 mm from apex, dark brown; coma 20—30 mm long. Figure 22. ECOLOGY AND DISTRIBUTION. Collected from southern Guatemala to northern Nicaragua, mostly at elevations of 800-1000 m but occasionally up to 1360 m and dOWn to near sea level. Figure 23. Apparently not found in forests but rather in moist or dry fields, thickets, fence- rows, streamsides, and roadsides. Mostly associated with rocky volcan— ic soils but once noted as occurring on a salt flat. Flowering mostly Figure 22. Matelea ceratopetala (drawn from Stevens 1245). A. section of flowering stem, x 0.6; B—C. flowers, x 1.2; D. pollinium, x 12; E. section of old stem, x 0.6. 134 mid-July to early October but as early as June and as late as November. Collected with mature—sized fruits from mid-September to December. The following collections showed significant chewing insect damage to the leaves of at least some of the specimens: MOlina M, 4999, 22511, Standley 12709, 26902, and Stevens 1245. In one case a lepidop— teran larva was pressed on the leaf and this is likely the kind of in— sect causing all the damage noted here. COMMON NAMES AND LOCAL USES. The following names have been applied to the fruits of this species, which are cooked and eaten when young: cochita (Standlgy 76325), cochitos (Standley 75729), corazon_ggi1g£g (Moreles 3, 1293), cuchampgl (Standlgyw g 91. 547), chanchitos (Molina .3. 1084), chununa (Standley 76678 and 9919919149). The names sombreros and sombreritos (Stand1gy 76662) apply to the flowers. "Leoncia picuda" is written on the label of the Chavez collection but I do not know the intent of this entry. DISCUSSION. Until now, the plants I consider to be Matelea dicty— antha have been included in this taxon. The two species are actually quite distinct and, in fact, Matelea ceratopetala has its closest af— finities with M, pavonii. For a further discussion of this problem, see the discussion under Matelea dictyantha. Matuda 1778 is only tentatively included under this species and is neither mapped nor included in the description. It differs in several respects but probably would still key to this species. The most sig- nificant differences are 1) it is outside the otherwise known range of the species, but would not be an unrealistic extension of the range, 2) the calyx lObes are larger and proportionately wider than those on any other specimen of this species, 3) the corolla is as large as the 135 largest measurements for this species, 4) the corolla reticulations are thinner and more widely and uniformly spaced than in any other speci— men, 5) the thin wall of the corona lobe apparently lacks a tooth, and 6) the apiculum of the gynostegium apex is less than 1 mm long and is apparently exceeded by the corpuscula. In some ways these characters approach Matelea pavonii but 1) it is also out of the known range of that species and represents a less likely range extension, 2) the leaves are not pusticulate and in general size and shape much more nearly resemble those of M, ceratopetala, 3) the peduncle is too short, and 4) the pattern of lines within the corolla tube is reticulate rath— er than of parallel vertical lines. The single specimen of Matuda 1119 has but one open flower and I consider it inadequate for a more defi— nite taxonomic disposition at this time. I attempted to recollect this plant in 1971, but the lack of wet season roads prevented an effective search. Additional searches for Dictyanthus in southern Chiapas and southwestern Guatemala are much to be desired and may help clarify the status of this specimen. SPECIMENS EXAMINED. MEXICO. CHIAPAS: Fca, Fuarez, Esc. [?Escuint— la], 12 Aug 1937 (fl), Matuda 1119_(MICH). GUATEMALA. ESCUINTLA: SE of Escuintla, 7 Sep 1964 (fl), 91199 99 91. 999_(MO): Puerto de S. José, 5 m, Nov 1929 (fl), Morales 9, 1199 (F). JALAPA: vicinity of Jalapa, ca 1360 m, 7—18 Nov 1940 (fl), Standley 19991 (F), 7—18 Nov 1940 (fr), 19919 (F); Finca El Ingenio, 3900 ft, 27—28 Sep 1944 (fl), flh199‘9149 (MICH). JUTIAPA: vicinity of Jutiapa, ca 850 m, 24 Oct—5 Nov 1940 (fl), Standley 19199 (CAS, F, MO), 24 Oct—5 Nov 1940 (fr), 19119 (F), 19919 (F, MO); potreros between Trapiche Vargas and Asunci6n Mita, 500—600 m, 15 Nov 1939 (fr), Steyermark 31786 (F). SANTA ROSA: 136 Km 38, 16 km NW of Barberena, 22 July 1970 (f1), Harmon 9 Dwyer 3230 (UMO); Santa Rosa, 3000 ft, Aug 1892 (f1), Hgyde_9 Lux 91 9, 9, Smith 3999, type of Dictygnthus ceratgpetalus (C, CH, K, NY, US, 2 speci- mens); 12 km NNE of Barberena, ca 3 km from Santa Rosa de Lima, 15 July 1971 (fl), Stevens 1245 (MSC). DEPARTMENT UNKNOWN: near Rosario, 16 Aug 1860 (fl),Ing§§_§,9, (CH). 91_SALVADOR. LA LIBERTAD: Km 14 de la carretera hacia La Libertad, 8 July 1957 (fl), 19g9§.919 (USF). SAN SALVADOR: San Salvador, June 1922 (fl), Ca1der6n 824 (US). HONDURAS. EL PARAISO: drainage of Rio Yeguare (ca 87°W, 14°N), Llano de Lizapa, 900 m, 24 Aug 1948 (f1), Molina 9, 1994 (F, MO, US); drainage of Rio Yeguare (ca 87°W, 14°N), entre Mata Indio y Lizapa, 950 m, 25 July 1951 (fl), Molina 9, 4053 (F, CH); drainage of Rio Yeguare (ca 87°W, 14°N), Rio Lizapa, 1000 m, 19 Sep 1951 (fl), Molina 9, 4111 (F, CH); Las Casi— tas, 950 m, 4 Dec 1946 (fl & fr), Standley 91 91, 941_(F); along Rio Yeguare near Casitas, 900 m, 4 Dec 1946 (fr), Williams 9_Molina B, 11999_(MO). MORAZAN: drainage of Rio Yeguare (ca 87°W, 14°N), Jicarito River, 2600 ft, 15 July 1948 (fl), Glassman 1902 (F, ILL, MIN, NY); Rio Guacerique between Los Laureles and Las Tapias, NW of Teguci- galpa, 1000 m, 4 Nov 1966 (fl & fr), Molina 3, 18618 (F, C, CH, NY); 1 small streamside ca 2 km S of EAP campus, E1 Zamorano, Aug 1960 (f1), Pfeifer 1633 (US); along Rio Yeguare, E of El Zamorano, ca 750 m, Sep— Dec 1948 (f1 & fr), Standley 12709 (F, P); along rd from El Zamorano toward San Antonio de Oriente, 825—950 m, Sep—Nov 1948 (fl), Standley 13671 (F); vicinity of El Zamorano, near Rastro, 800—850 m, 5 Aug 1949 (fl), Standley 22152 (F, MO, US); along road from E1 Zamorano toward San Antonio de Oriente, 825-950 m, 9 Aug 1949 (fl), Standley 22377 (F); W of El Zamorano, along trail from Rio de la Orilla to El Pedregal, 137 1! 00¢ ocucobofif . Soaoaouocooi . Distribution of Matelea ceratopetala and M. dict antha. Figure 23. j—f—‘ 138 800-900 m, 6 Aug 1949 (fl), Standley 22683 (F); vicinity of El Zamo— rano, rd toward Chagfiite, ca 800 m, 30 Sep 1950 (f1), Standley 26902 (F, GH, US); Zamorano, 800 m, Aug, year not given (fl), Valeria 9,‘99 (F); Santa Inés, 850 m, Aug 1943 (f1), Valerie 9, 413 (F); El Pedregal, 800 m, 18 Sep 1943 (fr), Valeria M, 874 (F); Rio Yeguare, 800 m, 32 [!] Sep 1943 (f1), Valerio 9, 985 (F); Zamorano, 800 m, 4 Oct 1943 (fl), Valerio 9. 1082 (F), 25 July 1945 (f1), Valerio 9, 3140 (F, MO); along Santa Clara Creek, Zamorano Valley, 850 m, 9 Oct 1946 (f1), Williams_9 Molina 9, 10569 (F, MO). OCOTEPEQUE: along rd, vicinity of Antiguo Ocotepeque, 800 m, 31 Aug 1968 (fl), Molina 9, 22511 (F, NY). NICARAGUA. CHINANDEGA: vicinity of Chichigalpa, ca 90 m, 12-18 July 1947 (fl), Standley 11388 (F). DEPARTMENT UNKNOWN: "Leoncia picuda," 8 Oct 1927 (fl), Chaves [Chévez] 315 (US). 8. Matelea dictyantha Woodson, Ann. Missouri Bot. Card. 28: 236. 1941, based on Rytidoloma reticulatum Turczaninow. Bytidoloma reticulatum Turczaninow, Bull. Soc. Imp. Naturalistes Moscou 25(2): 319-320. 1852, non Matelea reticulata (Engel- mann g§_A. Gray) Woodson. Type: Jurggnsen 692 (K! probable isotype). Dictyanthus reticulatus (Turczaninow) Bentham & Hooker f. 91 Hemsley, Biol. Centr. Am. Bot. 2: 329. 1882. Plants erect to trailing or twining. Stems 15—70 (~150+) cm long, with a woody caudex to 5 cm long and 3 cm.wide, this with thin to thick corky bark, or occasionally with an elongate woody rhizome, also often with woody stems above caudex or rhizome, these usually with thin corky 139 bark and up to ca 5 cm long but occasionally up to 20 cm long, other— wise herbaceous and lacking bark, with dense short and glandular trich- omes and sparse to dense, mostly straight long trichomes to 3 mm long. Leaf blade wide—ovate to very-wide—ovate, 26—62 (—lO3) mm long, 24—52 (—76) mm wide, with uncinate long trichomes and also often glandular trichomes on veins below, surface smooth, smaller veins sharply raised below, apex acute to attenuate or rarely obtuse, base lobate, lobes mostly convergent to descending, with 1—6 (—8) acropetiolar glands, margin often somewhat thickened and revolute; petiole (11—) 14—34 (-57) mm long, with dense short and glandular trichomes and sparse to dense uncinate long trichomes. Inflorescence a simple cyme or sometimes re— duced to a single flower with or apparently without a rudimentary pe— duncle; peduncle absent—10 mm long, with indumentum of petiole; bracts linear to lanceolate, 2.5—7.0 mm long, abaxial surface with indumentum of stem, adaxial surface glabrous; pedicel 5—12 (—16) mm long, some- ! times markedly accrescent in fruit, with indumentum of stem. Calyx A lobes narrow—ovate or occasionally lanceolate or ovate, 6—11 mm long, 2.5—6.0 mm wide, apex acute to attenuate, with one gland below each h sinus or occasionally these somewhat above sinus near margin of lobe, abaxial surface with indumentum of stem, adaxial surface glabrous. Corolla campanulate, base to sinus length (7—) 9—16 mm, limb revolute; lobes 8—14 mm long, apex acute or sometimes rounded, plane to strongly reflexed, margin revolute; glabrous within except with sparse to dense short trichomes around corona lobes and on limb and bases of lobes or sometimes over whole surface of lobes, indumentum on outside of short trichomes and occasionally also with long trichomes on limb and bases of lobes or occasionally tube and tips of lobes nearly glabrous; tube 140 convoluted with raised parts opposite corona lobes and deep sacs formed between them; faintly to densely grayish-purple—reticulated, reticula— tions wider within tube. Corona lobes (5—) 6-8 mm long, linear in out- line with a raised margin, adnate to corolla and adnate by a thin wall to gynostegium, connate at base, wall to gynostegium continuing as a narrow ridge nearly the length of lobe. Gynostegium 3-4 mm high and 3.0-3.5 mm wide at apex, stipitate, apex broadly and shallowly concave with corpuscula as high points, terminal anther appendages covering margin of apex. Corpusculum 0.22—0.35 mm long, 0.09-0.17 mm wide, pol- len sacs 1.17-1.45 mm long, 0.29-0.38 mm wide. Follicles fusiform, (45-) 55-70 mm long, 10-22 mm wide, light green with a few dark green stripes, with short and glandular trichomes, with (30-) 50-110 straight to arcuate projections to 3 mm long. Seeds obovate, ca 5.5 mm long and 4.0—4.5 mm wide, with a raised, radially grooved margin, this weakly toothed distally, inside this margin flat or slightly concave and ver- rucate on one side, convex and verrucate on opposite side, flat side with slight ridge from apex to near center, light brown to brown; coma 25—30 mm long. Figures 24 and 25. DISTRIBUTION AND ECOLOGY. Collected in the mountains of four more or less discrete areas: around Cuernavaca (Morelos and adjacent state of México), around Chilpancingo (Guerrero), around Oaxaca (Oaxaca), and in southwestern Puebla. Figure 23. Found at elevations of about 1500— 2500 m. About a third of the collections are noted as being on or as— sociated with limestone and many of the other localities are in lime— stone areas, but it cannot be determined at this time if the species is restricted to this substrate. Mostly found in low, open oak, pine, or pine-oak forests, especially where disturbed. Flowering mid-June to Figure 24. 141 Matelea dictyantha (A—D drawn from Stevens C—lOS, cultivat- ed specimens of Graham 1231, and E from Stevens 1311). A. section of flowering stem, x 0.6; B—C. flowers, x 1.7; D. pollinium, x 18; E. caudex, x 0.6. Figure 25. 142 Regional variation of Matelea dictyantha. A—B represent plants from Morelos and Guerrero and C—D rep- resent plants from Oaxaca, all scales approximately equal- A. Stevens C-105 (greenhouse—grown plant of Graham 1231); B. Rose 11065 (US); C. copy negative taken from color posi- tive, courtesy of Dr. Robert Cruden; D. Iltis 25 El. llgé (WIS). 143 144 mid—September. Mature-sized fruits collected August-December, but with mature seeds only in December. Four of my collections from Oaxaca (Stevens 1191, 1211, 1144, 114g) have a few insect-chewed leaves and Graham 1211) 11£1§_g£.a1. 1194, and Rg§g_119§§ have some of the flower buds forming galls. The galls were removed from the Graham specimen at the time of collection (by me) and the insects were collected as they emerged. They are an undetermined dipteran species. One adult cerambycid beetle, probably Tetraopes, was collected on Stevens 1111. The larvae of many species of Tetraopes live in the stems and rhizomes of Asclepias (Chemsak, 1963). Matelea dictyantha as well as the other species of Matelea with thick rhizomes or caudices might be suitable hosts for Tetraopes but I have no direct evidence of this. Most plants of Stevens 1111 were infected with a rust. Associ- ated with the rust, but probably only feeding on the fungal spores or secretions, were numerous small dipteran larvae. COMMON NAMES AND LOCAL USES. None known. DISCUSSION. As mentioned in the discussion under Matelea macvaugh— 1gna, Woodson provided the epithet for g, dictyantha quite by accident. He did not distinguish between this species and Matelea ceratopetala, a species to which it bears some resemblance in the shape and color pat— tern of the corolla and the size and shape of the leaves but from which it is clearly distinct. The major characteristics distinguishing this species from Matelea ceratopetala include the woody caudex and predomi— nantly shorter habit, the smaller flowers, the lack of a tooth on the wall connecting the corona lobe to the gynostegium, the concave rather than apiculate gynostegium apex, and the smaller follicles with smaller 145 and more numerous projections. The geographic ranges of the two spe— cies are also distinct. The distinctiveness of this species from Matelea ceratopetala has been noted on two specimens, Conzatti g1§§ at F (unsigned and undated) and Pringle 312% at GH (J. M. Greenman, 18 Sep 1890). Standley (1924) considered Rytidoloma reticulatum to be synony— mous with Matelea pavonii. The holotype for the Turczaninow name has not yet been located. It may be at CWU (Charkow University, U.S.S.R.), to which was given the Turczaninow herbarium after his death, but I have not attempted to bor— row their material. The one specimen of Jurgensen §21_which has been examined is at K and is probably an isotype. The form of Turczaninow's protologue (1852, pp. 319—320) has also led to some confusion. The ap— parent generic description is actually a description of the genus and its single species and cites Jurgensen Q21. Immediately following the genus—species description is the entry, "18. R. reticulatum. Altera species hujus generis, quantum e flore unico, Ptino corrupto, dijudicare possum, adest in collectione Galeottiana ex Oaxaca sub n. 1563." This led Langman (1964, p. 748) to state that Rytidoloma EEEEET ulatum was based on a Galeotti collection. Turczaninow was actually indicating that he recognized another species in his new genus, but he neither named nor described it. Galeotti 12g; is indeed a different species, Matelea standleyana. As with Matelea hemsleyana, two distinctive elements of this species can be recognized, but I do not believe they warrant taxonomic recogni— tion of the basis of currently available material. The flowers of the Morelos—Guerrero element of this species tend to be larger, more shal— lowly campanulate, and much lighter in color than the Puebla—Oaxaca 146 element (Figure 25). In this case the ranges are apparently distinct but I have found no non-subjective way of describing the difference in dried specimens. Perhaps with additional spirit—preserved collections and specimens with descriptions of the flower colors, taxonomic recog— nition will become more feasible. In this connection, the Specimens of Ghiesbreght s13. from province d‘Oaxaca are of the Morelos—Guerrero type and were most likely collected at Cuernavaca, Morelos (this mixing of labels apparently occurred often with Ghiesbreght specimens). It is also interesting to note that the distribution of this species largely resembles that of Crusea calcicola (Anderson, 1972), another species growing on calcareous soils. SPECIMENS EXAMINED. MEXICO. MEXICO: 4 mi S of Ixtapan on Hwy 55, 16 Aug 1972 (fl), Dunn 35 31. 19421 (MSC, UMO); cercania de Ixtapan de la Sal, 1800 m, 16 Aug 1953 (f1), Matuda g£_a1, 1§§Q§_(MO). MORELOS: "pres de Cuernavaca," Sep [7], year not given (f1), Ghiesbreght e§_JDP §Z_(P, mixed with Matelea pavonii); near Yautepec, 12-13 July 1905 (fl), BEES gggg (US); near Cuernavaca, 14 Aug 1906 (fl & fr), Rggg 11965 (GH, NY, US). GUERRERO: Mpio. de Chilpancingo, camino a1 Cerro Alquitrén, cerca de Mazatlan, 1500 m, 5 July 1966 (fl), Chavez 1Q (ENCB); hill 8 of Chilpancingo, 5.7 mi S of S entrance to town on Hwy 95, 20 Aug 1972 (f1), QEEE;EEufll°.g9§9§.(MSC: UMO); small rocky moun— tain valley 14 mi from Mex. 95 on rd to Chichihualco, 26 Oct 1970 (fr), Graham 1231 (MICH); Mpio. de Chilpancingo, veriente E del Cerro Alqui— trén, cerca de Mazatlan, 1500 m, 5 July 1966 (fl), Rzedowski 22677 (ENCB, MICH, MSC, WIS). PUEBLA: Cerro de Castillo, near Coatepe [Coatepec], July 1907 (f1), Purpus 2620 (F, GH, MO, NY, UC, US); cerros near San Luis, July 1907 (fl), Purpus 2620A (UC). OAXACA: Valley of 147 Etla, July 1895 (fl), Alvarez g§_1, g, §Ei£h.£lg (GH); lower slopes of Cerro San Felipe, S side, above town of San Felipe del Agua, 1700—2000 m, 20 July 1968 (f1), Anderson §_Anderson 4854 (MICH); Dist. Centro, Hacienda de Guadalupe, 1600 m, 14 June 1908 (fl), Conzatti 1148 (F); Cafiada de San Gabriel Etla, 1930 m, 8 Aug 1897 (fl), Conzatti 9.9227 E§1g§_113 (CE); Sierra, 7000-8000 ft, 184? (fl), Galeotti 1§§Z_p,p. (US); Yavezia [Santa Maria Yavesia], 184? (fl), Galeotti 1§§Z_p,p, (P); without precise locality and date (fl), probably actually from Cuerna— vaca, Morelos, Ghiesbreght s32, (K, P, mixed with Matelea pavonii); base of Cerro San Felipe, ca 9—11 km NNE of Oaxaca, along rd from San Felipe, ca 2000 m, 21 Aug 1960 (fl & fr), 11£1§_§£.a1. 1128 (W18, 4 specimens), 21 Aug 1960 (f1), 1194_(MEXU, MICH, MO, UC, WIS, 2 speci— mens); "Sierra San Pedro Nolasco, Talea, etc.," 1843—1844 (fl), Jurgensen.§21, type of Rytidoloma reticulatum (K); Valley of Oaxaca, 5100—5800 ft, 8 Sep 1894 (fl), Nelson 114Z_(US); Valley of Oaxaca, 5000—5300 ft, 20 Sep 1894 (f1), Nelson 1124 (US); hills above Oaxaca, 6000 ft, 6 Aug 1894 (apparently f1 & fr), 26 Dec 1894 (apparently fr), Pringle 4168 (BKL, ENCB, F, G, 3 specimens, CH, ISC, MASS, MEXU, MICH, MIN, MO, MSC, ND, NY, P, PH, UC, US, VT); near city of Oaxaca, 16—21 June 1899 (fl), Ro§g_4§1§ (US); 9 km a1 S de Sola de Vega, sobre la carretera a Puerto Escondito, 1600 m, 30 Sep 1965 (fr), Rzedowski 11119 (ENCB); along rd to microwave tower ca 3.6 mi S of Matatlén on Hwy 190, ca 1 mi 8 of Km 595, 6 July 1971 (fl), Stevens 1191 (MSC), 22 July 1971 (fl), Stevens 1111_(MSC); ca 5.8 mi N of Telixlahuaca along Hwy 131, 25 July 1971 (fl), Stevens 1§4§_(MSC); ca 3.8 mi NE of Sola de Vega along Hwy 131, 26 July 1971 (f1), Stevens 1144 (MSC); ca 4.9 mi SW of Sola de Vega along Hwy 131, 26 July 1971 (fl), Stevens 1346 (MSC). STATE 148 UNKNOWN: "Eugenio," "Sierra de la Cruz," "Vera Cruz to Orizaba," 1 Aug 1853 (f1), Muller 1108 (K, NY). 9. Matelea aenea (Woodson) W. D. Stevens, comb. nov. Dictyanthus aeneus Woodson, Amer. J. Bot. 22: 691, pl. 1, fig. 4. 1935. Type: Steere 3005 (M0! holotype; MICH! isotype). Plants twining vines. Stems woody below, with thin to thick corky bark, herbaceous and lacking bark above, with dense short and glandular trichomes and dense uncinate long trichomes to 2.5 mm long. Leaf blade wide-ovate to very—wide—ovate or occasionally ovate, 35—98 mm long, 26— 70 mm wide, indumentum of uncinate long trichomes and also glandular trichomes on veins below, surface smooth, apex acute to acuminate, base lobate, lobes convergent to widely divergent, with 4—11 acropetiolar glands, margin often crispate; petiole 21—62 mm long, with indumentum of stem. Inflorescence a simple cyme; peduncle 4—11 mm long, with in- dumentum of stem; bracts linear to lanceolate, 3—4 mm long, with indu— mentum of stem or occasionally glabrous on adaxial surface; pedicel ca 4 mm long, with indumentum of stem. Calyx lobes lanceolate to narrow— ovate, 6—9 mm long, 2.0—4.5 mm wide, apex acute to attenuate, with one or occasionally two glands below each sinus, abaxial surface with dense short, glandular, and uncinate long trichomes, adaxial surface gla— brous. Corolla campanulate, base to sinus length (6-) 8—12 mm, limb ascending to nearly plane, not revolute; lobes 5—9 mm long, length to width (sinus—sinus) ratio 0.67-0.78, apex acute or rounded, ascending to slightly reflexed at tip, margin not revolute; glabrous within ex— cept with dense short trichomes on limb and lobes, indumentum on 149 outside of short, glandular, and uncinate long trichomes; tube convo— luted with raised parts opposite corona lobes and deep sacs formed be— tween them; pale yellowish—green, sometimes with faint reticulations, these often drying darker. Corona lobes 6-8 mm long, linear—spathulate in outline, with tip deeply rugose and glistening purplish—black, otherwise yellowish-green or tinted purple, connate at base, adnate to corolla and adnate by a thin wall to gynostegium, this wall continuing as a narrow ridge nearly the length of lobe. Gynostegium ca 3 mm high and ca 2.5 mm wide at apex, stipitate, apex broadly and shallowly con— cave with corpuscula as high points and occasionally also slightly con- vex in center, terminal anther appendages hardly covering margin of apex. Corpusculum 0.21—0.28 mm long, 0.12-0.15 mm.wide, pollen sacs 1.08-1.18 mm long, 0.35-0.37 mm wide. Young follicles with short pro— jections, mature specimens unknown on specimens but said to be lO—15 cm long and "tuberculosa" (Flores_1). Seeds unknown on specimens but said to be comose (Flores 1). Figure 26. DISTRIBUTION AND ECOLOGY. Collected only in the vicinity of Pro- greso, on the tip of the Yucatan Peninsula, at near sea level. Figure 27. Growing in low scrub vegetation in thin, limestone derived soils, apparently where not particularly saline. Collected in flower June- August and in December. The only insect damage found was a few chewed leaves on Stevens 1145, COMMON NAMES AND LOCAL USES. Gaumer 1173 232, gives the name xbockin. Flores 1 gives the common names cabeza 43 cocodrilo, appar— ently referring to the follicles, and mata chivo on the label as well as an attached note as follows: "Parece ser la Philiverta Lindeleana. Figure 26. 150 Matelea aenea (A—D drawn from Stevens C—157, cultivated Specimens of Stevens 1145, and E from Stevens 1145). A. section of flowering stem, x 0.6; B-C. flowers, x 2.3; D. pollinium, x 24; E. base of stem, x 0.6. 151 Apoccinia silvestre que abunda en el camino del ferrocarril de via angusto, algunos 1e dicen mata chivo porque aunque la come el ganado, sus hojas trituradas y molidas en la comida matén a los Perros, lo mismo que a los chivos que las comen. Su fruto de 10 a 15 ctms. es una capsule tuberculosa y lechosa cuando verde, sus granos llevan pelos I largos.‘ It is unfortunate that more collectors do not make such use— ful and interesting observations. DISCUSSION. See also the discussion under Matelea yucatanensis. In describing Dictyanthus aeneus, Woodson (1935) summarized the differ— ences as follows: "_1. aeneus differs from 1, yucatanensis Standl. [sic] superficially in the smaller and more shortly petiolate leaves with paler color and hispidulous or strigillose surfaces, and smaller, paler corollas with a more pronounced campanulate tube. Structural differences of the corolla and corona are conspicuous as well." Later, Woodson (1941) considered the species to be synonymous with Matelea ypcatanensis (Standley) Woodson and so annotated the type specimen. I concur with his original recognition of Dictyanthus aeneus, but unfor— tunately his characters are not particularly diagnostic and, in fact, his drawing of the flower (Woodson, 1935, p. 1, fig. 4a) has the shape of Matelea yucatanensis and the size of M, aenea. The best characters for separating the two species are most easily observable in fresh flowers. The corolla limb and lobes of Matelea yucatanensis form es— sentially a right angle with the tube and have revolute margins while those of M, aenea are ascending and do not have revolute margins. (Under greenhouse conditions, Matelea aenea occasionally produces flow— ers which have slightly revolute margins and lobes which are not strongly ascending.) This difference can often still be seen in dried 152 specimens when, ironically, they are not carefully pressed. The corol— la color is also strikingly different in fresh flowers, yellowish— green with faint reticulations in Matelea aenea and densely grayish— purple-reticulated in.M, yucatanensis. Unfortunately, Matelea aenea sometimes darkens in drying and the difference is partially obscured. The floral characters which are most dependable in pressed specimens are the size and shape of the corolla lobes, but even with these the flowers often require boiling to be accurately measured. As noted in the descriptions, Matelea aenea has shorter and proportionately wider corolla lobes. There seem to be certain vegetative differences as well, but the variation within each species is large and there are too few specimens to make possible any meaningful conclusions. Matelea 33323 tends to have denser long trichomes on the vegetative parts and smaller, wider, more crispate, and less purple-pigmented leaves. Finally, it can be seen in Figure 27 that the ranges of the two species seem to be allopatric; Matelea aenea appears to be restricted to the coastal area immediately around Progreso, while M, yucatanensis is found at scattered, more inland, localities. Both species are too poorly collected, however, to support much conjecture on their relative distributions. As can be partially noted in the descriptions and Figures 9, 26, and 28, the pollen sacs of the two species differ somewhat in size and shape. More specifically, the pollen sacs of Matelea aenea have a mean length of 1.13 mm (1.08—1.18) and a mean width of 0.36 mm (0.35—0.37) while those of M, yucatanensis have a mean length of 1.18 mm (1.11- 1.26) and a mean width of 0.33 mm (0.28—0.35). The mean length to width ratio of Matelea aenea is 3.18 (3.04—3.35) and that of M, 153 [F E .9. ‘2 0 rm 4‘ g (B 4- Q 5 cu m 1; O C m 3 0 > (U 2 2 O 4 0 o I. + O \ Lg \ + .\ \ \_\‘ *1 .-. Figure 27. Distribution of Matelea aenea and M. yucatanensis. 154 yucatanensis is 3.56 (3.20—4.22). Of the several potential isolating mechanisms directly associated with pollination, one of the most easily detectable is a change in pollen sac shape which makes hybridization mechanically difficult or impossible. Holm (1950) found that closely related species of Sarcostemma with sympatric or contiguous ranges had markedly different pollen sac shapes. Although the differences in this case are not great and are based on a very inadequate sample, it is in— teresting to speculate that changing pollen sac shape is a factor in the divergence of these two closely related species. SPECIMENS EXAMINED. MEXICO. YUCATAN: Progreso, along railroad near port, Dec 1932 (fl), Flores 1 (F); Progreso, without date (fl), Gaumer 1173 2,2, (F, GH, NY); Progreso, Km 31, Mérida rd, 21 July 1938 (fl), Lundell §_Lundell 8012 (MICH); Progreso, 11-15 Aug 1932 (fl), Steere 3005, type of Dictyanthus aeneus (MICH, M0); at Km 28 on rd from Mérida to Progreso, 28 June 1971 (fl), Stevens 1145 (MSC). 10. Matelea yucatanensis (Standley) Woodson, Ann. Missouri Bot. Card. 28: 237. 1941. Dictyanthus yucatanensis Standley, Publ. Field Columbian Mus., Bot. Ser. 8: 37. 1930. Type: Gaumer 933 (F! holotype; G! fragment of holotype). Plants twining vines. Stems woody below, with thin to thick corky bark, herbaceous and lacking bark above, with dense short and glandular trichomes and dense uncinate long trichomes to 2.5 mm long. Leaf blade ovate to wide—ovate or occasionally very-wide-ovate, (39-) 45—95 mm long, 24-81 mm wide, indumentum of uncinate long trichomes and also 155 glandular trichomes on veins below, surface smooth, apex acuminate to attenuate, base lobate, lobes mostly convergent to descending, with 4— 10 acropetiolar glands, margin often crispate; petiole (22-) 42—57 (—82) mm long, with indumentum of stem. Inflorescence a simple cyme; peduncle 2—9 mm long, with indumentum of stem; bracts linear to lanceo— late, 3—5 mm long, with indumentum of stem or sometimes with long trichomes on margin only and glabrous on adaxial surface; pedicel 3-5 (—7) mm long, with indumentum of stem. Calyx lobes lanceolate to nar— row—ovate, 7~1O mm long, 2.0-3.5 mm wide, apex acute to attenuate, with one gland below each sinus, abaxial surface with scattered short trich— omes, dense glandular trichomes, and scattered uncinate long trichomes or occasionally nearly glabrous, adaxial surface glabrous. Corolla campanulate, base to sinus length (7-) 10-11 mm, limb plane,.revolute; lobes 7—12 mm long, length to width (sinus-sinus) ratio 0.83—1.20, apex acute, plane or slightly reflexed at tip, margin revolute; glabrous within except with sparse to dense short trichomes on limb and lobes, indumentum on outside of short, glandular, and uncinate long trichomes; tube convoluted with raised parts opposite corona lobes and forming deep sacs between them; densely grayish-purple-reticulated, reticula— tions wider in tube. Corona lobes (4-) 5-7 mm long, linear-spathulate in outline, with tip deeply rugose and glistening purplish—black, otherwise deep purple, connate at base, adnate to corolla and adnate by a thin wall to gynostegium, this wall continuing as a narrow ridge nearly the length of lobe. Gynostegium 3.0—3.5 mm high and 2.5—3.0 mm wide at apex, stipitate, apex broadly and shallowly concave with cor— puscula as high points and occasionally also slightly convex in center, terminal anther appendages hardly covering margin of apex. Corpusculum 156 0.24—0.26 mm long, 0.12-0.15 mm wide, pollen sacs 1.11-1.26 mm long, 0.28—0.35 mm wide. Follicles fusiform, ca 95 mm long, ca 15 mm wide, finely mottled green and white, with scattered short and glandular trichomes, with ca 55 thick projections to 4 mm long. Seeds obovate, ca 4.5 mm long, ca 3.5 mm wide, with a raised, faintly radially grooved margin, this entire, inside this margin essentially flat on one side and convex on the other side, both sides deeply verrucate to deeply rugose, dark brown; coma ca 35 mm long. Figure 28. DISTRIBUTION AND ECOLOGY. The identifiable collection localities are scattered in inland areas of the state of Yucatan at elevations probably well below 200 m. Figure 27. Almost certainly to be expected in the adjacent areas of Campeche and Quintana Roo. Growing in low forests and second growth and probably always in limestone—derived soils. Collected flowering June—July. Three collections, Lundell §_Lundell 7885, Steere 2120, and Stevens 1168, have insect damage to the leaves, probably caused by large lepi- dopteran larvae, but none of these insects have been found. COMMON NAMES AND LOCAL USES. The only information available is the name boochin given on the label of Gaumer 933. DISCUSSION. Most closely related to Matelea aenea. For a compari- son of the two species, see the discussion under Matelea aenea. These two species form a distinct unit morphologically, and are geographical— ly isolated from the other species of subgenus Dictyanthus. They are obviously related to the several species grouped with Matelea pavonii but have no clear affinities with any one of the species. Their most conspicuous innovation, besides occupying a unique region (Yucatan Pen— insula) and a unique environment (karst limestone), is the highly Figure 28. Matelea yucatanensis (drawn from Stevens C—158‘ 157 U / a cultivated specimen of Stevens 1168). A. section of flowering stem, x 0.6: B-C. flowers, x 2.3; D. pollinium, x 18; E. base of stem, x 0.6; F. base of adaxial surface of leaf blade, showing acropetiolar glands, x 6. 158 modified tip of the corona lobe. These tips glisten as if they were wet but apparently produce no secretion. They may function as a "pseudo-nectary" in the attraction of pollinators, probably dipterans. It should be noted, however, that the sides of the corona lobes are secretory in apparently the same manner as the species grouped with Matelea pavonii. The "pseudo-nectaries" then, may function as a visual attractant or may have some other function, such as producing an 01— factory attractant, but the major attractant is probably still the nec— tar produced by the corona lobes. These two species are also unique in having predominantly uncinate long trichomes on the internodes, the other species considered here having either entirely straight or only occasionally a few uncinate long trichomes on the internodes. SPECIMENS EXAMINED. MEXICO. YUCATAN: without precise locality and date (fl), Gaumer 933, type of Dictyanthus yucatanensis (F, G, fragment of F specimen); Buena Vista Xbac, without date (fl), Gaumer 1173 ELB- (F, 2 specimens); Chichankanab, without date (fl), Gaumer 1544 (F), 1968 (F, CH); without precise locality and date (f1), Gaumer 31131. 803 (MICH); Chichén Itza, near Xocenpich, June—July 1938 (fl), Lundell 1 Lundell 7885 (MEXU, MICH); Chichén 1:25, 13 June 1932 (fl), Steere 1294 (MICH); Chichén Itzé, near Xnaba cenote, 25 June 1932 (f1), Steere 1548 (MICH, 2 specimens); Chichén Itzé, 25 June 1932 (fl), Steere 1621 p,p, (MICH), 29 June 1932 (fl), Steere 1621 2,2, (MICH, M0); Uxmal, 20—21 July 1932 (fl), Steere 2082 (MICH); Muna, on high ridge, 22-23 July 1932 (fl), Steere 2120 (MICH, MO); Peto, 26-27 July 1932 (fl), Steere 1198 (MICH); along rd from Dzités to Valladolid, ca 4.0 mi NW of Uayma, 1 July 1971 (fl), Stevens 1168 (MSC); greenhouse—grown specimen of Stevens 1168, Aug 1974 (fr), Stevens C—158-3 (MSC). IIIIIIIIIIIIIIIIIIIIIIlIIIII_______________“_“___’_—___F" 159 Species treatments (subgenus unassigned) ll. Matelea altatensis (Brandegee) Woodson, Ann. Missouri Bot. Card. 28: 236. 1941. Gonolobus altatensis Brandegee, Zoe 5: 244. 1908. Type: Brandegee ‘§13,, 10 Sep 1904 (UC! holotype; MO, 2 specimens, fragments of holotype!). Plants twining vines. Stems woody below, with thin to thick corky bark, sometimes with a weakly developed woody caudex with thick corky bark, herbaceous and lacking bark above, with sparse to dense short, glandular, and long trichomes, the long trichomes to 3 mm long and mostly straight. Leaf blade wide—ovate or rarely veryswide-ovate, (35—) 45-75 (-87) mm long, (25—) 35-77 mm wide, with sparse glandular trichomes and sparse to occasionally dense, mostly uncinate long trich— omes, surface pusticulate, especially above, apex acute to attenuate, base lobate, lobes mostly convergent, with (0—) 2—5 acropetiolar glands, margin occasionally somewhat thickened and revolute; petiole 25—52 mm long, with sparse to dense short, glandular, and long trich- omes, long trichomes mostly uncinate. Inflorescence a simple or more often a compound cyme; primary peduncle (15-) 30—135 mm long, with sparse to dense short, glandular, and long trichomes, long trichomes straight or uncinate; bracts linear to lanceolate, 1.5—4.0 mm long, with indumentum of leaf; pedicel (6-) 12-28 mm long, with indumentum of peduncle. Calyx lobes lanceolate to ovate or occasionally elliptic, (2*) 4—6 mm long, l.5~2.5 mm wide, apex acute to attenuate, with one gland below each sinus, abaxial surface with sparse glandular trichomes and sparse to dense, mostly uncinate long trichomes, adaxial surface 160 glabrous or with scattered glandular trichomes. Corolla shallowly cam- panulate, nearly rotate, base to sinus length 4-6 mm, limb not dis— tinct, margin slightly or not at all revolute; lobes (2—) 4—7 mm long, apex acute to obtuse or rounded, slightly reflexed, margins slightly revolute; indumentum within of dense short trichomes except glabrous between corona lobes and especially dense around corona lobes and in a line above them, indumentum on outside of glandular and straight long trichomes, occasionally distal half of lobes nearly glabrous; tube con— voluted, with raised parts opposite corona lobes and shallow pockets formed between them, with corona lobes in distinct pockets in bases of raised parts; pale greenish~white or sometimes also tinted yellowish, especially at base, with very faint to moderately dark green reticula- tions, mostly drying pale brown. Corona lobes ca 2 mm long, basically triangular in outline above, appressed side to side, adnate to gynoste- gium and adherent but not adnate to corolla. Gynostegium ca 2 mm high and ca 2 mm wide at apex, not markedly stipitate, apex plane or slight- ly convex, with a low ridge from each corpusculum to center, this formed from adjacent margins of terminal anther appendages which nearly or completely cover apex. Corpusculum 0.20—0.26 mm long, 0.15—0.18 mm wide, pollen sacs 0.78—0.88 mm long, 0.28—0.35 mm wide. Follicles fus— iform, with a distinct basal flange on one side and apex often long and thin, 60—100 mm long, 13—20 mm wide, striped and mottled light and dark green, glabrous or with sparse short trichomes, with 18—34 (—44) arcu— ate to hooked projections to 8 mm long. Seeds obovate, 4—5 mm long, ca 2 mm wide, with a raised margin, this coarsely toothed, especially dis— tally, inside this margin slightly convex and sparsely verrucate on one side, the other side slightly concave, verrucate, and with a narrow 161 ridge from apex to near center, dark brown; coma ca 35 mm long. Figures 29 and 30. DISTRIBUTION AND ECOLOGY. Matelea altatensis has been collected from northern Sonora to central Sinaloa, but is to be expected farther south, southern Sinaloa being rather poorly collected. Figure 31. Most of the collections have been on the coastal plain at elevations of less than 50 m, but the northernmost localities are more inland and ap— parently up to about 500 m. This species is found in dry thorn forest in heavy clay soils or occasionally in sandy washes. Flowering speci- mens have been collected from late July to mid-September. Two collec— tions with nearly mature fruits were made in September, and one speci— men with completely mature fruits was made in February. Only two of the 12 collections of this species have been made since the 1940's and it is probably not at all common. Particularly with the increased deve10pment of irrigation systems, the coastal plains of this part of Mexico are rapidly being cleared for agricultural purposes, especially for growing cotton. Suitable habitats for Matelea altaten— §1§ are already difficult to find near highways. Although the true abundance of this species cannot be known until its range is better ex— plored, it may well be endangered. Small orange lepidopteran larvae were found feeding on the flowers of Stevens 2062. These have not been reared or identified. A few other specimens showed similar damage to the flowers but no insect dam- age to the foliage has been found. COMMON NAMES AND LOCAL USES. The only information available comes from the label on the specimen collected by Tays in 1912, which gives the local name maguey and the note, "young tender pods are eaten raw by 162 Figure 29. Representative specimen of Matelea altatensis (Wiggins 4 Rollins 140, A). 163 Mulch-n nWLMS (Wu) wands-n M meme.“ In" utuvllh'v nun-mu I)“ M \\ .un-n l‘mnul.“ \lI-\rll\ FLORA P0P THE SORORA]! 1338321 no! Sonora, lulu mam-m I7 mu nun-y nub-nu- .1 aunt-I nuns-a, 'utelea shat ans: 5 (Branden-ac) Woodson .- Jet. R. 5.. ”cease Jr HIHSAHU y ”m V110, mod at b a, 011 hi: ‘ 1 to 20 . ft. over .931 la. llowls‘-cree ~ UM leave: ver rJ fhxlz ;.1luno, 2'? miles “:12 01‘ "er 031110 on the road to 1'1sz 38.1% ‘3. Alt. 720 ft. 1_. .fl' [“LWIDGIIIIRIDC.KIIJ.DII I140 28ml,“ Figure 30. Representative features of Matelea altatensis. A. leaves and old inflorescence, Stevens 2062 (MSC); B. fruit, Wiggins §_Rollins 259 (M0); C. inflorescence, Wiggins §_Rollins 140 (ARIZ); D. flower, Wiggins §_RolllnS 140 (A) . 165 4 In :7: 4 Hill law I . 1 M! l 1 \u C l IMLYQV 166 the natives." DISCUSSION. Although Woodson (1941) included this species in his subgenus Dictyanthus, it lacks the major character that has been used to distinguish the group, the adnation of the digitate corona lobes to the corolla. The corona lobes of this species (in size, shape, and position) are much like those of Matelea tuberosa and M, hemsleyana, but in these species the corona lobes are adnate for their full length to the corolla. The corona lobes of Matelea altatensis are appressed to or perhaps even connivent with the corolla and it may be a small step to complete adnation. I have made a preliminary examination of the developmental stages of Matelea hemsleyana and it will be interest— ing to carefully study the three species together to see if there are any basic developmental differences. Three other characters can more readily be used to distinguish this species from subgenus Dictyanthus. (1) This is the only species considered here in which the terminal anther appendages essentially cover the style apex. In dried flowers these appendages often shrink somewhat, leaving an uncovered spot in the center, but because of the drying become white and easily observ— able (Figure 3OD). In fresh flowers the appendages are translucent and more difficult to see. (2) Also unique among the species considered is the distinct basal flange on one side of the follicle (Figure 303). (3) This species differs from those I have included in Dictyanthus, but compares with the other two species I am treating but have not includ— ed, in having an inflorescence which is, or at least can become, a com— pound rather than a simple cyme. Although this species could be added to subgenus Pachystelma, I prefer not to make a descision until the status of that subgenus is better understood. I i 167 Within the geographic range of Matelea altatensis, there are two species of Matelea not treated here which could be confused with it in fruiting condition. Matelea pringlei (A. Gray) Woodson, which may ac— tually be restricted to Baja California, differs in having longer, thinner, and straight rather than arcuate projections on the follicles and the follicles lack the basal flange. Matelea caudata (A. Gray) Woodson differs in having shorter and thicker follicles which again lack the basal flange. Matelea caudata also tends to be shrubby rather than viney. SPECIMENS EXAMINED. MEXICO. SONORA: Torres, 6 Feb 1903 (fr), Coville 1611 (US); 0.2 mi N of Km marker 2231 and ca 0.2 mi N of side rd to Querobabi, Hwy 15, 28 July 1969 (fl), M§§22.3§2§ (ARIZ, CAS, NY); Bacum Station, near Rio Yaqui, 30-40 m, 7 Sep 1935 (fl), Pennell 16161 (GH, MICH, NY, PH, us); ca 2.2 mi NE of Hwy 15, ca 6.9 mi SE of Ciudad Obregon, 12 Sep 1973 (fl & fr), Stevens 1661_(MSC); 27 mi W of Hermosi— llo on rd to Kino Bay, 720 ft, 28 Aug 1941 (f1), Wiggins 6_Rollins 146 (A, ARIZ, DS, MO, ND, NY, TEX, UC, US); 5 mi N of Suhuoral, 18 mi W of El Camino Nacional (Hermosillo-Guaymas), 3 Sep 1941 (fl & fr), Wiggins 6_Rollins 162_(A, ARIZ, DS, MO, 2 specimens, ND, NY, UC, US). SINALOA: vicinity of Culiacan, Yerba Buena, 10 Sep 1904 (f1), Brandegee ExE-a type of Gonolobus altatensis (MO, 2 specimens, fragments of UC speci— men, UC); Culiacan, 17 Sep 1904 (fl), Brandegee ELE' (POM); Culiacan and vicinity, volcanic cerro and valley, 150—500 ft, Sep 1944 (f1), Gentry 1966-(CH); Maraton, 12 mi W of Culiacan, 100 ft, 21 Sep 1944 (f1), Gentry 1966 (GH, MICH, NY); Los Mochis, July 1912 (fl), 1§y§_§,6. (US). STATE UNKNOWN: without locality and date (f1), Sessé, Mocifio, §£_§1, 1301 (F, fragment, MA, not seen, photo from F neg. 41451 at 168 1x 001 £0052... .s. . 63:82.6 .5. . Distribution of Matelea altatensis and M. se icola. Figure 31. — i 169 MSC), 3570 (F, fragment, MA, not seen, photo from F neg. 41452 at MSC). l2. Matelea sepicola W. D. Stevens, Phytologia 32: 387—392. 1975. Type: Stevens 1436 (MSC! holotype). Plants twining vines. Stems essentially herbaceous and lacking bark except with a woody caudex with thick corky bark, with dense short and glandular trichomes and moderately dense to dense, mostly straight long trichomes to 3 mm long. Leaf blade wide—ovate or occasionally ovate or very—wide—ovate, 35—85 (—105) mm long, 23—85 mm wide, indumentum of dense, or occasionally sparse above, uncinate long trichomes, surface pusticulate to minutely pusticulate or occasionally nearly smooth, smaller veins often slightly to sharply raised below, apex aCuminate to attenuate, base lobate, lobes convergent to divergent, with 0—3 (—5) acropetiolar glands, margin often slightly thickened and revolute; pet- iole (19—) 28—72 (—88) mm long, with dense short and glandular trich— omes and sparse to dense, mostly uncinate long trichomes. Inflores— cence a simple or more often a compound cyme; primary peduncle mostly 2—4 mm long, but occasionally with an inflorescence branch originating . at or near the base of the apparent peduncle (Figure 330), with dense 3 short and glandular trichomes and moderately dense to very sparse, straight or uncinate long trichomes, or occasionally long trichomes : absent; bracts linear to lorate or lanceolate, 1-2 mm long, abaxial surface with dense short and glandular trichomes and moderately dense to dense, straight or uncinate long trichomes, adaxial surface gla— brous; pedicel 1.5—3.5 mm long, with indumentum of peduncle. Calyx lobes narrow—ovate, 3—5 mm long, 1.5—2.5 mm wide, apex attenuate, with 170 one gland below each sinus, abaxial surface with sparse to moderately dense uncinate long trichomes, adaxial surface glabrous. Corolla urce— olate, base to sinus length 3—5 mm, limb slightly reflexed and slightly revolute; lobes 2.5—4.5 mm long, apex acute to obtuse, slightly re— flexed and slightly revolute; glabrous within, indumentum outside of moderately dense straight long trichomes on limb and lobes; tube with a pair of ridges inside opposite each corona lobe; with reddish-brown vertical lines within tube, these becoming circular and reticulated on limb and lobes but partially obscured by the green or greenish—brown background. Corona lobes 1.5—3.0 mm long, shape elaborate but basical— ly thickly laminar and rhombic in lateral view, adnate or tightly con- nivent along axis to corolla (between ridges) for part of length but tip free above, adnate to gynostegium along axis by a narrow wall, loosely to tightly appressed side to side, lateral tips sometimes slightly thickened, giving lobes a trilobed appearance from above. Gynostegium 1.5—3.0 mm high and 1.5—2.0 mm wide at apex, slightly stip- itate, apex broadly and shallowly concave and slightly convex and bi— lobed in center, corpuscula slightly exceeding convex center, terminal W anther appendages covering nearly half of apex. Corpusculum 0.20—0.25 mm long, 0.08—0.09 mm wide, pollen sacs 0.62—0.72 mm long, 0.34—0.43 mm wide. Follicles fusiform, (44-) 54—74 mm long, 12—20 mm wide, green with white stripes, with dense short trichomes and occasionally with very sparse glandular trichomes, with 22—37 (~48) projections, these thick, straight or arcuate, to 4 mm long. Mature seeds unknown; im- mature seeds obovate, to 4 mm long, to 3 mm wide, irregularly toothed distally, both sides verrucate to rugose, dark brown; coma to 30 mm long. Figures 32 and 33. Figure 32. Matelea sepicola (drawn from Stevens 1436). A. section of flowering stem, x 0.6; B—C. flowers, x 3.5; D. pollinium, x 24; E. fruit, x 0.6; F. caudex, x 0.6. 1 172 Figure 33. Representative features of Matelea sepicola. A. inflorescence, note parallel vertical lines within co— rolla, Stevens 1436 (MSC); B. inflorescence, note method of adnation of corona lobes, Stevens 1436 (MSC); C. inflores- cence, one petiole removed, note inflorescence branch orig- inating near base of apparent peduncle, Stevens 1895 (MSC); D. immature fruit, Pringle 5439 (F). 173 \ I. l IHHHHHW HUI 1 H HHHHHH hPrh __,_._._._4._.__::____:.___:_______________..___EF.__ FPbthFFVPLVFDFFP’DrDE C I i 174 DISTRIBUTION AND ECOLOGY. The six known collection localities range from southern Sinaloa through Nayarit to Jalisco. Figure 31. The four localities in Nayarit and Sinaloa are apparently at elevations of 30 m or less and the two in Jalisco at about 1300 m. Flowering August- October. Mature—sized fruits collected in September and November and old, dehisced fruits collected in June and September. The known habi— tats are fencerows, roadsides, and thickets, hence the origin of the epithet. A few leaves of Stevens 1416 and 1616 had minor chewing insect dam— age. One fruit of Pringle 1416 had the seeds partially eaten but I could not determine whether this occurred before or after the specimen was collected. No pollination activity was observed but this species appears to produce significantly more nectar than the other species considered and also appears to produce more fruits, only one of the seven collections lacked fruits in some stage of development. A higher level of seed production might be expected from a species occupying such an ephemeral habitat. i COMMON NAMES AND LOCAL USES. The label of Mexia 1060 notes the name taguarinde for this species. No other names or uses are known. DISCUSSION. Woodson recognized this as an undescribed species by his undated annotation "Matelea (Macroscepis) n.sp." on Ferris 1166 at DS, but annotated another collection, Pringle 1411 at F, as Matelea reticulata (Engelmann) Woodson, probably because the label determina— tion was "Gonolobus reticulatus, Engelm., (with short peduncles)." Matelea reticulata is quite a different species both in morphology (Woodson [1941] placed it in his subgenus "Eumatelea") and in range, being found in northeastern Mexico and adjacent United States. As 175 certainly as this species is distinct from Matelea reticulata, it also does not belong to subgenus Macroscgpis, even as Wbodson (1941) con- ceived it. The six species Woodson referred to Macroscgpis form an unnatural assemblage of which the type element probably deserves gener— ic status. There are at least two other distinct groups represented in the subgenus, but all the other species are probably appropriately placed in the genus Matelea. Matelea sspicola appears to have closer affinities with subgenus Dictyanthus than with any species Woodson in— cluded in Macroscepis. Despite the differences in corolla shape, urce— olate rather than campanulate, and the proportionately broader corona lobes, the flowers are much like those of Dictyanthus, especially with respect to the position and mode of adnation of the corona lobes. The vegetative features, including the indumentum, and the nature of the fruit are identical with those of Dictyanthus and distinct from most of the rest of Matelea. Until Matelea is better studied, I prefer not to assign this species to any of Woodson's subgenera. SPECIMENS EXAMINED. MEXICO. SINALOA: ca 2.0 mi SW of Hwy 15 along rd to Chametla, ca 5 mi 5 of Rosario, 10 Sep 1973 (fl & fr), Stevens 1616 (MSC). NAYARIT: vicinity of San Blas, first hill on old Spanish rd to Tepic, 13 Oct 1925 (fl), Ferris 5506 (DS, US); Tuxpan, Palapar Redondo [labelled as state of Jalisco], 20 m, 6 Nov 1926 (fr), Mexia 1060 (UC); Acaponeta, 23-30 June 1897 (fr),Rose 3122 (US). JALISCO: hills near Tequila, 26 Sep 1893 (fl & fr), Pringle 5439 (F, MO, US); ca 6.9 mi SW of Hwy 15 along rd to Ameca, near dirt rd leading N, 23 Aug 1971 (fl), Stevens 1436, type of Matelea sepicola (MSC), 2 Sep 1973 (f1) , 1895 (MSC) . ugh—.— 176 13. Matelea aspera (Miller) W. D. Stevens, Phytologia 32: 396. 1975. Cynanchum asperum Miller, Card. Dict., ed. 8, no. 6. 1768. Type: Houstoun gig. (BM, not seen, holotype, photos from BH neg. 5251 at MICH! & US!). Gonolobus littoralis Decaisne 16 de Candolle, Prodr. 8: 596. 1844. Type: Galeotti 1545 (P! holotype; G! isotype, F! fragment of G isotype! photo from F neg. 26924 of G isotype at M0!). Vincetoxicum littorale (Decaisne 16 de Candolle) Standley, Contr. U. S. Natl. Herb. 23: 1188. 1924. Vincetoxicum megacarphum Brandegee, Univ. Calif. Publ. Bot. 4: 381. 1913. Type: Pugpus 6014 (UC! holotype; F! G, 2 speci— mens, l a fragment of F specimen! GH! MO, 3 specimens, 2 are fragments, probably of UC specimen! NY! P! isotypes). Matelea mggacarpha (Brandegee) Woodson, Ann. Missouri Bot. Gard. 28: 236. 1941. Pachystelma cordatum Brandegee, Univ. Calif. Publ. Bot. 7: 330. 1920. Lectotype: Purpps 8508 [UC no. 204968, not Purpus 8008 of protologue] (UC! lectotype, mixed with sterile Matelea sp.). Dictyanthus brachistanthus Standley, Publ. Field Columbian Mus., Bot. Ser. 8: 38. 1930. Lectotype: 11316661613351. 6. 6111116 6146_(F! lectotype, mixed with sterile, probably apocyn- aceous, vine, photo from F neg. 51447 of F specimen at F!; C! CH, mixed collection! K! MO! NY! US, 2 specimens, 1 a mixed collection! isolectotypes). ,, 177 Plants twining vines. Stems woody and with thick corky bark below, at least at the base, herbaceous and lacking bark above, with dense short and glandular trichomes and dense to essentially absent, mostly straight long trichomes to 2 mm long. Leaf blade ovate to wide-ovate or occasionally narrow—ovate or very—wide—ovate, 31—98 (~122) mm long, 18—75 (—102) mm wide, indumentum above of sparse to moderately dense uncinate long trichomes and short trichomes on major veins, indumentum below of moderately dense to dense uncinate long trichomes, surface pusticulate to smooth, smaller veins slightly raised below or not, apex acuminate to attenuate, base lobate, lobes convergent to widely diver— gent, with 2—8 acropetiolar glands, margin often slightly thickened and revolute; petiole 15-70 mm long, with dense short and glandular trich— omes and dense to essentially absent, mostly uncinate long trichomes. Inflorescence a simple or more often a compound cyme; primary peduncle 3—31 (~65) mm long, with dense short and glandular trichomes and dense to essentially absent, straight or uncinate long trichomes; bracts lin- ear to lanceolate or lorate, 2-5 (—6) mm long, abaxial surface with dense short and straight or uncinate long trichomes, adaxial surface H with sparse short trichomes or glabrous; pedicel 5-13 mm long, with in— 1 dumentum of peduncle. Calyx lobes lanceolate to narrow—ovate or rarely ; ovate, (3.5—) 5-10 mm long, (l—) 2.0—3.5 (—4) mm wide, apex attenuate, one or occasionally two glands below each sinus, abaxial surface with sparse to dense straight or uncinate long trichomes, surface often pusticulate, adaxial surface glabrous. Corolla shallowly campanulate, nearly rotate, base to sinus length 4—9 (—11) mm, limb broad, hardly distinct from short tube, margin slightly revolute; lobes 3.5-8.0 mm long, apex rounded or occasionally acute or obtuse, plane to somewhat 178 reflexed, margin slightly revolute; indumentum within of dense, very small short trichomes except glabrous at base between corona lobes and especially dense in lines above corona lobes, glabrous outside or with sparse to moderately dense short and sparse straight long trichomes on distal part of limb and bases of lobes, with shallow pockets alternate with corona lobes; background color very pale green (drying cream— white) to deep brownish—green, reticulations from essentially absent on palest backgrounds to dense on darker backgrounds, reticulations purple to brownish-purple. Corona lobes (1.5—) 2—3 (—4) mm long, ovate in outline from above, inflated, with a small tooth on inner surface, lower half adnate to corolla, free above, adnate to base of gynostegi— um, connate at base and forming a fleshy disc partially distinct from lobes. Gynostegium (1.5—) 2—3 (—3.5) mm high and (1.5—) 2—3 mm wide at apex, slightly and shortly stipitate, anther wings prominent, apex es— sentially flat, terminal anther appendages covering ca half of apex. Corpusculum 0.23—0.29 mm long, 0.20—0.26 mm wide, pollen sacs 0.89—1.04 mm long, 0.34—0.43 mm wide. Follicles fusiform, 62—87 (—ll3) mm long, ; 12—18 (—27) mm wide, apparently dark purplish—red or nearly black when I mature but drying to lighter colors, with moderately dense to dense 1 short trichomes, with 18—34 (~46) projections, these straight or slightly arcuate, mostly thick and blunt, to 5 or rarely even 7 mm long. Seeds obovate, 4-5 mm long, 2.0—3.5 mm wide, with a raised, ra- dially grooved margin, this entire or irregularly toothed, especially distally, one side convex and shallowly to deeply verrucate to rugose, other side concave and deeply rugose, with a shallow ridge from apex to near center, dark brown; coma 25—30 mm long. Figure 34. 179 Figure 34. Matelea aspera (drawn from Stevens 1296). A. section of flowering stem, x 0.6; B—C. flowers, x 2.3; D. pollinium, x 18; E. base of stem, x 1.2. 180 DISTRIBUTION AND ECOLOGY. Occurring from Jalisco and Colima south- eastward to central Nicaragua. Figure 35. The areas where this spe— cies has been collected are rather widely spaced northwest of the Isth— mus of Tehuantepec but are more continuous southeastward. Found from sea level to about 1000 m. Tolerant of a variety of substrates, in— cluding limestone derived soils and beach sands, and a variety of com— munities, including pine forests at the highest elevations, but most commonly collected in disturbed thorn forests with clay soils. Flower— ing mostly June to October, but collected flowering once in December and once in February. Considering the number of collections of this species, I doubt the validity of the isolated flowering date in Febru— ary, especially since the specimens concerned appear to be in early stages of flowering. The later flowering dates tend to be from plants of southeast of the Isthmus of Tehuantepec. Mature—sized fruits col- lected July-March. The following collections showed significant chewing insect damage to the leaves: Davidse 88%8401, Emrick 24, 1131338353911. 8. 851£8_8848, Hinton 8182, §1gg_1188, 1155, 1828, MacDaniels 451, McVaugh 18885, 15814, Molina 8..4885, Nelson 1814, Purpus 8914, Standley 11414, 11848, 18851, 18811, Stevens 1158, 1488, and Valerie 8, 5418. Many of the fruits of McVaugh 15814_had conspicuous scars on the outside and the developing seeds were being eaten by weevil larvae. A number of these fruits were brought back to the United States (by me) and after several months the adults emerged. Dr. Daniel H. Janzen has identified the weevil as belonging to the genus Rhyssomatus. I also collected similar weevils from Sarcostemma pannosum (McVaugh 14181) and Marsdenia coulteri? (McVaugh 25285). This weevil is apparently also a seed _ __ _____', .___--4 181 I! 00v Sodom .2 . Distribution of Matelea aspera. Figure 35. 182 predator of Asclepias in Michigan (Janzen, pers. comm.). No pollination activity has been observed but many pollinia were missing from the flowers I examined and this species appears to pro— duce relatively more fruits than most of the other species considered. COMMON NAMES AND LOCAL USES. The name g£g§_has been recorded from the state of México (Hinton 8188). From Honduras the names champerra (Standley 11199), chanchito (Standley 11181), and siempreviva (EEEE 8581) have been noted. No local uses are known. DISCUSSION. Matelea aspera has the greatest geographic range of the species treated here and is the most variable in appearance. Most of the variation, however, is in the size and coloration of the corolla. There is nearly a three—fold difference in the range of corolla sizes; the largest—flowered specimens are found in the Pacific coastal low— lands of the Isthmus of Tehuantepec, and the plants from the mountains of Chiapas southeastward are rather uniformly small—flowered. Corolla color varies considerably, even within populations, but the palest . corollas with faint or no reticulations are all found on Mexican plants and again the plants from the mountains of Chiapas southeastward are rather uniformly dark-colored. The northwestern part of the range 1 tends also to have more substantially woody plants while to the south— east the plants tend to perennate from near ground level. Most of the other characters of the species have less well-marked regional varia- tion and the corona, in particular, appears to be remarkably uniform throughout the range. The type of Cynanchum asperum is apparently from a specimen culti— vated by Miller from seeds sent from Veracruz by Houstoun in about 1730. I have seen only photographs of the holotype, but Dr. Garrett E. f. ‘fiflw"‘__-'_ -— ' ~ 183 Crow compared the specimen, at BM, with samples of my material and con- firmed its identity. Through an apparent printer's error, the protologue of Pachystelma cordatum gives the type collection as Purpus 8008, but the UC accession number given corresponds to the marked type specimen, Purpus 8508. Un— fortunately, Purpus 8508 is a mixed collection. The majority of the sheet is Matelea aspera, but there is a sterile shoot and an isolated leaf of a second species of Matelea. The sterile specimen apparently did not influence the type description and I have therefore chosen the fertile element as the lectotype. It is interesting to note that even the fertile element has only very immature flower buds; this might ex- plain why Brandegee failed to recognize that his new genus and species were the same as Vincetoxicum megacarphum, which he had described seven years earlier from another Purpus collection from near the same local- ity. Brandegee (1922) augmented his type description of Pachystelma cordatum by describing the fruit of a specimen from the type locality; he did not cite the collection number and I have seen no specimen of this species, other than the holotype, from the type locality. In the case of the mixed collection of Dictyanthus brachistanthus, it appears that Standley based the vegetative aspects of the descrip— tion on the sterile, probably apocynaceous, vine and based the descrip— tion of the inflorescence and flowers on the element representing Matelea aspera. The name could probably be rejected on the basis of Article 70 of the International Code of Botanical Nomenclature (Stafleu et a1., 1972), but on the basis of Standley's apparent intent, I have chosen to follow Article 9 and designate the fertile element as the lectotype. 184 This is the type species of Woodson's (1941) Matelea subgenus Pachy— stelma. The previous two species, Matelea sepicola and M, altatensis, could be loosely allied with M, aspera, but the other two species Wood— son included in the subgenus appear to be more distantly related. I am reluctant to add Matelea sepicola and M, altatensis to subgenus Pachy— stelma primarily because all these species have clear affinities with subgenus Matelea section "Reticulatae" and the subgenus Heliostemma. An adequate assessment of the subgenera of Matelea must await careful studies of more of the constituent species. SPECIMENS EXAMINED. MEXICO. JALISCO: steep ravines in gorge of Rio Cihuatlén, below bridge 13 mi N of Santiago, Colima, 175-200 m, 3 July 1957 (f1), McVaugh 15941 (MICH, 2 specimens); mountains 3 mi above (S of) La Huerta, rd to Barra de Navidad, 500—550 m, 3 Oct 1960 (f1 & fr), McVaugh 19805 OMICH, 2 specimens); near new rd ca 25 km NW of Rio San Nicolas and 20 km SE of Tomatlén, 90-150 m, 11—12 Dec 1970 (fr), McVaugh 25314 (MICH, MSC); ca 7 km S of Tomatlén toward Manzanillo- Puerto Vallarta rd, 30-50 m, 15 Feb 1975 (fr), McVaugh 26304 (MICH); 0.5 mi N of La Resolana, 22 mi SSW of Autlan, ca 1000 ft, 11 Aug 1949 (f1), Wilbur 8 Wilbur 2253 (MICH). COLIMA: Paso del Rio, Nov 1906 (fr), Emrick 224 (F); Colima, Aug 1897 (fl), Palmer 164 (MICH, US). MEXICO: Dist. Temascaltepec, Bejucos, 610 m, 26 Aug 1932 (f1), Hinton 1476 (GH, US); Dist. Temascaltepec, Chorrera, 7 Mar 1934 (fr), Hinton 5741 (K), 19 Aug 1935 (f1), 8189 (K, US). VERACRUZ: "dunes de Vera Cruz," June—Oct 1840 (f1), Galeotti 1545, type of Gonolobus littoralis (F, fragment of G specimen, G, photo from F neg. 26924 of G specimen at M0, P); "E. Vera Cruz," 1730 (fl), Houstoun §,n,, type of Cynanchum asperum (BM, not seen, photos from.BH neg. 5251 at MICH & US); vicinity 185 of Palmar, ca 3200 ft, 3 Sep 1935 (fl), MacDaniels 452 (F); Bafios del Carrizal, Aug 1912 (fl), Purpus 6014, type of Vincetoxicum megacarphum (F, G, 2 specimens, 1 a fragment of F specimen, GH, M0, 3 specimens, 2 are fragments, probably of UC specimen, NY, P, UC); Acax6nica, Aug 1919 (fl), Purpus 8508, lectotype of Pachystelma cordatum (UC, mixed with sterile Matelea sp.); ca 4.5 mi W of Palmilla along hwy through Hua— tusco, 10 Aug 1971 (fl), Stevens 1406 (MSC). OAXACA: on Hwy 190, 1.5 mi SE of Niltepec, ca 50 m, 11 July 1972 (f1), Denton 1776 (MICH, MSC, WTU); 9 mi W of Zanatepe [Zanatepec], 17 Aug 1971 (f1), wag£_ E 21, 155_(M0); 5 mi E of Temascal (10 mi W of Veracruz border), ca 45 ft, 25 Oct 1963 (fr), Janzen §,£, (MICH); along Hwy 190, 2 km S of Niltepec, 50 m or less, 17 July 1959 (f1),.§1ggn11g§ (TEX), 1151_(TEX),'1155 (MICH, NY, TEX, US); along Hwy 190, 2 km E of Zanatepec, 50 m or less, 21 July 1959 (f1), 81ggn1821_(MICH, NY, TEX, UC, US); Santa Efigenia, 500 ft, 18 July 1895 (f1), Nelson 2824 (GB); 70 km (by rd) SE of Pino- tepa Nacional on rd to Puerto Escondido, ca 150 m, 23 July 1965 (f1), Roe SE 81. 521 (WIS); near bridge ca 4.0 mi SE of Zanatepec on Hwy 190, 21 July 1971 (f1), Stevens 1296 (MSC); along Hwy 131 ca 3.6 mi N of river bridge near Juchatenango, 27 July 1971 (f1), Stevens 1363 (MSC). CHIAPAS: slopes on bank of Rio Lagas 4 mi SW of Soyala [?Soya16] along rd to Pan American Hwy, 3400 ft, 26 July 1964 (f1), Breedlove 6557 (DS, F, MICH, US); slopes S of Tapanatepec, near Oaxaca-Chiapas state line, 200 ft, 25 Aug 1967 (fl), Clarke 462 (DS); Miramar, 11 Aug 1937 (fl), Matuda 1624 (MEXU, MICH, 2 specimens, MD, NY); Aguas Calientes, Es- cuintla, 21 June 1947 (fl), Matuda 16628 (F, MD); Jalapa, Triunfo, Es- cuintla, 900 m, 10 July 1948 (fl & fr), Matuda 18103 (F); Playa Cin- talapa, Escuintla, 2 June 1949 (fl), Matuda 18657 (F); Valley of Jiquipilas, 2200—2800 ft, 16-18 Aug 1895 (fl), Nelson 1881 (US, mixed with Matelea guirosii); plains near Monserrate, July 1925 (f1), Purpus 18181 (US); rocky plains, Monserrate, June [71930] (fl), Purpus 18815 (UC); Monserrate, June [?l930] (fl), Purpus 18888_p,p, (US); rocky plains, Monserrate, June—July [?1930] (f1), Purpus 18888 pip. (UC). STATE UNKNOWN: without locality and date (fl), §E§§§J Mocifio, E£.§l° 1888 (F, fragment, MA, not seen, photo from F neg. 41465 at MSC), 8588 (MA, not seen, photo from F neg. 41466 at MSC). GUATEMALA. EL PRO— GRESO: along rd between San Ger6nimo and Morazén, near Baja Verapaz line, 1000 m, 9 Oct 1942 (fl), Steyermark 51188 (F, MO). GUATEMALA: 10 km NE of Motfifar, rdside, 15 July 1970 (f1), Harmon 8_8wyg£_8888 (UMO). QUICHE: without precise locality, 1942 (fl), Ignacio 8, 1888 (F). SANTA ROSA: plains of Llano Entero, SE of Chiquimulilla, ca 150 m, 30 Nov 1940 (fr), Standley 18851 (F); region of La Morenita, NE of Chiquimulilla, ca 400 m, 1 Dec 1940 (fl), Standley 18811 (F); along Avellana rd, S of Guazacapén, ca 150 m, 6 Dec 1940 (fr), Standley 18411 (F). SOLOLA: Atitlén, 600 m, Feb 1894 (fl), 8gygg_8_181lg§.1. 8. 82188.8848, lectotype of Dictyanthus brachistanthus (F, mixed with a sterile, probably apocynaceous, vine, photo from F neg. 51447 of F Specimen at F, C, CH, mixed collection, MO, NY, US, 2 specimens, 1 a mixed collection). 88 SALVADOR. MORAZAN: along ditch to reservoir, Monte Cristo, 9 Dec 1941 (fl), Tucker 481 (UC). SAN MIGUEL: NW of Hacienda Potrero Santo, ca 0.1-0.8 km, S side of Lake Olomega, 13°17'N, 88°04'W, ca 60 m, 2 Feb 1942 (fr), Tucker 881 (UC). SAN SALVADOR: San Salvador, 1922 (f1), Calder6n 181 (US). DEPARTMENT UNKNOWN: between San Sabastién and Aculhuaca, 1922 (fl), £3192£§E 1181 (US). 88888888. CHOLUTECA: vicinity of Pespire, 160—200 m, 18—27 Oct 1950 (f1), 187 Standley 11188 (F), 18—27 Oct 1950 (fr), 11181 (F). COPAN: along Copén river between Sta. Rita and Jaral, 700 m, 21 Aug 1971 (fl), Molina 8. 18188 (F, US). EL PARAISO: drainage of Rio Yeguare (ca 87°W, 14°N), entre Mata Indio y Lizapa, 950 m, 25 July 1951 (f1), Molina 8. 4885 (F, GE, US). MORAZAN: drainage of Rio Yeguare (ca 87°W, 14°N), Yeguare River, 2600 ft, 16 July 1948 (f1), Glassman 1818 (F, ILL, MIN, NY); drainage of Rio Yeguare (ca 87°W, 14°N), along Ji- carito Creek, near Jicarito, 950 m, 13 Aug 1947 (fl), Molina 8, 481 (F); vicinity of E1 Zamorano, 780—900 m, 3—17 Aug 1947 (f1), Standley 11588 (F), 11848 (F), 11118 (F); above E1 Zamorano, rd from Jicarito toward El Pedregal, ca 875 m, 14 Aug 1947 (fl), Standley 11188 (F); vicinity of El Zamorano, ca 800 m, 6 Oct 1948 (fl), Standley 11818 (F); near Santa Clara, valley of Rio Yeguare, E of E1 Zamorano, ca 850 m, 19 Oct 1948 (fl), Standley 18181 (F); trail from La Quince, El Zamorano, to El Jicarito, 800-900 m, 15 July 1949 (f1), Standley 11188 (F); near El Jicarito, along rd toward El Pedregal, ca 900 m, 24 July 1949 (fl), Standley 11881 (F); vicinity of El Zamorano, 800—850 m, 26 July 1949 (f1), Standley 11188 (F); region of Rio de Orilla, SE of El Zamorano, 900—950 m, 11 Aug 1949 (f1), Standley 11448 (F, GH); along Quebrada E1 Gallo above E1 Jicarito, 900—1000 m, 12 Aug 1949 (fl), §£22§121.ZZEIZ (F); vicinity of El Zamorano, 800-850 m, 16 Aug 1949 (f1), §£EE§1EX 11888 (F); along rd from E1 Zamorano toward Chagfiite, ca 800 m, 5 Aug 1950 (f1), Standley 18118 (F, GE, US); mountains above El Jicarito, 950 m, 21 Aug 1951 (f1), Standley 18888 (US); Camino Sn. Antonio, 850 m, 21 Oct 1943 (f1), Valerio 8, 1845 (F, MO); vicinity of E1 Zamorano, along rd to Chagfiite, ca 2200 ft, 23 July 1962 (fl & fr), EE§§£E£.§E.§1‘ 12523 (MO); drainage of Rio Yeguare (ca 87°W, 14°N), ca 3 km E of l. I i I 188 Chagfiite, 850 m, 25 Sep 1949 (f1), Williams 16873 (F, CH). VALLE: Salamar Beach, 2 km E of San Lorenzo, Fonseca Gulf, O m, 3 Oct 1968 (f1), Molina 8..8 Molina 11181 (DS, F, G, 2 specimens, MO, NY); San Lorenzo, 20 m, 13 Sep 1945 (f1), Valerio 1, 8418 (F, 2 specimens, GH, M0); lower slopes of El Tigre volcano, above Ampala [Isla El Tigre], 50 m, 16 Sep 1935 (fl), EE§£.2§§Z (GH). NICARAGUA. GRANADA: "Grenade de Nicaragua," Autumno 1869 (fl), 18!y_1811 (P). LEON: Volcén Santa Clara near Hwy 26 [?Volcén Rota], 600 m, 19 July 1970 (f1), Davidse 8_1881 1481 (MSC). CHINANDEGA: Ameya, near sea level, 19—21 June 1923 (fl), 53523.1122 (US); vicinity of Chichigalpa, ca 90 m, 12—18 July 1947 (fl), Standley 11111 (F), 11121 (F), 11411 (F), 11518 (F). DEPARTMENT UNKNOWN: "Leoncia 2.,” 16 Oct 1927 (fl), Chaves [Chavez] 325 (US). 81888 collections Alvarez, C., ex L. C. Smith 470 (8). Anderson, W. R. & C. Anderson 4856 (8); 5146 (4). Anderson, W. R. & C. W. Laskowski 3659 (4); 3692 (2). Anséne, G. 4799 (6); s.n. (1). Bércena, M. de la 536 (4). Barkley, F. A., J. B. Paxon, & C. M. Rowell Jr. 7499 (2). Bell, c. R. & J. A. Duke 16649 (4). Botteri, M. s.n. (6). Brandegee, T. S. s.n. (ll), s.n. (11). Breedlove, D. E. 6475 (l); 6557 (13); 7639 (l). Burch, D. 5270 (4). Calder6n, S. 781 (13); 824 (7); 1017 (l); 1182 (13). Chaves, D. see Chavez, D. 189 Chavez, D. 315 (7); 325 (13). Chévez, Z. 16 (8). Clarke, 0. F. 107 (l); 462 (13). Collins, G. N. & J. H. Kempton 33 (4). Conzatti, C. 2168 (8); 3760 (6); 4628 (4). Conzatti, C. & V. Gonzales 219 (8). Coville, F. V. 1627 (11). Cruden, R. W. 2153 (4). Davidse, G. & R. W. Pohl 2407 (13). Denton, M. F. 1776 (13). Detling, L. E. 8455 (4). Diguet, L. s.n. (4, 5); s.n. (2); s.n. (4). Diquet, L. see Diguet, L. Dressler, R. L. & Q. Jones 252 (1). Dunn, D. B., C. Dziekanowski, & —. Bolingbroke 20451, 20508 (8). Dwyer, J. D. Spellman, J. Vaughan, & R. Wunderlin 755 (13). Elias, T., et al. 800 (7). Emrick, G. M. 224 (13). Faberge, A. C. s.n. (5). Feddema, C. 378, 782 (4). Ferris, R. S. 5506 (12). Flores, R. S. 1 (9). Flores Crespo, J. 12 (4). Galeotti, M. 1394 (4); 1545 (13); 1563 (6); 1567 (8). Gaumer, G. F. 933 (10); 1173 (9, 10); 1544, 1968 (10). Gaumer, G. F., J. D. Gaumer, & G. J. Gaumer 803 (10). Gentry, H. S. 6554 (2); 7065, 7086 (11). 190 Gentry, H. S. & W. B. Fox 12017 (4). Ghiesbreght, A. B. 220 (4); 663 (l); s.n. (4); s.n. (l); s.n. (4); s.n. (4, 8). Ghiesbreght, A. B. ex J. D. P. 37 (4, 8); 38 (4). Glassman, S. F. 1902 (7); 1919 (13). Graham, W. L. 1231 (8). Harmon, W. E. & J. D. Dwyer 3066 (13); 3230 (7); 4208 (l). Hayes, 8. s.n. (7). Heyde, E. T. & E. Lux ex J. D. Smith 3999 (7); 6346 (13). Hinton, G. B. 1476 (13); 1498 (4); 1690, 4328, 4471 (l); 4590 (4); 5741 (13); 8008 (l); 8014 (4); 8189 (13); 8201, 10592, 13275, 14481, 15061 (4). Houstoun, W. s.n. (l3). Ignacio Aguilar, J. 1363 (13). Iltis, H. H., R. Koeppen, & F. Iltis 1198, 1204 (8). Janzen, D. H. s.n. (13). Jurgensen, C. 692 (8). Keck, -. 5 (6). 1": Kerber, E. 35 (6). 4 King, R. M. 1726, 1752, 1755, 1892 (13). Kruse, H. 523 (4). Lagos, J. A. 810 (7). Langlassé, E. 257 (3). Leavenworth, W. C. & —. Leavenworth 917 (4). Lévy, P. 1071 (13). Liebmann, F. M. 12053 (6). Lundell, C. L. & A. A. Lundell 7885 (10); 8012 (9). 191 Lyonnet, E. 1033 (4). MacDaniels, L. H. 452 (13). Martinez, M. 15116 (4). Martinez-Calderén, G. 111 (6). Mason, C. T., Jr. 2895 (11). Matuda, E. 1624 (13); 1778 (7); 16628, 18103, 18657 (13). Matuda, E., et al. 26961 (4); 28802 (8); 31345 (4). Maxon, W. R. 7159 (13). McVaugh, R. 13278 (2); 15826 (4); 15941 (13); 16349 (4); 16393, 16393A (2); 17340 (4); 19805 (13); 19834 (2); 24934 (5); 25314, 26304 (13). Mexia, Y. 589 (4); 1060 (12). Molina R., A. 481 (13); 1084, 4053 (7); 4065 (13); 4121, 18618, 22511 (7); 26209 (13). Molina R., A. & A. R. Molina 22762 (13). Molseed, E. & H. Rice 220 (5). Morales Ruano, J. 1293 (7). Mfiller, F. 1108 (8). Nagel, I. 8038 (4). Nelson, E. w. 1247, 1296 (8); 2824, 2937 (13). \ Oliver, R. L., D. F. Austin, & B. MacBryde 865 (1). Palmer, E. 113 (4); 164 (13); 251 (2). Paray, L. 2114 (4). Pennell, F. W. 19544 (2); 19842 (4); 20207 (11). Pfeifer, H. W. 1633 (7). Pringle, c. G. 2994 (4); 3568 (2); 4468 (4); 4482 (2); 4768 (8); 5431 (5); 5439 (12); 6376 (l); 7364 (4); 8629 (5); 11020 (4); 11627, 13112 (1); s.n. (4); s.n. (l). 192 Purpus, C. A. 206 (l); 2620, 2620A (8); 6014 (13); 8411 (1); 8508 (13); 9075, 9077 (1); 10232, 10615 (13); 10631 (1); 10638 (13); s.n. (l). Reko, B. P. s.n. (4). Roe, K., E. Roe, & S. Mori 521 (13). Rose, J. N. 1645, 1699 (4); 2009 (2); 3122 (12); 3222 (4); 3468, 3473 (2); 4615 (8); 6858 (4); 8586 (8); 10229 (4); 11065 (8); 12863 (2). Rzedowski, J. 20823 (1); 21370, 22677 (8); 26243 (4). Santos, J. V. 2318 (6). Sessé, M., J. M. Mocifio, J. del Castillo, & J. Maldonado 835 (4); 837 (1); 838 (4); 1253 (6); 1300 (13); 1301, 3570 (11); 3580, 3581 (4, 6); 3584 (2); 8568 (13); s.n. (4). Smith, L. C. 595 (6). Standley, P. C. 11217 (13); 11388 (7); 11395, 11474, 11526, 11593, 11648, 11726, 12236 (13); 12709 (7); 12878, 13187 (13); 13671 (7); 21286, 21637, 21736 (13); 22152, 22377 (7); 22446, 22517 (13); 22683 (7); 22686, 26279 (13); 26902 (7); 27100, 27181, 28638 (13); 75105, 75729, 76325, 76662, 76678 (7); 78852, 78871, 79422 (13). Standley, P. C., L. 0. Williams, & P. H. Allen 547 (7). Steere, W. C. 1294, 1548, 1621, 2082, 2120, 2208 (10); 3005 (9). Stevens, W. D. 1145 (9); 1168 (10); 1203 (8); 1245 (7); 1296 (13); 1311, 1343, 1344, 1346 (8); 1362 (4); 1363 (13); 1375 (4); 1392 (6); 1399 (1); 1406 (13); 1427, 1435 (4); 1436 (12); 1453 (4); 1458, 1473 (2); 1895, 2038 (12); 2062 (11); C-158-3 (10). Steyermark, J. A. 31786 (7); 32972, 50732, 51624 (1); 52133 (13). Tays, E. A. H. s.n. (11). Ton, A. S. 4033 (1). Tucker, J. M. 497, 881 (13). 193 Valerio Rodriguez, J. 63, 413, 874, 985, 1082 (7); 1345 (13); 3140 (7); 3473 (13). Villarreal de Puga, L. M. 853 (2); 1651 (4). Weterfall, U. T. 16368 (2). Webster, G. L., K. Miller, & L. Miller 12523 (13). West, J. 3537 (13). White, S. S. 5243 (7). Wiggins, I. L. & R. C. Rollins 140, 259 (11). Wilbur, R. L. & C. R. Wilbur 2079, 2155 (4); 2253 (13). Williams, L. 0. 16873 (13). Williams, L. 0. & A. Molina R. 10569, 11068 (7). Collector unknown (4, 4, 6). Cited scientific names Cynanchum asperum Miller = Matelea aspera l Dictyanthus aeneus Woodson = Matelea aenea D. brachistanthus Standley = Matelea aspera D. campanulatus Reichenbach = Matelea pavonii D. ceratopetalus J. D. Smith = Matelea ceratopetala l D. parviflorus Hemsley = Matelea hemsleyana D. pavonii Decaisne 18 de Candolle = Matelea pavonii D. prostratus Brandegee = Matelea hemsleyana reticulatus (Turczaninow) Bentham & Hooker f. 8§_Hemsley = Matelea dictyantha D. stapeliiflorus Reichenbach = Matelea pavonii . tigrinus Conzatti & Standley 18_Standley = Matelea standleyana D. tuberosus Robinson = Matelea tuberosa 194 Dictyanthus yucatanensis Standley = Matelea yucatanensis Gonolobus altatensis Brandegee = Matelea altatensis G. littoralis Decaisne 18_de Candolle = Matelea aspera Matelea aenea (Woodson) W. D. Stevens, 1229: = No. 9 M. altatensis (Brandegee) Woodson = No. 11 M. aspera (Miller) W. D. Stevens = No. 13 M. ceratopetala (J. D. Smith) Woodson = No. 7 M. dictyantha Woodson = No. 8 M. diffusa Woodson = Matelea hemsleyana M. hamata W. D. Stevens, EEEHJ = No. 3 M. hemsleyana Woodson = No. l M. macvaughiana W. D. Stevens, 1888, = No. 5 M. megacarpha (Brandegee) Woodson = Matelea aspera M. pavonii (Decaisne 12 de Candolle) Woodson = No. 4 M. sepicola W. D. Stevens = No. 12 M. standleyana Woodson = No. 6 M. stapeliiflora (Reichenbach) Woodson = Matelea pavonii M. tuberosa (Robinson) Woodson = No. 2 M. yucatanensis (Standley) Woodson = No. 10 \ Pachystelma cordatum Brandegee = Matelea aspera Rytidoloma reticulatum Turczaninow = Matelea dictyantha Stapelia campanulata Pav6n‘81_Decaisne 18 de Candolle = Matelea pavonii S. campanulata Sessé & Mocifio = Matelea pavonii Tympananthe suberosa Hasskarl = Matelea pavonii Vincetoxicum littorale (Decaisne 18_de Candolle) Standley = Matelea pavonii V. megacarphum Brandegee = Matelea aspera 195 Cited common names Atuz = Matelea aspera and Matelea pavonii Boneta de diablo = Matelea pavonii Bonete = Matelea tuberosa Boochin = Matelea yucatanensis Cabeza de cocodrilo = Matelea aenea Cochita = Matelea ceratopetala Cochitos = Matelea ceratopetala Corazon de loro = Matelea ceratopetala Cuchampel = Matelea ceratopetala Champerra Matelea aspera Chanchito Matelea aspera Chanchitos = Matelea ceratopetala Matelea hemsleyana Chinuna Matelea ceratopetala Chununa Hiedra del monte = Matelea pavonii Maguey = Matelea altatensis Mata chivo = Matelea aenea Pegapega = Matelea hemsleyana Siempreviva = Matelea aspera Sombreritos - Matelea ceratopetala Sombreros = Matelea ceratopetala Taquarinde = Matelea sepicola Xbockin Matelea aenea Yulpate Matelea hemsleyana APPENDIX APPENDIX (TAXONOMIC DATA—MATRIX) A large part of the data gathered in preparing this revision of Matelea subgenus Dictyanthus has been translated into a Morse taxonomic data—matrixl. A listing of this matrix, DICTlM, is shown in Figure 37 and a user—version of the character list is shown in Figure 36. The matrix is in the "general" format and contains data on 19 dichotomous, 14 multistate, and 29 quantitative characters. Although the design and uses of the taxonomic data—matrix have become relatively sophisticated, my use of them is still experimental; DICTlM is an early experiment and should not be used with the same confidence as the acompanying revi- sion. Considerably more effort is needed to most appropriately code the characters useful for asclepiads into data—matrix form. One of the possible uses of a data—matrix is for specimen identification. Using Morse's IDENT4 program on MSU's timesharing system, several sample identifications were made; a few of these are shown in Figure 38. The identifications proceeded easily and directly and were accurate in each case. Although the taxonomic data—matrix definitely warrants further experimental use, it will probably not be of much practical taxonomic value until it can practically be put to more of its potential uses. 1For an explanation of this type of taxonomic data-matrix and its potential uses consult: L. E. Morse, ”Computer programs for specimen identification, key construction and description printing using taxonom— ic data matrices," Publ. Mus. Michigan State Univ., Biol. Ser. 5: 1—128, 1974. 196 E .' i I 11 13 21 23 31 33 41 43 51 53 61 63 71 73 81 83 91 93 101 103 111 113 121 123 131 133 141 143 151 153 161 163 171 173 181 183 191 193 Figure 36. 197 Plant with a woody or fleshy caudex Plant without a caudex Long trichomes of internodes mostly straight Long trichomes of internodes mostly uncinate Leaf surface smooth Leaf surface pusticulate Small veins of leaf distinctly raised below Small veins of leaf not distinctly raised below Apex of largest leaf acute Apex of largest leaf acuminate to attenuate Leaf margin somewhat thickened and revolute Leaf margin not thickened and revolute Glandular trichomes of petiole sparse Glandular trichomes of petiole dense Long trichomes of petiole sparse Long trichomes of petiole dense Glandular trichomes of peduncle and petiole sparse Glandular trichomes of peduncle and petiole dense Long trichomes of peduncle and pedicel sparse Long trichomes of peduncle and pedicel dense Long trichomes of peduncle and pedicel straight Long trichomes of peduncle and pedicel uncinate Inflorescence bracts with glandular trichomes Inflorescence bracts without glandular trichomes Calyx lobe apex acute or acuminate Calyx lobe apex attenuate Calyx lobes with a mixed indumentum Calyx lobes with long trichomes only Corolla limb and lobes with revolute margins Corolla limb and lobes without revolute margins Corolla lobes plane or reflexed Corolla lobes ascending Corolla limb and lobes with short trichomes inside Corolla limb and lobes glabrous inside Inside of corolla tube with short trichomes around and/or above corona lobes Inside of corolla tube glabrous Follicles with glandular trichomes Follicles without glandular trichomes Character list for DICTlM (revised 14 Dec 1975). . ..._.-._ _... _._.._..._.—__ 2001 2002 2004 2101 2102 2104 2108 2201 2202 2204 2208 2216 2301 2302 2304 2308 2316 2401 2402 2404 2408 2501 2502 2504 2601 2602 2604 2701 2702 2704 2708 2801 2802 2804 2808 2816 2901 2902 2904 2908 Figure 198 Stems erect Stems trailing Stems twining Largest leaf narrow—ovate Largest leaf ovate Largest leaf wide—ovate Largest leaf very—wide—ovate Lobes of largest leaf overlapping Lobes of largest leaf convergent Lobes of largest leaf descending Lobes of largest leaf divergent Lobes of largest leaf widely divergent Inflorescence bracts linear or lorate Inflorescence bracts lanceolate Inflorescence bracts narrow—ovate or ovate Inflorescence bracts very—narrow—elliptic, narrow-elliptic, or elliptic Inflorescence bracts narrow—oblong Calyx lobes lanceolate Calyx lobes narrow—ovate Calyx lobes ovate Calyx lobes elliptic Corolla shallowly campanulate (nearly rotate) Corolla campanulate Corolla deeply campanulate (nearly tubular) Corolla lobe apex acute Corolla lobe apex obtuse Corolla lobe rounded Corolla tube with vertical lines within Corolla tube with circular lines within Corolla tube with a distinct reticulate pattern within Corolla tube with no distinct pattern inside Corona lobes subulate Corona lobes short—sagittate Corona lobes short—spathulate with acute tip Corona lobes linear or linear—spathulate, tip not specialized Corona lobes linear—spathulate, tip deeply rugose, dark purple, glistening Corona lobe without teeth Corona lobe with two lateral teeth Corona lobe with one tooth on upper margin Corona lobe with two teeth on upper margin 36 (continued) . .....—.._._.._.—._—_' ' -—H IIIIIIIIIIIIIZ_________________________———7'7’; 3001 3002 3004 3101 3102 3104 3108 3116 3132 3201 3202 3204 3208 3216 3232 3301 3302 3304 3308 3316 3332 34000 35000 36000 37000 38000 39000 40000 41000 42000 43000 44000 45000 46000 47000 48000 49000 50000 Figure 199 Style apex basically concave Style apex convex Style apex apiculate Flowering in June Flowering in July Flowering in August Flowering in September Flowering in October Flowering in November Plants of Sonora Plants of Sinaloa—Nayarit Plants of Durango Plants of Guanajuato—Jalisco-Colima Plants of México—Morelos Plants of Puebla Plants of Michoacén Plants of Veracruz Plants of Oaxaca Plants of Guerrero Plants of Chiapas—Guatemala—El Salvador-Honduras—Nicaragua Plants of Yucatén Erect or trailing stem length in cm Maximum length of long trichomes of stems in 0.1 mm Largest leaf (midrib) length in mm Largest leaf width in mm Acropetiolar gland number Length of petiole of largest leaf in mm Peduncle length in 0.1 mm (absent = 0) Length of largest inflorescence bract in 0.1 mm Pedicel length in 0.1 mm Calyx lobe length in 0.1 mm Calyx lobe width in 0.1 mm Corolla base-sinus length in mm Corolla lobe length to width (sinus-sinus) ratio X 100 Corolla lobe length in 0.1 mm Corona lobe length in 0.1 mm Gynostegium height in 0.1 mm Gynostegium width at apex in 0.1 mm 36 (continued) 200 51000 Corpusculum length in 0.01 mm 52000 Corpusculum width in 0.01 mm 53000 Pollen sac length in 0.01 mm 54000 Pollen sac width in 0.01 mm 55000 Follicle length in mm 56000 Follicle width in mm 57000 Follicle projection number 58000 Seed length in 0.1 mm 59000 Seed width in 0.1 mm 60000 Coma length in mm 61000 Elevation of collection in 10 m 62000 Maximum length of follicle projections in mm Figure 36 (continued) 201 0.9999999999 5» I 231009719500900 99999999990. “80376574501236455>433051944561960 99999999999 701001900 99999999999 gee-6006.......nonhoeoooooo-naco... 99999999999 (5... (( .Ir . . o . . . . . . . . 2274“7.323301123303323025921445950 .01000.00 'I'I’IPOYI' [ LLLLLLLLLLL 1]), . r. . r C 5207J970 1111.11.11.11! 102.11.. (I...) .71.: < .n. .1 . FFFFFFFFFFF H.561. 00000000000 canon...neoco-oooo.ences-noon“..- NNNNNNNNNNN 197500935 v 9 9 I v v I O I o. 1 10090510:47.011W7798665c011911221910 .0000". . 00 00000000000 ,D’IIOIIQDVDIV'OPODD'I'O'OO’IIOSI IDO'O'I'O'I 01057209>7 7109000900 16069377120 13N63163650661757m8592076970997900 10010000000|ETEFEC10 IO'D’OI’OI' none-no-cooncoca-nonco-ooonooooeo N N 77h2Q55Ql7.)\ 50 [ftrr.nq 0 0 N 00320311240.»291971857095960075990 S S 0 90891662224 .00000.00 n D S 1000001133091"’0’OOOPIFODIOVDDDO”OV’DOOO'D n O D vvovvvvivv 669667461009999999950047489398750 H O 0 0 7.. )099992320 O D 7 NY 9’90190990100000000000._.-.._00000000000000 H I EE 21196291119I’!V""DI"IOIUIUOI"I!'9OIIIIII NS M VL L )(0839000 \I INNS FD '9"0'..va5M70705V78001121321110985648601?0 NN F0 cNTN 111321141 [LL 1 00 EVSDOSA 00010000000coo-oooococooooooc-sconce-noon... 89 NF” IS vl ’0'9,0901!137991651169636339729038“.335607.701...) rJDN 510140)») b?100M886661107.2317.7.200011011001056“40551010 701513010N( (If 4llllll COBNA “ on 0'10099099999099099700 99’09'l'l'9’II'I'IIOO [HOFCI flHSS 444474.!““##004121129904605005432850 55055005500390c.87..75c.0 0 RVENAL DI 2103131101400100000000348290669902366466643011111?21110 QA‘EDANAAOS 21313111330 9.0),”!!t! 1N l‘tA'HOV 1131111111951677115510coho-oneconch-onco-c-houens-on... 409 HYer-HF lull. I.) 72238927...) «V... vlrn.LD\u(N 70.9!!!'19.!vvv!”'v'9"11763QP7VW0119H1412717000111710110 HFOAYULOI 47210111113161“). 7 31...)... .... 4|)VV :‘lLL 'QSI'OOOOltlftltttivOftvfil’IIIOI' CSFAOVNATFAIZ7133377229’9990’999' 06565060 IHPHVCAQP.N0L23331311214.510.1.3737135321131231.7.06940?6319007.2?l.7.54377.0 ”FUAAATFIEU 17.17.117.332... 10000 OHTHPHSCDAYNVl5991’199vvovvvtlv0070a.co... nu. .o.. no... co. . e... .0... 0 0727.7... 17.7.1)... 41.30.11.771.31.1.400000000009000000000069h?13fl.1.1.14c. 3AAAAAAAAAAX31229?23334 0510.8?9667011232442220 r. 57.11169076“01112111133533““7M77W7N5?1002m7.u10 00 34M087353300173b333770 055407010301011031317.‘ 7000000000 9'I900009I’.””I",I""' ’07,, 0'1. ’0'!”"”lr",',l"l' 01334567090912.5456?890913395670909123456789 234567890917345678903 . . \r 00 -(r), 0 1, r [ 000000000—0000000000. r 1.! . (rilf 017.37.56.789000123456789001230567890017.3“5678900123456759001234567590 (IDIIII (II .7959‘) [rr.01 455555555555666666666666666666|6666777777777777777:7.77777788888888888 DICTlM, a general format taxonomic data—matrix for Matelea subgenus Dictyanthus. Figure 37. 202 HDS =03u,rLAG= 'NONE’ "1:020'01 : v «(7.9193771990F7. 050957550005_.YN oqooaoo-nnocnl 339959059JQIODS 3790.70.37.93000 A 00..D.00.00:=SL 0000000000 . FR o F 3 H S 7 V1 UXN o 9 N11 H 1. NS ER H 1N IN CTN 3O E BAF 1 3! 0M UWV . .3 r31. 5 E U 7! NC A 1V IE AYE N IE HP EAV A OR AS L01. H :0 X F r 7 50L EF 1E! EVA 0 “NM 9 BUN S ‘17: r... OBI NC. I. 7 L .5 OR FOE R JI IE Urn. O P .0. '10. H .NRO CH NHE SNAP EU 0R8 H HF. .LN .10 F ’AH L SE 0 $18 .I 0E - INT 0.Mv CHO V0 HUIYT T1 Eco vi LEI. V. R E N KU 0 Y 00 FAACS .5 AUDH LF 0F V N DER? GHSLG F! U tOIOA YA)" r_LPF H“ NS no A LS 0 DCAIU NLX PUD NILH FOE FHNE ARLRR U! 0N D. 0...: SFCR ESHUERR<.EFHS DHIOHUSLHHODHQSD HTDFTHXITSFNT71H 0123656789M0011 ‘0.“0‘ '7'- 58888688688999? 7.70:? ‘ 7 7 7 ’ ?233h95566770599 qngqqqqggggqggq Figure 37 (continued) 203 00... M S U TAKUNOMIC DATA MATRIX PROGRAMS at... I PROGRAM IDENT4 DATA MATRIXDICTQM FILE REVISED IADLC75; RELIABILITY 0 MORE DOCUMENTATION-’YES MATELEA SUBGENUS DICTYANTHUS U! WARREN DOUGLAS STEVENS MORL DnCUMhNTATlOV--N0 IDENTIFICATION PROGRAM: I=INSTRUCTIUVS 2=STAR1 --2 ttttt NEW SPECIMEN! COLLECTOR AND NUMBER -- GAUMER FT AL. 1173 VARIABILITY LIMIT '- CHARACTER5261633332123 THE FOLLOWING 2 OF 10 TA‘A REMAIN! MATELEA AEVEA MATELEA YUCATANENSIS NEXT- USEFUL CHARACTERS! 46 26 I5 16 27 37 00 AZ NEXT- 1 CHARACTER5153:I63;27OS SUGGESTED IDENTIFICATION OF SPECIMEN GAUMFR ET AL. II73 MATELEA AENEA UNUSUAL CHARACTERS! 163 2708 2316 3332 50023027 CHARACTERS YOU USED! 23 153 163 2708 28I6 3332 NEXT- 4 ##1## NEW SPECIMEN: COLLECTOR AND NUflBER -- CONZATTI 3760 UARIABILITY LIMIT --1 OIARACTE8536077p36056:31p2004:2502:2702 3 a CF I0 TAXA REMAIN POSSIBLE. ’ NEXT- I CHARACTERSASOIdua7240 SUGGESTED IDENTIFICATION OF SPECIMEN CONZATTI 3760 MATELEA STANDLEYANA , UNUSUAL CHARACTERS: 2702 51045055 52023028 3002 3200 , CHARACTERS YOU USED! 31 2004 2502 2702 36056 45018 I A7240 NEXT- 4 ##t" NEW SPECIMEN! COLLECTOR AND NUMBER -- MOLSEED R RICE 220 VARIABILITY LIMIT --2 UHARACTLRSJbOSZ:3703012001:3205 U 0} IO TAXA REMAIN POSSIBLE- NEKT- I CHARACTERS2208:40050;42170p270| SUGGESTED IDENTIFICATION OF SPECIMEN MOLSEED 8 RICE 220 MATELEA MACVAUGHIANA UNUSUAL CHARACTERS: 2902 57IIKlafl 51023026 55080055 53lanlaH CHARACTERS YUU USED! 200] 2205 270] 3205 36052 37030 40050 42170 NEAT- a Figure 38. Samples of interactive specimen identification using IDENT4 and DICTlM. 204 nu. NEU spacmam COLLECTOR AND numnsn -- srsvanmnx swan VARIABILITY LIMIT --2 ' mARAcransaeoaz.31025.39013.2001.2Ioa i s or no 'rmm REMAIN POSSIBLE. é NDtT- 1 ' mAHACTl-znsaaoso. 430 so “IRE FOLLOWING 3 0F 10 1mm REMAIN: MATELEA HEMSLEYANA : MATELEA TUBEROSA 5 -2 MATaEA CERATOPETALA ' NEXT- 1 I mARAcmnsasox.asooa.27oa i SUGGESTED IDENTIFICATION OF SPECIMEN STEYERMARK 51624 MATELEA HEMSLEYANA UNUSUAL CHARACTERS! 2501 2504 49013015 3002 45003006 CHARACTERS YOU USED: 2001 2108 2501 2704 36022 37025 39013 42050 43050 45004 NEXT- 4 tittt NEU SPECIMEN: COLLECTOR AND NUMBER -- PRINGLE 4482 UARIABILITY LIMIT --2 CHARACTERSJ6O32:37026:39013n11:200112504 5 OF 10 TAXA REMAIN POSSIBLE. NEXT- 1 CHARACTERS3208;2701;4I THE FOLLOWING 2 OF 10 TAXA REMAIN: MATELEA TUBEROSA -I MATELEA MACVAUGHIANA NEXT- 0 USEFUL CHARACTERS! 26 28 42 47 29 48 49 50 NEXT- I CHARACTERS2802-42050 SUGGESTED IDENTIFICATION OF SPECIMEN PRINGLE 4482 MATELEA TUBEROSA UNUSUAL CHARACTERS! 2802 3300 49018022 2504 50018022 CHARACTERS YOU USED: 11 41 2001 2504 2701 2802 3208 36032 37026 39013 42050 NEXT- 4 tttti NEW SPECIMEN! COLLECTOR AND NUMBER -- PENNELL 19842 VARIABILITY LIMIT -- OIARACTERS39045n40180036060137060 6 OF 10 TAXA REMAIN POSSIBLE. NEXT- 1 CHARACTERSIU1p2701;45020:2502 01E FOLLOWING 2 OF 10 TAXA REMAIN: MATELEA PAVONII -I MATH.EA STANDLEYANA NEXT- 1 CHARACTERS42102;4312014404012604 NO PROGRESS NEXT' _ USEFUL CHARACTERS! 51 52 55 57 47 3% 14 30 NEXT- OIARACTERS47150 SUGGESTED IDENTIFICATION OF SPECIMEN PEVNELL 19542 MATELEA PAVONII UNUSUAL CHARACTERS! 2701 51031035 34000000 57029044 3004 CHARACTERS YOU USED! 151 2502 2604 2701 36080 37060 39045 40180 42102 43120 44040 45020 47150 NEXT- 7 STOP Figure 38 (continued) TII. .1..IIII-.. .- m u M m m u LITERATURE CITED Anderson, W. R. 1972. A monograph of the genus Crusea (Rubiaceae). Mem. New York Bot. Gard. 22: 1-128. Anonymous. 1852. Dictzanthus campanulatus, Jordan. Gard. Companion Florists' Guide 1: 20—21. ----- . 1853. Dictyanthus pavonii. Bot. Mag. 79: t. 4750. ————— [B. 0.]. 1857. Dictzanthus stapeliaeflorus Rchb. und 2, Pavonii Decne. Gartenflora 6: 65—66, t. 187. ----- [B. J.] & D. Beaton. 1852. Bell-flowered dictyanth. (Dicty— anthus campanulatus). Cottage Gardener 8: 236-237. Baum, H. 1949. Die Stellung der Samenlagen am Karpell bei Asclepias sxriaca, Cznanchum Vincetoxicum, und Erythraea centaurium. Oesterr. Bot. Z. 95: 251—256. Bentham, G. & J. D. Hooker. 1876. Genera Plantarum . . ., Asclepi— adeae. London. Vol. 2: 728—785. Boke, N. H. 1947. Development of the adult shoot apex and floral initiation in Vinca rosea L. Amer. J. Bot. 34: 433—439. Brandegee, T. S. 1922. Plantae Mexicanae Purpusianae, XI. Univ. , Calif. Publ. Bot. 10: 181-188. \ Brown, R. 1810. On the Asclepiadeae, a natural order of plants separated from the Apocineae of Jussieu. Mem. Wern. Nat. Hist. Soc. 1: 12-78. ----- 1833. On the organs and mode of fecundation in Orchideae and Asclepiadeae. Trans. Linn. Soc. London 16: 685—745, p1. 34-35. Brunaud, A. 1968a. Démonstration de la nature sympodiale des pousses fertiles chez le Vinca rosea L. Compt. Rend. Hebd. Séances Acad. Sci. 266(D): 92-94. ————— . 1968b. Mise en evidence de deux mécanismes différents dans le phénoméne d'usurpation, chez les plantes a pousse sympodiale. Compt. Rend. Hebd. Séances Acad. Sci. 266(D): 468—470. 205 206 Brunaud, A. 1968c. La pousse fertile de l'Ogypetalum coeruleum Dec. (Asclépiadacée a inflorescences "interpétiolaires") est construite suivant 1e type monopodial. Compt. Rend. Hebd. Séances Acad. Sci. 266(D): 11116—1119. ————— . 1968d. Sympodisation dans les pousses fertiles et structure de 1'inflorescence chez l'Hoya bella Hook. (Asclépiadacée a inflores— cences "interpétiolaires"). Compt. Rend. Hebd. Séances Acad. Sci. 267(0): 1709—1712. ————— . 1969a. Modifications spontanées du systéme des c6tes chez des Asclepiadaceae cactiformes. Mém. Soc. Bot. France 1969, pp. 29— 43. ————— . 1969b. Une pousse ramifiée de constitution exceptionalle chez les Plantes a fleurs: cas de l'Hoya multiflora Bl. Compt. Rend. Hebd. Séances Acad. Sci. 268(D): 1048—1051. ————— . 1970a. Une mise a fleur incomplete? La transformation en épine du méristéme terminal chez une Apocynacée: Carissa macro— carpa (Eckl.) A. DC. Rev. Gén. Bot. 77: 97—109. ----- . 1970b. Sur l'organization différentielle de l'inflorescence chez les Asclépiadacées monopodiales et sympodiales. Bull. Soc. Bot. France 117: 250—258. ————— . 1971a. La notion de rameau occupant un site foliaire et le cas des Rauwolfia. Mém. Soc. Bot. France 1971, pp. 125—132. ————— . 1971b. Continuité des hélices foliaires dans la pousse des Asclépiadacées monopodiales a inflorescences "extra—axillaires." Mém. Soc. Bot. France 1971, pp. 133—142. Carlquist, S. 1959. Glandular structures of Holocarpa and their ontogeny. Amer. J. Bot. 46: 300—308. Chauveaud, G. 1891. Recherches embryogéniques sur l'appareil l laticifére des Euphorbiacées, Urticacées, Apocynées et Asclepi adées. Ann. Sci. Nat. Bot., sér. 7, 14: 1—162. Chemsak, J. A. 1963. Taxonomy and bionomics of the genus Tetraopes (Cerambycidae: Coleoptera). Univ. Calif. Publ. Entomology 30: 1-90, pl. 1-9. Corry, T. H. 1883. On the mode of development of the pollinium in Asclepias cornuti Decaisne. Trans. Linn. Soc. London, Bot. 2: 75—84, pl. 16. ----- . 1884. On the structure and development of the gynostegium, and the mode of fertilization in Asclepias cornuti. Trans. Linn. Soc. London, Bot. 2: 173—207, pl. 24—26. 207 Cronquist, A. 1968. The Evolution and Classification of Flowering Plants. Houghton Mifflin. Boston. Davis, G. L. 1966. Systematic Embryology of the Angiosperms. Wiley. New York. Davis, H. B. 1936. Life and Work of Cyrus Guernsey Pringle. 1838— 1911. Univ. of Vermont. Burlington. Decaisne, J. 1844. Asclepiadeae. In A. P. de Candolle, "Prodromus Systematis Naturalis Regni Vegetabilis." Vol. 8: 490—665. Demeter, K. 1922. Vergleichende Asclepiadeenstudien. Flora 115: 130— 176. ————— 1923. fiber "Plasmoptysen”-Mykorrhiza. Flora 116: 405—456, c. o Deshpande, B. D. & P. Joneja. 1962. Morphology and embryolo y of Leptadenia pyrotechnica Dcne. Phyton (Buenos Aires) 19: 73—84. Dop, P. 1902. Sur le développment de l'ovule des Asclépiadées. Compt. Rend. Hebd. Séances Acad. Sci. 135: 800-803. _.. :. __...__..__ -- .- ————— . 1903. Sur l'ovule et la fecondation des Asclépiadées. Compt. Rend. Hebd. Séances Acad. Sci. 136: 250—252. Drapilik, D. J. 1970. A biosystematic study of the genus Matelea in the southeastern United States. Ph.D. Thesis. University of North Carolina at Chapel Hill. Dyer, R. A. 1971. New and interesting records of African plants. Asclepiadaceae. Bothalia 10: 363-365. Ehrlich, P. R. & P. H. Raven. 1964. Butterflies and plants: A study . in coevolution. Evolution 18: 586—608. ‘ Esau, K. 1965. Plant Anatomy. (second edition) Wiley. New York. Federov, A. A. (ed.) 1969. Khromosomnye Chisla Tsvetkovykh Rasteny. Leningrad. [In Russian]. Francke, A. 1927. Zur Kenntnis der Exodermis der Asclepiadaceen. Planta 3: 1-26. Frye, T. C. 1902. A morphological study of certain Asclepiadaceae. Bot. Caz. (Crawfordsville) 34: 389*412, pl. 13-15. Galil, J. & M. Zeroni 1969. On the organization of the pollinium in Asclepias curassivica. Bot. Caz. (Crawfordsville) 130: 1—4. l‘ffi ”— 208 Gilbert, L. E. 1971. Butterfly—plant coevolution: Has Passiflora adenopoda won the selectional race with heliconiine butterflies? Science 172: 585-586. Goldman, E. A. 1951. Biological investigations in México. Smith- sonian Misc. Collect. 115: xiii, 1-476, pl. 1—70, frontisp. Good, R. 1952. An atlas of the Asclepiadaceae. New Phytol. 51: 198— 209. ————— 1956. Features of Evolution in the Flowering Plants. Longmans, Green and Co. London. Groom, P. 1893. On Dischidia rafflesiana (Wall.). Ann. Bot. (London) 7: 223-242, pl. 10. Guignard, L. 1903. Remarques sur la formation du pollen chez les Asclépiadées. Compt. Rend. Hebd. Séances Acad. Sci. 137: l9~24. Hallier, H. 1912. L'origine et le systeme phylétique des angio- spermes. Arch. Néerl. Sci. Exact. Nat., ser. 3B, 1: 146-234. Handa, T. 1936. Anatomical observations on the internal cambium of the stem of Marsdenia tomentosa Morr. et Decne. Jap. J. Bot. 8: 59—64 0 Hemsley, W. B. 1882. Asclepiadaceae. 12_F. D. Godman & O. Salvin (eds.), "Biologia Centrali—Americana. Botany." Vol. 2: 318-338, t. 56. Hickey, L. J. 1973. Classification of the architecture of dicotyledonous leaves. Amer. J. Bot. 60: 17—33. Holm, R. W. 1950. The American species of Sarcostemma R. Br. (Asclepiadaceae). Ann. Missouri Bot. Card. 37: 477-560. Jonsson, G. 1968a. Etude ontogénique du Stapelia variegata L.: Le point végétatif de la germination a la plante adulte. Rev. Gén. Bot. 75: 329-376. ----- . 1968b. Etude de la ramification chez trois Stapelia. Rev. Gén. Bot. 75: 441—464, pl. 1—2. Krishnamurthy, K. H. & B. Kannabiran. 1970. Histomorphology of foliar epidermis and pharmacognosy in Asclepiadaceae. J. Indian Bot. Soc. 49: 105-114. Langman, I. K. 1964. A Selected Guide to the Literature on the Flowering Plants of Mexico. Univ. of Pennsylvania Press. Philadelphia. Lawrence, G. H. M. 1951. Taxonomy of Vascular Plants. Macmillan. New York. 209 Leach, L. C. 1970. Stapelieae from South Tropical Africa, V. Bothalia 10: 45—54. Levin, D. A. 1973. The role of trichomes in plant defense. Quart. Rev. Biol. 48: 3-15. Liskins, H. F. & M. L. Suren. 1969. Die Entwicklung des Polliniums von Asclepias curassavica. Ber. Deutsch. Bot. Ges. 82: 527—534. The travels and botanical collections of Eugéne McVaugh, R. 1951. Candollea 13: 167— Langlassé in Mexico and Colombia, 1898—1899. 211. ————— 1956. Edward Palmer. Plant Explorer of the American West. Univ. of Oklahoma Press. Norman. Mayberry, M. W. 1938. An anatomical study of Asclepias pumila, A, incarnata and A. guadrifolia. Trans. Kansas Acad. Sci. 41: 127— 137. Metcalfe, C. R. & L. Chalk. 1950. Anatomy of the Dicotyledons. Oxford University Press. London. Moore, R. J. 1946. Investigations on rubber—bearing plants. III. Development of normal and aborting seeds in Asclepias syriaca L. Canad. J. Res., Sect. C, Bot. Sci. 24: 55—65. ----- 1973. Index to Plant Chromosome Numbers 1967—1971. Regnum Veg., Vol. 90. Utrecht. ————— 1974. Index to Plant Chromosome Numbers 1972. Regnum Veg., Vol. 91. Utrecht. Morren, C. 1952. Le Tympananthe suberosa, magnifique Asclépiadée de pleine terre pour l'été. Belgique Hort. 2: 73, t. 11. Mulay, B. M., D. P. Deshpande, & U. Tolani 1965. Studies in Asclepiadaceae 11. Floral morphology and gametogenesis in certain members of the Asclepiadaceae. J. Indian Bot. Soc. 44: 95-104. Nolan, J. R. 1966. Xeromorphic leaf adaptations in Arizona milkweeds. Plateau 39: 94—101. ————— . 1967. Developmental morphology of the Asclepiadaceous inflorescence. Ph.D. Thesis. Cornell University. ----- 1969. Bifurcation of the stem apex in Asclepias syriaca. Amer. J. Bot. 603—609. Payer, J. B. 1857. Traité d'Organogénie Comparée de la Fleur. Paris. Planchon, J. E. & L. Van Houtte. 1852—1853. Dictyanthus pavonii. Fl. Serres Jard. Eur. 8: 55—56, pl. 6, 770. 210 1912a. Etude anatomique de quelques eSPECes d'Asclépiadées Puech, G. aphylles de l'Ouest de Madagascar. Thesis. Paris. . 1912b. Etude anatomique de quelques espéces d'Asclépiadées aphylles de l'Ouest de Madagascar. Rev. Gén. Bot. 24: 298-312, 329-343. Rao, V. S. & A, Ganguli. 1963. The floral anatomy of some Asclepiadaceae. Proc. Indian Acad. Sci. 57: 15—44. 1974. Angiosperm biogeography and past Raven, P. H. & D. I. Axelrod. Ann. Missouri Bot. Card. 61: 539-673. continental movements. ----- . & -—-—-. 1975. History of the flora and fauna of Latin America. Amer. Sci. 63: 420-429. Richharia, R. H. 1934. The number of microsporangia in each stamen in Asclepiadaceae. Curr. Sci. 2: 340—342. fiber die Gattung Marsdenia R. Br. und die Stamm~ Bot. Jahrb. Syst. 52: 354-434. Rothe, W. 1915. pflanze der Condurangorinde. Roy Tapadar, N. N. 1964. Cytotaxonomic studies in Apocynaceae and delineation of the different evolutionary tendencies operating within the family. Caryologia 17: 103-138. Safwat, F. M. 1962. The floral morphology of Secamone and the evolution of the pollinating apparatus in Asclepiadaceae. Missouri Bot. Card. 49: 95-129. Ann. Sass, J. E. 1958. Botanical Microtechnique. (third edition) Iowa State College Press. Ames. Schaffstein, G. 1932. Untersuchungen an ungeliederten Milchréhren. Beih. Bot. Centralbl. 49: 197—220. IE_A. Engler & K. Prantl, "Die Schumann, K. 1895. Asclepiadaceae. 189-306. natfirlichen Pflanzenfamilien." Vol. 4(2): Scott, D. H. & G. Brebner. 1891. On internal phloem in the root and stem of dicotyledons. Ann. Bot. (London) 5: 259-300, p1. 18—20. ————— . & E. Sargent. 1993. On the pitchers of Dischidia rafflesiana (Wall.). Ann. Bot. (London) 7: 243-269, pl. 11-12. Sessé, M. & J. M. Mocifio. 1887-1890. Plantae Novae Hispaniae. Mexico [published in 9 parts as appendices to Naturaleza (Mexico City) ser. 2, Vol. 1]. The origin and distribution of inter- and intraxylary Singh, B. 1943. Indian Acad. Sci. 18: 14-19. phloem in Lgptadenia. Proc. 211 [translated by L. A. Boodle & F. E. Fritsch, Systematic Anatomy of the Dicotyledons. Solereder, H. 1908. revised by D. H. Scott] Clarendon Press. Oxford. Las colecciones botanicas de C. A. Purpus en Sousa S., M. 1969. Univ. Calif. Publ. Bot. 51: ix, 1—36, Mexico. Periodo 1898-1925. frontisp. Stafleu, F. A. 1967. Taxonomic Literature. Regnum Veg., Vol. 52. Utrecht. ----- ., C. E. B. Bonner, R. McVaugh, R. D. Meikle, R. C. Rollins, R. Ross, J. M. Schopf, G. M. Schulze, R. de Vilmorin, & E. G. Voss (eds.). 1972. International Code of Botanical Nomenclature. Regnum Veg., Vol. 82. Utrecht. Standley, P. C. 1924. Trees and shrubs of Mexico. Asclepiadaceae. Contr. U. S. Natl. Herb. 23: 1166-1194, 1680. -—-—-. & L. 0. Willimas. 1969. Flora of Guatemala. Apocynaceae and Asclepiadaceae. Fieldiana, Bot. 24(8): 334—472. New York. Stearn, W. T. 1966. Botanical Latin. Hafner. 1971. The Asclepiadaceae, Apocynaceae, Loganiaceae, Stevens, W. D. M.S. and Buddleiaceae in Hassan District, Mysore State, India. Thesis. Michigan State University. . 1975. Notes on the genus Matelea (Apocynaceae, §.l.). Phytologia 32: 387-406. A new name, Marsdenia lanata (Apocynaceae, §,l,), ----- . In press. Southw. Naturalist. and notes on the Species. Synopsis of a putatively phylogenetic Thorne, R. F. 1968. A1180 6: 57—66. classification of the flowering plants. Inclusion of the Apiaceae (Umbelliferae) in the ————— . 1973. Notes Roy. Bot. Gard. Edinburgh 32: 161—165. Araliaceae. Tomlinson, P. B., M. H. Zimmermann, P. G. Simpson. 1970. Dichotomous and pseudodichotomous branching of monocotyledonous trees. Phytomorphology 20: 36—39. Ueber den anatomischen Bau des Stammes der Treiber, K. 1891. Bot. Centralbl. 48: 209—218, 241—249, 273—281, 305- Asclepiadeen. 313, tt. 1—2. 1932. Note sur l'anatomie de la feuille d'une Trochain, J. Asclépiadée: 1e Leptadenia lanceolata Decne. Bull. Soc. Bot. France 79: 28—32. 212 Asclepiadeae quaedam hucusque indescriptae. Turczaninow, S. 1852. Bull. Soc. Imp. Naturalistes Moscou 25(2): 310—325. In W. Zimmermann & P. G. Ozenda 1962. Plant hairs. Abt. I, Vol. 4(5): 1— Uphof, J. c. T. (eds.), "Handbuch der Pflanzenanatomie. 206. Vesque, J. 1885. Caractéres des principales familles Gamopétales tirés de l'anatomie de la feuille. Ann. Sci. Nat. Bot., sér. 7, 1: 183-360, pl. 9—15. Zur Kenntnis der Pollinien der Asclepiadaceen. 62: 68—72. In H. Melchior (ed.), " A. Vol. 2: 414-417. N a Volk, O. H. 1949. Ber. Deutsch. Bot. Ges. wagenitz, G. 1964. Asclepiadaceae. Engler's Syllabus der Pflanzenfamilien watson, S. 1887. Contributions to American botany. XXI. 1. List of plants collected by Dr. Edward Palmer in the State of Jalisco, Daedalus 22: 396-465. Mexico, in 1886. Tropical American plants, X. Fieldiana, Bot. Williams, L. 0. 1968. 32: 35—61. woodson, R. E., Jr. 1933. The floral anatomy and probable affinities of the genus Grisebachiella. Bull. Torrey Bot. Club 62: 471-478. Amer. J. Bot. 22: New Apocynaceae and Asclepiadaceae. -—---. 1935. 684-693. ———--. 1941. North American Asclepiadaceae. I. Perspective of the genera. Ann. Missouri Bot. Gard. 28: 193—244. ————— . 1954. The North American species of Asclepias. Ann. Missouri Bot. Card. 41: 1—211, p1. 1—3. The vascular anatomy and comparative Bull. Torrey Bot. Club 65: ---—-. & J. A. Moore. 1938. morphology of apocynaceous flowers. 135-165, p1. 3—5. Anatomische Untersuchungen an Pflanzen der Namibwfiste Flora 133: 365—416. Zemke, E. 1939. CDeutsch-Sfidwestafrika). Mlcllmlmmm “WWW”!film/AIR!“ 31293 03175 4298