REPRCDUCTIVE PROCESSES AFFECTING THE TAXONCMY OF SOME MEXICAN MID CEI‘I‘IRAL AMERICAN SPECIES OF EUPATCRIUM Thesis for the Degree of 9M. S. MICHIGAN. STATE UNIVERSITY IEROLD LEE GRASHOFF 1969 LIBRA P Y MiChig‘“‘ NC I Univt V“! " Ii‘g‘f. i I I...‘--fiy-"M~zi.w‘ THESI‘ I! LmHAHY"""’ ‘g‘ smuma:_____ ABSTRACT REPRODUCTIVE PROCESSES AFFECTING THE TAXONOMY OF SOME MEXICAN AND CENTRAL AMERICAN SPECIES OF EUPATORIUM BY Jerold Lee Grashoff Eupatorium is one of the two largest genera in the shrubs, Compositae. It consists of an estimated 1200 herbs, vines, and trees which are distributed chiefly in the American tr0pics. The genus has not been treated system- atically for over one hundred years and is often considered to be taxonomically difficult. The four parts of this thesis may ultimately c0ntribute to a systematic treatment of the genus. Cytological and histological studies were made on young ovules of EUpatorium muelleri Sch. Bip ex Klatt, section section Subimbricata. Eximbricata, and g, macrophyllum L. f., The former is a tetraploid, obligate apomict. AgamOSpermy in this species was observed to be that of diplospory by somatic division. Sizeable embryos are formed before anthe— sis begins. The latter, a diploid, sexual species was observed to have normal meiosis resulting in a linear tetrad, the chalasal cell of which usually develops into the mature Jerold Lee Grashoff gametOphyte. Antipodal cells and nuclei in g, macrophyllum are variable in number. An investigation of the contents of anthers of pre— anthesis florets of 1,048 specimens representing 192 Species was made to provide an estimate of the extent of apomixis in Mexican and Central American species of EUpatorium. On the basis of this examination apomixis is thought to occur in 35 (18.2 per cent) of the Species studied. Three sections of the genus contain possible apomicts. It is hypothesized that apomixis has deve10ped independently in each of these sections. Concentrated systematic effort in a genus in which apomixis occurs has often resulted in taxonomic confusion. Much of the taxonomic confusion in Mexican and Central American Eupatoria may be the result of an inadequate under— Some sug- standing of the extent of apomixis in the genus. gested guidelines are given for the monographic treatment of apomictic species in a genus such as EUpatorium. Wind pollination in the Compositae has heretofore been known to occur only in the tribe Anthemideae and the sub- tribe Ambrosinae of the Heliantheae. Eight species of Eupatorium were observed to have morphological adaptations thought to be indicative of anemOphily. These species generally have copious, small, relatively smooth pollen, elongated inflorescences, increased stigmatic surfaces on Jerold Lee Grashoff the style branches, and reduced apical anther appendages. On the basis of morphological evidence it is concluded that anemophily has developed in the Eupatorieae independ— ently from that of the other tribes and that, as exhibited by these eight species, the adaptation to anemOphily in Eupatorium is in its incipient stages. REPRODUCTIVE PROCESSES AFFECTING THE TAXONOMY OF SOME MEXICAN AND CENTRAL AMERICAN SPECIES OF EUPATORIUM BY Jerold Lee Grashoff A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Botany and Plant Pathology 1969 «f; :26 /? 7 é—zé-w ACKNOWLEDGMENTS I wish to express my sincere gratitude to my major professor, Dr. John H. Beaman for his enthusiastic guidance during this study. I appreciate also the helpful suggestions concerning the thesis provided by Dr. Kenneth C. Sink and Dr. John E. Cantlon. I am also indebted to the curators of the Field Museum Herbarium, The University of Michigan Herbarium, and the Michigan State University Herbarium for the use of their collections. A collecting expedition to Mexico and Central America was sponsored by the Institute of International Agriculture and Nutrition. ii TABLE OF CONTENTS 0 O O O O C GENERAL INTRODUCTION. . . . . . . . . CYTOLOGICAL AND EMBRYOLOGICAL OBSERVATIONS. . Introduction . . . . . . . . Eupatorium muelleri. . . . . . MicrOSporogenesis . . . . . Megasporogenesis. . . . . . . . Megagametogenesis in Eupatorium macrophyllum A SURVEY OF THE CONTENTS OF ANTHERS OF MEXICAN AND CENTRAL AMERICAN EUPATORIA . . . . . . . . Introduction . . . . . . Materials and Methods. . . . . . Remarks. . . . . . . . . . . . . PROBLEMS IN THE TAXONOMIC TREATMENT OF APOMICTS E UPAT 0 RI UM C C O O O C O O O O O 0 Introduction . . . . . . . Suggestions Concerning the Taxonomy of Apomicts APPARENT WIND POLLINATION IN EUPATORIUM . . . Introduction . . . . . . . . Description of Putative Anemophilous Species Morphological Evidence . . . . . . Discussion . . . . . . . . GENERAL DISCUSSION. . . . . . . . . BIBLIOGRAPHY. . . . . . . . . . . APPENDIX-—Specimens Used and Pollen Quality in the Investigation of Pollen Quality Observed of Mexi- can and Central American Species Of EUpatorium . iii Page NI“ 28 29 37 59 59 42 46 46 47 48 57 61 68 LIST OF TABLES AND FIGURES TABLE Page 1. Summary of pollen quality in Mexican and Central American Species of Eupatorium. . . . 52 FIGURE 1. EUpatorium muelleri. MicrOSporocyteS . . . . 11 2-7. Eupatorium muelleri. Embryo sac development. 16 8-13. Eupatorium muelleri. Embryo sac develOpment (continued) and early embryogenesis . . . . . 19 14-20. Eupatorium macrOphyllum. Megagametogenesis . 22 21—24. Eupatorium macrOphyllum. Megagametogenesis (continued) . . . . . . . . . . . . . . . . . 24 25-29. Eupatorium macrOphyllum. Chalazal portion of embryo sacs Showing variation in the number of antipodal cells and nuclei . . . . . . . . 26 50. Apical anther appendages. . . . . . . . . . 52 51. Apical portion of style branches. . . . . . . 52 32. Representative pollen grains photographed with a scanning electron microsc0pe . . . . . 56 iv REPRODUCTIVE PROCESSES AFFECTING THE TAXONOMY OF SOME MEXICAN AND CENTRAL AMERICAN SPECIES OF EUPATORIUM GENERAL INTRODUCTION Eupatorium iS one of the two largest genera in the Compositae. It contains an estimated 1200 Species of herbs, Shrubs, vines, and trees which are distributed chiefly in the American trOpics but range from southern Canada to Patagonia. The greatest diversity appears to be in South America. About 500-400 Species occur in Mexico and Central America; 50-75 are found in the United States and Canada; approximately 100 are located in the West Indies. Four species are native to EurOpe, one to Africa, and several to Asia. Eupatorium is economically important as the result of some undesirable attributes of several species. Livestock have been poisoned by Eupatorium rugosum (see Kingsbury, 1964) and humans have died from the ingestion of the milk of the sick cattle. Milksickness, as the disease is called in humans, was a major problem during the settlement of the mid-western states of the United States where it often reached epidemic proportions. The disease is said to have I I been the cause of death of Abraham Lincoln's mother. EUpatorium wrightii of the southwestern United States and northern Mexico is said to kill cattle quickly and without any visible symptoms. Research concerning the poisonous properties of some species of Eupatorium is summarized by Kingsbury (1964). Several species of Mexican origin have become distributed throughout the tropics where they have become pernicious weeds and are now a significant problem to tropical agriculture to the extent that eradication pro- grams have been initiated in several countries. No comprehensive taxonomic work has been done in Eupatorium Since De Candolle (1856) . He divided the genus into three series: Imbricata (with four sections), Subimbricata, and Eximbricata. The series were distinguished by the number of rows of phyllaries and their degree of imbrication. Bentham and Hooker (1876) eXpanded Eugatorium with the inclusion of twelve previously described genera. Hoffmann (1889) described the eight sections of the genus which are in current usage, again basing the sections largely on characteristics of the involucre. Several of the sections are, however, poorly defined and are in need of re— investigation. Six of the eight sections are represented in Mexico and Central America. The greatest amount of informa— tion on the tropical Species was published by B. L. Robinson who, between the years of 1895 and 1954, approximately doubled the number of described species of Eupatorium. Three factors confound investigation in EUQatorium: hybridization, apomixis, and morphological or phenotypic plasticity. Hybridization has been documented in species of Eupatorium from the southeastern United States (Fryar, 1964) and is currently being investigated by Dr. R. K. Godfrey. Hybridization probably also occurs in the tropical Species but these have not been adequately studied because the great number of relatively obscure species make hybrid- ization difficult to detect. Conversely, apomixis is known from the tropical species (Holmgren, 1919; Sparvoli, 1960) but is only suspected in some temperate Species from the southeastern United States (Grant, 1955). Morphological plasticity is reported by Baker (1965, 1967) in several self— compatible weedy species from Mexico and Central America. An even greater amount of phenotypic plasticity is displayed in some presumably apomictic Mexican Species but this is in need of further investigation. Any comprehensive work on the tropical species of EUpatorium awaits a thorough systematic treatment of the genus. It is imperative, however, that the systematist be aware of the biological phenomena which tend to confuse the species boundaries. This paper is meant to illustrate several reproductive processes which have evolved in Mexican and Central American Eupatoria. It is eXpected that a better understanding of the reproductive processes will permit a more accurate taxonomic treatment of the genus. The following observation by Ornduff (1969) seems eSpecially relevant to the taxonomy of Eupatorium. Many of the diverse floral characteristics used by taxonomists in assessing relationships among taxa represent adaptations to specific pollinators or pol- linating methods. Therefore, the diversity of reproductive methods that occurs within a phylad has a strong influence on the number of taxa that are generally recognized in the phylad. . . . It is sug- gested that taxonomists Should make an attempt to understand the reproductive methods of the plants with which they work, Since such an understanding will strengthen the foundation upon which taxonomic judg— ments are made. The thesis is in four parts, each of which contains its own introduction and literature review. The individual parts may ultimately contribute to a systematic treatment of the genus, which is beyond the sc0pe of the present study. CYTOLOGICAL AND EMBRYOLOGICAL OBSERVATIONS Introduction Cytological investigation in Eupatorium has been, for the most part, restricted to studies of apomictic Species and to chromosome counts. At present, chromosome numbers have been reported for over 100 Species of the genus. Several counts have not been reported in the literature but are indicated on voucher Specimens only. The meiotic number of p_= 4 has been reported in g, Sinclairii and g, micro— stemon. ,This number is considered by Baker (1967) to be derived from the more common condition of p_= 10. Crosses between plants with p_= 4 and p_= 10 produce highly fertile and vigorous offspring. The number g_= 17 is common in section Eximbricata (as is p_= 10) but the higher base number is apparently restricted to this section. Over 25 per cent of the Species thus far counted are polyploids or contain polyploid pOpulations. Most common are triploids (17 spp.) and tetraploids (10 Spp.) with a few others: hexaploids (4 Spp.), octaploids (1 Sp.) and four species at an undetermined ploidy level. Synapsis is never seen in the triploids and may or may not occur in the tetraploids. Information is not available concerning synapsis in the higher polyploids. Apomixis consists of two basic methods of asexual repro- duction. Vegetative reproduction, or asexual reproduction by vegetative prOpagules, occurs commonly in both sexual and asexual plants. Vegetative reproduction may be a natural process of the plant or it may be induced by various horti- cultural practices. Agamospermy, or asexual reproduction by seeds, was not discovered until the mid-nineteenth century (J. Smith, 1841), and since that time agamospermy has been demonstrated in a wide array of vascular and non—vascular plants. Agamospermy can be divided into two subcategories. In adventitious embryogeny the embryos arise in the ovule from nucellar or integumental outgrowths, omitting the gametophyte from the life cycle altogether. The other subcategory is that in which gametophytes develop. This subcategory, in turn, is divisible into diplOSpory, in which the gametophyte arises directly from a megaspore mother cell (EMC); and apOSpory, in which the gametOphyte develops from some cell other than the EMC. Three types of diplospory are recognized. Diplospory by means of a restitution nucleus is a proc— ess in which the chromosomes never pair. The first diviSion of meiosis is arrested at anaphase and a nuclear membrane enve10ps the entire mass of chromosomes. The restitution nucleus is characteristically elongated. At metaphase of the second division all the chromosomes are gathered on a common plate. After telophase II daughter nuclei contribute to the binucleate embryo sac. This type of diplospory is not known in Eupatorium. Diplospory by means of a pseudohomeotypic division is a process in which the EMC nucleus has the appearance of a pre-meiotic nucleus. The chromosomes remain unpaired at diakinesis and are scattered in the cell. The metaphase univalents collect at the equatorial plane, divide longi— tudially and separate, forming two nuclei each with the unreduced chromosome number. The resulting dyad may or may not undergo a somatic division before one of the cells be— comes dominant and forms an embryo sac at the expense of the others. This type of diplospory has been observed by Sparvoli to occur occasionally in Eupatorium riparium. Diplospory by means of a somatic division is a process in which the EMC never assumes the characteristic aspect of a pre—meiotic cell. Except for its great Size, the nucleus of the megaspore mother cell resembles nuclei of surrounding somatic cells. Division is typically mitotic, resembling that of any somatic cell. This type of diplospory is the rule in E, adenophorum and g, riparium. There is some doubt (cf. Beaman, 1957) whether the first two types of diplospory, noted above, are distinct from one another. No information concerning this question was obtained in the present study. Apomixis in Eupatorium was first reported by Holmgren (1916, 1919). He investigated 8 sexual species and g, adenophorum (reported as g, glandulosum), a triploid, obligate apomict (2g = 51). In E. adenophorum an embryo sac develops by diplospory with the division of the EMC resembl— ing a somatic division. The unreduced egg develops into the 22_embryo and the polar nuclei unite and develop into the 4x_cellular endosperm. A similar case was reported by Sparvoli (1958,1AMKD in another triploid, obligate apomict, g, riparium (2p_= 48). In addition to somatic diplospory, Sparvoli observed occasional diplOSpory by means of a pseudo- homeotypic division of the megaspore mother cell. Eupatorium muelleri Eupatorium muelleri Sch. Bip. ex Klatt (section Eximbricata) is a tetraploid with 68 univalents and is closely related to at least ten other wholly or partially apomictic Species. It is an herb of low stature with an erect or ascending stem one meter or less in length. The foliage is confined to the lower portion of the plant and the inflores— cence is a few—headed open panicle with heads of 80 or more white florets. No sexual pOpulationS of the species are known. Unlike the two apomicts studied by previous investi- gators, g, muelleri is not weedy. It has instead a rather limited habitat preference, being restricted to pine or oak- covered slopes from central Mexico to Honduras. It is usually found in semi-Shady areas. 9 Microsporogenesis Seeds were obtained from a herbarium specimen (McVaugh 21246, MICH) and were germinated on moist filter paper in petri dishes. The plants were grown to maturity, at first in a growth chamber, later in the garden, and then in the greenhouse. Flowering was delayed due to cool autumn weather but the plants flowered shortly after they were placed in a greenhouse. Flowering occurred nearly a year after germi- nation. Probably as a result of the delayed flowering the plants became atypically shrubby and grew unusually tall. Buds were fixed for 24 hours in Carnoy's solution of 6 parts 100% ethanol, 5 parts chloroform and 1 part glacial acetic acid. They were maintained at ca. 400 F in 70 percent ethanol until used. Anthers were crushed in aceto-carmine for chromosome analysis. MicrOSpore mother cells at diakinesis were observed to have 68 univalent chromosomes (Figure 1). This number was observed repeatedly in cells from several florets. A voucher specimen is filed in the Michigan State University Herbarium (MSC). After flowering, the plants died back and the areal shoots were cut off. Some of the plants died at this point while others sent Up new but weak Shoots from the base of the stem. These also eventually died before reaching flowering stage. Plants were also grown from seeds obtained from Grashoff .188 and 2§§_from Jalisco and Chiapas, respectively. The florets were prepared primarily for investigation of the 10 Figure 1. Eupatorium muelleri. MicrOSporocytes with 68 univalents (ca. x 9500). 11 ‘ 0 " ’ W I ~. ‘: ‘ t - R.‘ k.‘ :7? 1...“ Figure 12 megagametophyte but since the whole floret was usually sectioned, there was opportunity to examine microsporogenesis as well. The culture of these plants and the preparation of the heads is described in the section on megasporogenesis. When the megasporocyte is young and nearly indistinguish- able from the other cells in the ovule, the microsporocytes have already enlarged and distorted the tapetal layer. The microsporocytes have the appearance of large somatic cells. The first division of meiosis occurs soon afterward, and, although some cells appear normal at metaphase I. others con- tain lagging chromosomes. After the first division the two halves of the dyad often separate and become spherical. Many cells abort at this point, some form pollen walls, and the rest enter the second stage of meiosis. Nuclei at this time all look very unorganized with numerous dark-staining bodies which appear tangled. The second metaphase is more abnormal than the first. Often so many lagging chromosomes are present that there is no well—formed metaphase plate. Telophase, if it is reached before abortion, is characterized by unequal groups of chromosomes forming the daughter nuclei. Micro— nuclei occur occasionally. The four resulting nuclei (occasionally there appear to be but three) never form walls about them but begin to disintegrate almost immediately. The cells may abort at any stage of meiosis. Those cells which form pollen walls lose them later in a process of resorption. All the products of microsporogenesis begin to 15 disintegrate before the first division of the EMC nucleus. The only remnants of the process are bodies of variable diameter which are clear and oil-like when seen in fresh condition. They are often preserved in the anthers of dried specimens and can usually be regarded as an indication of apomixis. Megasporogenesis Material for the investigation of megasporogenesis was obtained in Mexico (Grashoff 188 and ggg). Seeds were germinated in petri dishes and the seedlings were removed to flower pots shortly after germination, which occurred in a period from 3 days to 5 weeks after sowing. The young plants were kept near a window at first and after a few weeks were transferred to a growth chamber. The plants appeared to grow slowly until the temperature was increased to 900 F by day and 700 F by night. After the temperature was raised, the plants grew rapidly; however, difficulty in maintaining a moisture supply was corrected by lowering the day tempera— ture to 800 F. Under these conditions and with a ten-hour day length the plants began flowering about 5 months after germination. Heads were killed and fixed in Navashin's solu- tion of 10 parts 1% chromic acid, 7 parts 10% acetic acid and 5 parts commercial formaldehyde. They were kept in this solution at ambient temperature until use. The heads were then washed in water, the florets removed, and the involucre 14 and receptable discarded. Sections were cut on a rotary microtome at a thickness of 10 microns. The sections were stained according to a haematoxylin-safranin staining pro— cedure adapted from Esau (1944) and were mounted in Canada balsam. The slides are in the possession of the author. Voucher specimens are filed in the Michigan State University Herbarium. DipIOSpory by means of somatic division was the only type observed in ovules of Eupatorium muelleri, and the process is essentially identical with that described by Holmgren (1919) and Sparvoli (1960). When the ovule is young, the megaspore is nearly indistinguishable from the surrounding tissue (Figure 2). At length the larger Size and the central position of the megaspore clearly distinguish it from the rest of the cells (Figure 5). The megaspore continues to enlarge at the expense of the nucellus. -At this point it can be considered a one-nucleate embryo sac. Associated with this increase in size is the vacuolization of the embryo sac. The uninucleate stage is one of long dura— tion, often lasting until Shortly before anthesis. The normal duration of the one-nucleate stage corresponds roughly to the time in which sexual species would undergo meiosis, form tetrads and, at length, develOp a one-nucleate embryo sac (see discussion of E, macrophyllum, below). Two divisions occur rapidly to give a four-nucleate embryo sac (Figures 4-7). A final division results in an eight-nucleate embryo sac 15 Figures 2-7. Eupatorium meulleri. Embryo sac develop- ment. 2 and 5, Megasporocyte stage; 4, Uninucleate stage resulting from the enlargement of the mega- Sporocyte. The nucellus is disintegrating. 5 and 6, Binucleate stage; 7, 4-nucleate stage (all ca. x 1500). 17 which, at first, has the nuclei arranged in groups of four at each end of the sac (Figure 8). One nucleus from the micropylar end and one from the chalazal end migrate toward the center of the embryo sac to become the polar nuclei. Unlike the sexual Species studied, in which the polar nuclei unite Shortly after migration to the center of the embryo sac, the polar nuclei in E, muelleri do not unite until after division of the diploid egg. At length the egg cell is dif- ferentiated, the antipodals begin to degenerate, and the synergids become less conspicuous. Division of the diploid egg begins before anthesis and a Sizeable embryo can be found in unopened florets (Figure 13). The embryo which is formed is unquestionably the result of apomixis. No viable pollen is ever produced, no pollen tubes can therefore reach the egg (none were ever seen), yet embryos are developed in pre- anthesis florets. The polar nuclei unite at a common meta- phase plate by the time the embryo consists of a few cells (Figure 11). The endosperm cells are very large and, for a time, divide in synchrony with the cells of the embryo. Cells of both endosperm and embryo can contain two nuclei each (Figure 12), cell wall formation apparently coming later. Abortion of the embryo sac is observed occasionally. This apparently can happen at any stage of develOpment but the highest frequency of abortion seems to be during the uninucleate stage. 18 Figures 8-13. Eupatorium muelleri. Embryo sac development (continued) and early embryogenesis. 8, Early 8- nucleate stage; 9, Later 8-nucleate stage showing 5 antpodal cells, 2 polar nuclei, egg (nucleolus not Shown), and 2 synergid nuclei; 10, Deliniation of egg cell, synergids somewhat obscured; 11, Embro sac con- taining 4-nucleate embryo, polar nucleus is dividing: 12, Young embryo and surrounding endOSperm; some cells are binucleate; 15, Embryo and endosperm shortly before anthesis (all ca. x 600). 'e lopmezt ing 3 not of egg ,c con- iding: 2 cells before Figures 8-15 20 Megagametogenesis in Eupatorium macrophyllum Eupatorium macrophyllum L. f. (section Subimbricata) is a moderately large herb occurring from Mexico to Brazil. It is generally found in areas of high rainfall. It appears to reproduce only sexually throughout its range, and has the chromosome number of §_= 10. Material was obtained from seeds of specimens collected in southwestern Costa Rica near San Vito de Java (Grashoff 105).- The plants were grown, fixed, sectioned, and prepared like those used for megagametophyte studies in E, muelleri. Only pre-anthesis florets were used. The development of the female gametOphyte in E, Egggg- phyllum is comparable to that of most other Compositae. The megasporocyte, when first distinguishable from surrounding cells, is seen in leptotene (Figure 15). As the stages of prOphase progress the megaSpore increases slightly in size and at diakinesis 10 chromosomes are visible (Figure 16). The first division of meiosis is rapidly followed by the second and dyads are rarely found (Figures 17 and 18). The result of meiosis is a linear tetrad, the chalazal cell of which generally (but, apparently, not exclusively) becomes the one‘uninucleate embryo sac (Figure 19). A series of three mitotic divisions forms the eight-nucleate embryo sac. The antipolal cells are separated by cell walls from the remainder of the gametOphyte. Two polar nuclei, one from each end of the embryo sac, migrate toward the center and 21 Figures 14—20. Eupatorium macrOphyllum. Megagametogenesis. 14, Megasporocyte and surrounding tissue before meiosnm 15, MegaSporocyte at leptotene; 16, Megasporocyte at diakinesis, g = 10; 17, Dyad; 18, Tetrad; 19, Uninucleam embryo sac, nucellus and 3 Of the megaSpores degenerat- ing; 20, Binucleate embryo sac, nucellus and 5 mega- spores degenerating (all ca. x 1400). 22 S ”U t C r - e :1 n“ a u e a 14-20 igures 25 Figures 21-24. Eupatorium macrOphyllum. Megagametogenesis (continued). 21, 4-Nucleate embryo sac; 22, Early 8-nucleate stage; 25, Later 8-nuc1eate stage Showing 5 antipodal nuclei, 2 polar nuclei, egg and 2 synergidS; 24, Mature embryo sac, antipodals are beginning to degenerate; polar nuclei have fused; egg cell has enlarged and obscured the 2 synergids (all ca. x 1500)- 24 1 e o w n t S 1 o y .1 . . 21-24 Figures 25 Figures 25—29. Eupatorium macrophyllum. Chalazal portion of embryo sacs Showing variation in the number of antipodal cells and nuclei (all ca. x 1500). 26 mm 27 25-29 Figures 27 fuse. The egg cell then enlarges and becomes vacuolized. In later stages of development the synergids are somewhat obscured by the egg cell and the antipodals begin to de- generate (Figure 24). The number of antipodal nuclei and cells is apparently loosely governed for they often pro— liferate. The usual case is to have two antipodal cells, one with two nuclei, the other with but a single nucleus. However, variations are frequent and range from two cells each with a single nucleus to three cells each with two or three nuclei (Figures 24-29). Shortly before fusion of the two polar nuclei, ten dark bodies (prochromosomes?) are visible. The resultant polar nucleus has a larger Size and often shows light portions in its nucleolus (Figure 24). A SURVEY OF THE CONTENTS OF ANTHERS OF MEXICAN AND CENTRAL AMERICAN EUPATORIA Introduction Proof of agamospermy usually has rested upon breeding exPeriments or intensive cytological study of the megagame— tophyte. Both methods of investigation are time-consuming and neither can be adapted to accommodate a large amount of material in a limited time. To support my hypothesis that apomixis is a major factor contributing to taxonomic confusion in Eupatorium, a method was utilized in which large numbers of species could be examined with reference to breeding behavior. Agamospermous plants often have abnormalities in micro- sporogenesis resulting in production of little or no viable pollen. No pollen is produced by the three apomictic species of Eupatorium thus far studied. Abortive pollen of irregular size and shape is produced by apomicts of Townsendia (Beaman, 1957) and Erigeron (Holmgren,1919). An examination of pollen was therefore conducted to provide an indication of the pos- sible presence of agamospermy in Mexican and Central American Species of Egpatorium. Pre-anthesis florets from Specimens 28 29 of the Field Museum (F), the University of Michigan Herbarium (MICH), and the Michigan State University Herbarium (MSC) were used in the survey. Materials and Methods Florets from 1,048 Specimens representing 192 Species were macerated in drOps of aceto-carmine to liberate the pollen. The preparations were examined directly with a com— pound microsc0pe at 150-420 x magnification. Normal pollen grains are uniform in shape and size and stain reddish in aceto-carmine. After several minutes, the nuclei stain dark red but they are often obscured by the highly sculptured pollen wall. Two kinds of abnormal pollen are observed in Eupatorium. In Specimens of section Cylindrocephala the grains are frequently of two sizes, the larger stained, the smaller usually clear. The other type of abnormality is found most frequently in Specimens of section Eximbricata. In these Specimens, the pollen is more or less uniform in size but the grains appear empty and collapsed or distorted. The extreme condition is one in which no grains are formed at all and, in a few Specimens, even the spherical residue bodies are missing, the anthers being empty. Many precautions must be taken when judging the condition of the pollen because other factors than meiotic abnormalities may cause pollen grains to look abnormal. First, the florets must be of the pr0per stage of develOpment, preferably 50 immediately before anthesis. Anthers which are too young have immature grains which do not stain well. Post-anthesis florets often have lost too many grains to permit an accu- rate analysis. Examination of immature or post—anthesis florets is therefore not useful. Second, while most Speci- mens are useable regardless of age, some, possibly due to slow drying, have grains which appear to have fermented. In one instance the contents of the anthers were found to be fungal spores. Third, some pollen abortion is normal even in sexual plants and the amount of abortion often increases with higher levels of ploidy. Fourth, interSpecific hybrids, if they exist, can be expected to Show a high degree of pollen sterility. Fifth, plant collectors often select unusual examples from a population resulting in perhaps a higher pr0portion of abnormal specimens in herbaria than actually exist at any one time in nature. Sixth, environ— mental factors such as drought, insect damage, disease, or herbicides can cause the production of abnormal pollen grains. Pollen quality was judged "bad" (Table I) when over 50 percent of the grains were unstained, when great irregularity in size and Shape occurred, or when few or no grains were produced. Pollen quality was judged "?" when about 50 to 50 percent of the grains were unstained. Pollen quality was judged "good" if the grains, though unstained, were regular in size and shape. This was frequently the case with Speci- mens which appeared to have fermented. 51 From one to 60 Specimens were examined for each species. Lack of adequate material often limited the number of specimens which could be observed. Generally, judgments concerning the probable presence of agamospermy in a Species were only made on those Species for which five or more speci— mens were observed. However, some species which are closely related to suspected apomicts and which showed pollen abnor- mality were also judged likely to be apomictic even if fewer than five Specimens were used. When the survey was initiated, two florets selected from each specimen were examined separately. As the study progressed it became evident, however, that the degree of pollen normality varied little, if at all, from floret to floret. It was felt that the uniformity of the anther con- tents and the nature of the survey did not require duplication of observations but that by limiting the number of observa— tions per Specimen to one, a greater number of Specimens could be examined. Often if the contents of the anthers ap- peared dubiously good or dubiously bad a second floret was prepared, but in the majority of cases, the preparation of the second floret appeared identical to the first. Table 1 summarizes the survey and the judgments made on each species. The Appendix lists the Specimens examined and the conclusions on each specimen. 52 UQSCHDEOO Ho O O N N E3H>um0m>uco .nw HO O O n n meHonuwom>u£o .mm | m D NH O N wH EDHGLQOOHHOLU .mm iemrmasnsmcsc HH AH u c m o H w , wmcwmero .,m m o o m m wmcmemmnu .wm HO o o m m HHcouHumu .mm I m o o A A agnnszmoHoumu .sm AnhmH .ucanuv ON u cm H0 o o m m ESHHONHHHHQMU .nn H0 o o m m wwcwcumemu .NW N o o m w ESHaxsmonu .H0 0 O O m m ESHHOwwmnuCHEmHmu .0w 0 o o N N meHowmo .mN m 0 m NH HN mwaH>muQ .mm 0 O o H H HHmwcmun .n u o o H H Escmmmfiocmun .mN o o o N N wamHn .mN m 0 H s m Enuumegn .wN O O O N N EDCMHmuHOQHHQ .MN I o h m m m m flfl>oHauh .mm Awme .mc4x Ucm nocuzev OH H c m o 0 MH mH ESUHCOuwQ .HN I U n H o w m HHOHO£uumn .nN RewcmflHnsmczv HH mwosw n c m D NH O o NH EDHHONHUHHHOQ .mH m 0 o A n “HuumHuumn .wH m o o m m Esmusum .IH ANme ..Hm um umcussv .mmmnw +ON u m m D em H NN ow EscchuOQcmnomm .wH m o o N N mwooucunm .mH m 0 O OH OH mumHomum .wH m 0 O m m EDHHONSMHMHM .MH 0 o O H H EsnosUHucm .NH AwwnH .mcHx Ucm umcusev NH so u m w o o NH NH EDEoncuochm .H AHmnH .mcsx Ucm umcusev HH 9N u m w o O NH NH muMHsmcm .0H m o o m m Sausownasmcm .m w 3 HH 0 0 HH EswumHLocm .m m 0 o n s ESCHHMUm>Em .m U NO H o O H EocflummEm .m m mo N o H m mndmso>sh5m .m m o H 6H NH stmanHm .w m 0 H H m EDEummmocmem .m imums ‘coumesomv em a mN m n m o m Ha asuocm0cmnm .N Ho 0 o N N Esflcmmnuncwbm .H A>Hco meummm pom m @000 HMDOB muuHEOQm mchuwocou ,NDHHmmU cmHHom mxnmeom m>flumusmv ucmEmpsb meome aflwmmmmmmm :ofiuoom ewcHdew mcmEHommm mo .02 Esfluobnmsw No meoon chHumE4 Hmuucmo Ucm cmOwaZ CM >uHHmso cwHHom No wumEESw .H mqmdh 55 nozzaucnu O O H N m ESCOH>L .c N No N O O N Enumuswcsmc .mm m 0 n m m mHmuHQmOL .mn m o C H H wmcmusococ .he 0 o O N N Eocm>m3H0£ .m« m o o m m mmcmosmoun .me m o C H H Escmm0>m£ .Hw O H O C H EntomOandHEms . n o o o H H ESHHONmmuwpr .Nw m o c m m Ez>quDwnwL .Hr m 0 O m m. @mcmcm>m£ .O« m o o r m wHHummn .mw m o o H H Escmecwms .mw m H o m m Eommflum .Im lhswms ..Hm so nwcussc 6H u m m o o mH mH Huwmmnw .mm iwmmu .uocuss ncm st3oav 6H n m o o o H H wH5m0uumnm .mw U o O H H HHNmeucom .Hw u o o H H agnnumusm .mw o O O N N EDUDmHm .Nw AmemH ..Hm um umcnzav NH u m m o H HN NN Enumunmam .Hm u o o H H EDEHuumanm .mm o o c H H duuowaamm .mm m o O OH OH EnumchHamm .mm AomcmHfinsmcsv HH Hmnom u d m o C m m Esaamaounuxum .rm 0 o o H H EsmumuoHnm .mm m o o w m Enumaw .mm o o o H H Humnmncwucm .Hm m o o w w HHucmusp .mm 0 O O N N ESHH>£QO>HU .Nm I m o c m w EsmUHODHwU .Hm Anan .ucmuov CN n :N m o N mH NH mmoHowHwU .Qm H0 o C N N ESHHAUOUOuu .mH u 0 H H N Esnummewuo .mH m o o w w eswEmunmmnu .rH m D N H m HH Humuasou .wH o o o N N HHuumNcoo .mH o o H H N EsnofiuflquHumcoo .H. m n H o m « Eszonamcou .nH o o o N N wmeoHHou .NH m D H H OH NH ESCHHHOU .HH A>Hco meUmmm omm m U000 Hence muoHEOQm mchuwocoo >DHHmSU cmHHom mxumemm m>Humusmv ucmEmUSh meowam ESHHOummsm :oHuomm pwcHmew mcmEHummm No .02 Smashucou I H mqmHmE mHmeOUmHmE EscmHuwuHmE EsuumE . EDHchmouumE .OHH ESHmLQmUOuomE .mo HHst .mo ESUHUSH .IO EscmocmNoH .m souHoCuocoH .m HHuwcmmoH .H ’73 x—I \ 'I I I .4 fl r—I ‘4 x—I “.‘I ‘» I.JI. pix”- roozomcrz ; “Lcmmcmzv m I m. OJ III (‘0 U ('1 U) N U ‘1‘“)ch U) (0 r» CD H—I~—{ ~.-—{‘-—I\-I\I\'I ‘ . Amend .anoHoov as n C o cram chm wrap qu NIHco>woN3mr~r~$.40JNuoH60ume .OmH HHcoEEwH .mm stmuHcmH .mm Enumoa>mmH .Im Humaasx .wm EscmemcH3umx .mm Escshwn .Hm coomHuonH .on ESHHOmmm>H .Nm mHawHOmH .Hm Eammuua .Om mmUHocmwnH .mm EDDQEOUCH .m EnumcumocH .nm menuHEH .wm Eschomm>£ .mm EDHHONHmomm>£ .Hm EnumHmEOQ>L .mm co>uUHUom>£ .Nm mHQmHOCmE>L .Hm IL] 9 UQJUrifi m U 00 Insmss ..Lc no umcusev cm I m U .NLDNHLDNEONd‘fi‘I‘” AtOLwHHQSQSSV HH mN u m flown” .ocmuov no u :N .ON u mN m pl U GUUNI‘CIUUIUUUNQUGJUU OOOOfiOOO0.00000000000000 OOOr-IONOOOOOOx—IOVI‘O OHfiOOOOOOOx-II‘OOOOLOONOOOO OOOOOOOOOvIOx-IHOHO ‘$C)HOcuUDN<4CDNzHOJ¢DHHmDU cwHHom mxuoEwm ®>Humbodv DCOEmesh mmHoQO ESHHOummsm :ofiuomm ewcHmem mcwEHummm mo .oz DmscHucou I H mqmde 5&5 ewscHucou :oHuoom UmcHmem mCMEHqum mo .02 ANwHH ..Hm um nmcussv OH H d m o o m m mmcmHmmmmn .me AUOLmHHQSchV HH on u m m o o H n wumHsvcmuUwsv .mmH AanH ..Hm um nocnsz ON.O+ H H Anme .mcHx Ucw Hszomv Om u d m o N ON NN EsHmnmmuocuxm .HmH immmH .ucmuov Hm." mm M o o H H mmnHoncamuocusd .an o o m m HHmsmnsm .mmH ADHQNH ..Hm um nmcusevHH mm H d U Q m o o m EDHHwnuHsm .HmH xaHemH ..Hm um umcnseo H on u m AeoLmHHhsmcsv HH AH I m m h HH H H mH EsHHommeHmcsum .omH I m o o H H HmecHnd .mHH U o o H H EsmochHquQ .NHH u o o H H esHHoanucmuuomHm .HHH w o o H H SSHermsumHm .mHH meH an I I H o o H EscHuoc .mHH ANme ..MW “w.wwc ::V «H I W m O O m m mHQmHOOchOLQ .HHH uses 4H n c m o o m w mumHoHuwm .mHH IaHsm ‘. m I u o o H H oumHsmchwa .NHH H H be umcuoev HH mN u c m Q NH 0 HH mN mmcwumzuwmm .HHH U o o H H EsuonHozmm .OHH m o o m m quumm .mmH U o o H H wmcmeEmmcmm .NMH m o o m m EQHOHNHHM>0 .HmH m D N O m m mmmwuuo .mmH U o o H H Eanmmuo .mMH m o o H w mmUHoHnmwuo .HmH m D m o o m mmHmcuHmuo .an m n m o o m EsHmcmmoomHHo .NMH leogmasnsacso...maum Icm I : .Nw I mm m o o m m SscmHowumumo HnH ANme .uocomcmz a Docoocmzv Hm u mN AmeH .Lwozov mm .mm .Hw n dN O Q m o m mH ESDMHOUO .omH U o o H H meDcprouo .mNH HO O O H H Escmensc .mNH s . AmeH .mcHx tam nmcusfiv HHNH u d N No H o o H Esansumn c HNH AnmeH ..Hm Hm Hospskv OH H d O O H N n ESmOuOEwc .mNH o o o H H HHcOmHmc .mmH O o H H N BauwuumeuHsfi .HNH InocusHco meuwmm vmm m 0000 HMDOB muoHEoma mchnmocou NJHHMSU :mHHOm mxumEom o>HumD3mv ucwEmbsb mmHommm EsHuOummdm fimDEHDCOU I H Ndmdk 5&3 “MIIIIIIJIIJI “wcHWWMHMM%me COHuuwmm "mHm: OUOHUCH . O a . ,I. ‘ I III _ cmEHOme uwzusom Rom nI Um: Nb n EN mo § 0 H O coHuommO new . I. I ucsou HON COEHO msoaummmofimmm NHanfirmmq uwxmm mwmv onnn>r 0D >mE SSH r OMQII .Uwucwmmummu meumsv Uwucmmmumwu >H®umsvw .mbmuHunEHflsm coHuuwmm HHOB UM; ucsou mEOmOEOHLU HON cwEHowmm umnuso>..I “cmHHo UwEnow HHm3 Um: .wmum+hN M N No ucsou HON em Doc meWMMMm CWHWMW 5mm: mHHmsmsco nuH3 mcmEHuwmm- “puma CH ummmH am one uoc meuwamo “mnwDEmE waOEuomwOEm u .H mm Hnmnouqupmucwmmuamu >Hmumsqum uoc meoQOHu HHOQ UOEHONI mam nw£030> IIIIIIIIIIIIIIIIIIIIII m mchHmbcoo >Hflmnoum mmHuwmmn ”Hmsxwm NHQMQOHQ meuwmmm AmeH .ucmuov Hm u mN MU wmcmHmEh .NmH HHurmHns .HmH o EDUHcmoHS> .OmH Ho ESHHOHHCH> .mmH mo mthmuH> .mmH meHocuan> .ImH mHmcum> .me EnumcHnusu .mmH HumEHwtxoumsu .HmH EDH>uwcHuu .mmH EDHkucmEOD .NmH SsnonHmuscu .HmH Escommuuwu .omH ESH>umchcwmnsm .mIH EnumwucHnsm .mrH EDthuooQDm .IHH EsuoHnum .wHH anHQHOm .mIH mmnHocmeeHHom .HHH ESHHONHSmetHHOm .WHH HHLousxm .NnH HHHHmHoch .HrH EscHuouom .OnH H meHoHcooououm .me HHNuHscom .mmH Humcwmmnuw .rmH mmcwHHuHmm .me EsumuuHmmm .mmH mHouHmsu .HwH stmuHuQSn .mmH HowsmHHnsmczv HH wH u d AmHmH .cmeEHomv ON n m N Anan .ucmnov ow ..IAU®LmHHDDQC3v NH HmeH .mcHK Ucm umcusfiv H Anan .ucmuov om n AmemH ..Hm um chusev HH u CI u ddIcHflN CI OOOOVJ‘ONOOWOHOOOOOOOOOOOOOOOOOO V‘OUDNOWCDx—INNLDU'JwHNNNNfi'v'bLOHIO'd‘NPOHV‘LO f6UmUQmDUUmeUMUUUUUMIUmeMVI AmmmH .ucMuoC Om+ mm AHmmH .uCMnOV Hm u CN m II Hquouzuou .NwH mHm>Hu .HmH ESHummHn .OmH EzmpHonEocn .mmH EsmpwEmHoouonu .mmH mmUHoHsucsmmn .MmH Aman .HHo>umdmC mH mm m fiOHOOO OHOOr—IONOOOOHHHOOOOOOOOOOOONOOO {ONE‘HH‘d‘ fi'x-‘ILDNLDLDWHNCDLOFI‘NNNNNfi‘d‘t‘LDv-IMd‘me-Iv‘m (00090.0 C>O<3v40~# InoawcarIO A>Hco mmHoQO Umm m @000 Hauoe muoHEOQm mchnmocoo >DHHMJU :wHHom mxumEmm m>HDMuDQC ucwemnsh mequm EDHHOummsm coHuowm owcHmem mcwEHqum mo .02 fiwanucoo I H mqm¢9 57 Remarks It must be emphasized that no taxonomic decision is implied either in Table 1 nor the Appendix. Every attempt was made to be taxonomically accurate but the inclusion of a species name in Table 1 does not indicate its acceptance as a taxonomic entity. Nor does the inclusion of a specimen in the Appendix imply that it has been properly identified. The majority of specimens are listed under the Species name to which they were assigned in the various herbaria. In a few cases a collection number can appear under more than one Species name, although misdeterminations were corrected inso- far as possible. The degree of inaccuracy in the determina— tions probably is not great enough to seriously alter the conclusions which have been drawn. 0f the 192 Species examined, 55 (18.2 per cent) were observed to have pollen abnormalities thought to be indica- tive of agamospermy. Twenty-four of the presumed apomicts are in section Eximbricata, 7 in section leindrocephala, and 4 in Section Subimbricata. Because of the limited number of Specimens examined, no attempt has been made at this time to relate apomixis to phytogeography or ecology. More Specimens covering the total range of each of the suspected apomictic Species need to be observed before conclusions concerning phytogeography and ecology can be made. 58 The survey of pollen quality in Eupatorium is not in— tended to replace the study of living material. Proof of reproductive behavior must still be obtained by intensive cytological investigation and breeding eXperiments. The survey would not be useful in detecting species which are pseudogamous because in pseudogamous Species, the pollen is functional. Nonetheless, the survey has provided an indica- tion of the approximate extent of apomixis in Mexican and Central American Species Eupatorium. Furthermore, during the course of the survey, several species were found to have unusually small and smooth pollen grains. This led to a discovery of possible anemophily (see below) which is a method of reproduction never before reported in the genus. The investigation of pollen quality in herbarium Specimens pro- vides a tool whereby the investigator can ascertain which Species need more intensive study. PROBLEMS IN THE TAXONOMIC TREATMENT OF APOMICTS IN EUPATORIUM Introduction Taxonomic confusion has often been the result of con— centrated systematic effort in a genus in which apomixis occurs. Studies in Crataegus, Hieracium, and Rubus, for example, have led to the description of an enormous number of microspecies. The micrOSpecieS usually represent units of variation which in sexual Species would be no greater than the degree of variation between individuals. In apomictic grouPs, however, there is replication of variants into sizable clones. Hundreds of apomictic biotypes have been given formal taxonomic recognition in these genera. The description of numerous clones in an apomictic complex is hardly the taxonomic solution required by the biological community. The necessity for dealing with an unwieldly number of taxa may tend to discourage further investigation in a genus. If taxonomic studies are to serve as the foundation for other kinds of investigation, the recognition of vast numbers of poorly defined microspecies may, in fact, defeat the purpose of the classification. 59 40 Furthermore, meaningful evolutionary relationships are often obscured when a genus is so subdivided. By giving formal recognition to clones, one risks grouping morphologically Similar organisms which have develOped independently. This has happened in Festuca vivipera (L6ve and L6ve, 1956). A great many opinions concerning the classification of apomicts have been published without any particular one gain— ing general wideSpread acceptance. Davis and Heywood (1965) provide a discussion of these theories. In Eupatorium there are Species which are known entirely from apomictic material and Species which contain both sexual and agamospermous individuals. Apomixis may be obligate, as in E, muelleri, or facultative. Some Species which are totally apomictic have relatively little variability. Species such as E, muelleri, E, bellid- folium, and E, anchisteum are distinct, morphologically well— defined apomictic Species which present no unusual taxonomic problems. Some totally apomictic species exhibit a high degree of variation. Eupatorium choricephalum is one of these.1 The plants appear to respond to environmental factors with a great deal of phenotypic plasticity. In such cases as this the monographer should determine what variation is the effect of environment and what results from genetic differences. 1The determinations are doubtful of the two specimens of E, choricephalum with "good" pollen noted in Table 1. 41 He would also have to judge the taxonomic significance of any genetic difference he detected. This type of informa- tion is not readily available from herbarium specimens. Species containing both sexual and apomictic individuals and species which are facultatively apomictic are frequently taxonomically difficult. Sexuality permits incorporation of new gene combinations through hybridization, mutation and segregation while apomixis permits new gene combinations to be replicated. Often the result of facultative apomixis is the formation of a highly polymorphic group of plants which is difficult to define. Furthermore, hybridization is be- lieved to be a factor contributing to the development of apomixis (Stebbins, 1950). Powers (1945) has illustrated how hybridization could lead to the initiation of apomixis. Certain species of Eupatorium (e.g., E, aschenbornianum, E, odoratum, E, pazcuarense, E, prunellaefolium) appear, on the basis of the foregoing survey, to contain both sexual and asexual plants. The range of variability is great in these Species and it is not unlikely that more than one species is present under any one of these names. This may certainly be the case in Eupatorium aschenbornianum. At least three partially sympatric, morphologically dissimilar groups are currently represented by herbarium specimens so named. One such group occurs in southern Mexico and southward through Central America. Originally described as E, donnell-smithii, it has since been treated (Robinson, 1928) as a synonym of 42 .E- aschenbornianum. Two other groups exist in Mexico, one on the east coast, centered in Veracruz, and the other covering many of the western states. These groups would deserve recognition as separate Species were it not for the apparent hybridization between them. The putative hybrids are more than ephemeral entities for they seem to reproduce by agamospermy. This results in one large complex which has its true nature obscured in part by the application of the single name Eupatorium aschenbornianum. In a situation such as that of E, aschenbornianum a workable classification can be achieved only through experi- mentation. However, given the use of cytology, chemo- taxonomy, and breeding experiments, the monographer Should be able to resolve these problems satisfactorily. Suggestions Concerning the Taxonomy of Apomicts When monographing a genus in which apomixis occurs, there are several suggestions which might be followed in order to achieve a more practical and accurate treatment of the apomicts: 1. One Should indicate which Species are known or suspected to be apomictic and to what extent. If pos— sible, phytogeographical data concerning the distribu— tion of the apomictic biotypes and the sexual biotypes (if any) and information concerning the kind of apomixis involved should be included. 45 2. One should try to relate apomictic grOUpS to their probable sexual precursors. In many instances the apomicts may be triploid (or otherwise polyploid) derivatives of a sexual Species (cf. Grant, 1955). In other instances the relationships may be more ob- scure due to alIOpolyploidy or disappearance of the immediate sexual precursors. 5. In such instances where hybridization and sub— sequent apomixis have obscured specific limits, it may be advantageous to use the concept of the species complex or aggregate Species. The Species complex may be regarded as a super-Specific taxonomic category which is used more for the sake of convenience than accuracy. In using a monograph which incorporates the Species complex one could first key a specimen to the appropriate Species complex. This may be as accurate a determination as the user requires and he would stop at this point. If a more accurate determination is required the user would then turn to the citation of the Species complex in the text where he would find a description of the complex and a key to the included taxa. It is possible that the key would fail to dif- ferentiate many of the intermediates but this is to be expected. 4. With apomixis many hybrid intermediates of two sexual taxa may be replicated until they comprise a 44 Sizeable portion of the flora. It may, therefore, be expedient to recognize formally some of the more conspicuous biotypes. However, care Should be taken not to obscure the relationships between the members of the complex. 5. Certain biological phenomena may need to be regarded differently if they occur in apomictic Species than if they occur in sexual Species. The naming of new species on the basis of different chromosome numbers is inadvisable in apomictic groups where polyploidy and aneuploidy are frequently encountered. Species are frequently described solely from herbarium specimens and dried Specimens usually do not permit an accurate appraisal of the degree of phenotypic plasticity in a Species. Consequently, especially when few Specimens are available, it is possible to describe two or more "Species" which are, in fact, each a different phenotypic eXpression of the same species. The greater the pheno- typic plasticity of a species, the greater is the chance that it could be divided into several taxa by the unsuspecting taxonomist. Clausen, Keck, and Heisey (1947) reported that the phenotypic plasticity of hybrids is frequently greater than that of either parent. The same is probably true for apomicts which contain genetic compliments of two or more species of differing habitat requirements. Thus the systematist Should 45 determine experimentally the extent of phenotypic plasticity in members of an apomictic complex. APPARENT WIND POLLINATION IN EUPATORIUM Introduction Three types of pollination have been observed in the Compositae; ornithophily (Fries, 1905), entomophily, and anemophily. EntomoPhilous species constitute the vast majority while anemophilous species have heretofore been known only in the Anthemideae and the subtribe Ambrosinae of the Heliantheae. The occurrence of wind pollination in EuEatorium, which apparently has not been recognized previously, illustrates the independent evolution of a pre— sumably adaptive trait in what is considered (Cronquist, 1955) to be a distantly related segment of the family. Pollination studies on two species of Eupatorium were undertaken by Cross (1897) who concluded, on the basis of floral morphology and fruit set, that the Species were insect pollinated although pollen not removed by insects was eventually blown away by the wind. Her studies Showed that while wind pollination was possible, it did not occur in the two species studied. 46 47 Description of Putative Anemophilous Species In the course of the survey of pollen characteristics in Eupatorium, several species were found with Short—spined pollen grains. Eleven Species of 192 examined thus far Show this characteristic. .Some of these Species obviously represent relatively unrelated segments of the genus, but three Species, E, solidaginifolium Gray, E, solidaginoides H.B.K., and E, monanthum Sch. Bip., have many features in common. Although some of these similarities are not ob- viously correlated with particular adaptive traits, several of their common characters appear to be specializations for wind pollination. Eupatorium solidaginifolium ranges from Arizona and western Texas to Michoacan and Colima in Mexico. It is a calciphile found at elevations from sea level to 1500 meters in tropical deciduous forests, usually in association with Quercus or Bursera. The Species has often been reported as locally abundant and with flowers ranging in color from white to green to reddish brown or cream. It blooms from October to February. Eupatorium solidaginoides is a widespread, semi-woody Species occurring in South and Central America and as far north in Mexico as San Luis Potosi. It is a calciphile and has a tendency to climb. In Mexico it occurs from 1000 to 2000 meters, blooming from December to February with white to green flowers. 48 Eupatorium monanthum is known from Sinaloa to Oaxaca, occurring at altitudes ranging from 150—1500 meters. It is reported as locally abundant in tropical deciduous forests often with Quercus or Bursera. The Species has been variously described as an arching or scandent Shrub or as a vine. The flowers are fragrant, greenish white or greenish yellow and the heads may have reddish involucres. It blooms from February to April. Robinson (1926) placed E, solidaginifolium and E, monan- thum in section Subimbricata and E, solidaginoides in section Eximbricata. His key, however, provides for identification of E, solidaginifolium also with section Eximbricata. A comprehensive reappraisal of Eupatorium will be necessary before the sectional boundaries are clarified, but for the present it appears that these species are more closely related than Robinson's treatment would suggest. In any case, they seem somewhat transitional between sections Subimbricata and Eximbricata as presently understood. Morphological Evidence In the trend toward anemophily in these species, .E- solidaginifolium appears to be least Specialized, E, solidaginoides intermediate and E, monanthum most Specialized. Faegri and van der Pijl (1966) and Whitehead (1969) have noted (1) ecological and phytogeographical conditions which favor anemophilous plants and (2) morphological modifications 49 of anemophilous plants. The Species of Eupatorium fit the patterns of ecological requirements and morphological adaptations typical of anemphilous Species as described by these authors. Furthermore, Payne (1965) has described a number of modifications of the inflorescence of Ambrosia which appear to be correlated with increasing degrees of Specialization for anemOphily. Some of these, notably an elongation of the flowering axis and placement of the heads in an outward or downward position are paralleled in Eupatorium. Eugatorium solidaginifolium has the largest heads of the three Species, with 10-15 florets, and thickest inflores— cence branches. Its florets are also largest, being 6 mm from the base of the achene to the tip of the corolla. The heads are aggregated into secondary glomerules such that the heads fan outward and the entire inflorescence is shorter and more dense than those of the other two Species. Eupatorium solidaginoides has heads with 9—11 florets which are 5 mm long. In this Species the heads are not aggregated but are borne on extremely Slender pedicels about 5 mm long. The inflorescence is about twice as long as that of E, solidagini— folium and the main branches of the inflorescence are well separated, giving the thyrse a more Open aSpect. Eupatorium monanthum has the least number of florets per head (1 or, occasionally, 2) and the heads are closely aggregated into Spherical glomerules such that they all radiate from a common 50 point of attachment. Its florets are about 4.5 mm long, but the inflorescence is the longest of the three species, with clusters of heads on the secondary branches as well as the secondary branches themselves well separated from one another. Separation is further increased by the occasional change from opposite inflorescence branches to an alternate arrangement. The anther appendages of these species deviate from the majority of EUpatoria. The more "typical" condition is illustrated by E, calonhyllum Robins (Figure 50, a), an ap— parently insect-pollinated species. Eupatorium solidagini- folium (Figure 50, b) has anther appendages about one—half the size of E, calophyllum and they are not as well differen- tiated from the rest of the anther. Eupatorium solidaginoides (Figure 50, c) has anther appendages which are truncated and have a thickened margin of stronger texture. These anther appendages turn outward, a condition, to my knowledge, not reported in other members of the Eupatorieae. They seem to function in keeping the uncommonly wide corolla from collaps- ing. In E, monanthum (Figure 50, d) the appendages have been lost altogether except for a small rim of cells. The Style branches of E, solidaggnoides (Figure 51, b) are clavate, different in shape than E, calophyllum (Figure 51, a) but not unlike those of many other Species of Egpatorium. The apical half of the style branches of E, ggii- daginifolium are thickened and with a smooth surface; Sweep- ing hairs are absent. The remainder of the style branch is 51 Figure 50. Apical anther appendages: 3, Eupatorium calOphyllum, Rzedowski 8445 (MSC); E, E. solidagini- folium, Pringle 942 (MSC); _c_, E. solidaginoides, Pringle 5956 (MSC); g, E. monanthum, Rzedowski 21947 (MSC). (All ca. x 100). Figure 51. Apical portion of style branches: _a_l, EupatOrium c_alOphyllum; Q, E. glidaginifolium; g, E. solidaginoi£§§7 Q, E. monanthu_m. Stippled area represents non-stigmatic portion. Voucher specimens as in Figure 1. (All ca. X 100) . 52 Figure 50 Figure 51 55 ribbon-like and covered with papillae. In E, solidaginoides (Figure 51, c) the receptive area extends further along the style branch to the extent that only the tip is nonstigmatic. This type of style has been found also in E, coelestinum (Cross, 99, EE£,). It appears that the proximal portion is more able to retain pollen grains than the sub-apical portion, judging from adhering pollen grains. In E, monanthum the entire length of the style branch appears to be stigmatic (Figure 51, d). The clavate nature of the styles is partially obscured by the increase in width of the greater portion of the style branch. In the dried condition, however, the edges of the branches tend to curl inward revealing the underlying clavate shape. Most pollen grains in the Compositae are equipped with a more or leSs elaborate development of the exine into spines or Spiny ridges. Spineless pollen has arisen (presumably independently) in several tribes or subtribes, i.e., Mutisieae, Cynareae, Anthemideae, and Ambrosinae. Although lack of spines is not necessarily correlated with wind pollination, the "reduction of Spinyness is entirely in keeping with the anemophilous habit" (Wodehouse, 1926). Pollen from all three presumably wind-pollinated Species as well as that from two presumably entomophilous species of Eupatorium were examined with a scanning electron microsc0pe. Pollen grains of ragweed (Ambrosia artemesiifolia E,) were also examined for comparison with a well—known anemOphilouS 54 species. When so examined, the pollen grains of Eupatorium solidaginifolium, E, solidaginoides, and E, monanthum are not unlike one another nor are the surfaces very different from that of the Ambrosia pollen. Eupatorium solidaginifolium and E, monanthum (Figure 52, a) have grains which measure 15H in diameter. Eupatorium solidaginoides and Ambrosia artemesiifolia (Figure 52, e) have grains 20u in diameter, and E, calophyllum (Figure 52, c) and E, calaminthaefolium have grains about 50u in diameter. The florets of Eupatorium monanthum, which are one-half the size of those of E, calo- phyllum, produce more pollen than the latter. The spines of E, monanthum (Figure 52, b) are about 0.5u high and 1.5u in basal diameter. Those of Ambrosia artemesiifolia (Figure 52, g) are 1.5u high and 2u in basal diameter. Those of E, caloggyllum (Figure 52, d) are 5.5u high and 4n in basal diameter and are often curved or hooked. The pollen of the presumably wind—pollinated Eupatoria is abundant, powdery, relatively smooth and small and thus it is Similar to the pollen typical of most anemOphilous plants (cf. Whitehead, 1969); however, the diameter of the pollen of E, monanthum and E, solidaginifolium (15u) is just slightly less than that of the "typical" anemophilous plant (20-40u). When the pollen of the Eupatorium species is dry it is elipsoidal in Shape (Figure 52, a) but in humid air the grains swell and eXpand in the region of the colpi to become Spherical. 55 Figure 52. Representative pollen grains photographed with a scanning electron microsc0pe. _a_, E, E. monanth_u_nlI Rzedowski 21947 (MSC) (presumed wind-pollinated specieS)I g. Q, E. calOphyllum, Rzedowski 8445 (MSC) (insect pollinated): g. L Ambrosia artemesiifolia, Beauvais 2L9 (MSC) (wind pollinated); a, e, ca. x 5050; b,d,f, ca. x 6060; c, ca. x 1200. 56 Ied wit IDIIIIIEII d spear sect Figure 52 57 Discussion The shape of the inflorescence, the loss of anther ap— pendages, the increase in stigmatic area of the style branches, and the size and sculpturing of the pollen grains are considered to be adaptations to wind pollination. The narrow, whip-like inflorescence would cause the pollen to be released in the wind. The aggregation of heads into clusters orients the florets such that the corollas face outward and the styles are well separated. In E, solidaginoides the filamentary pedicels would allow the heads to hang pendent in the breeze thus shedding the pollen and allowing the style branches to swing freely back and forth. The mechanism of pollen release in the Compositae was first recognized by Cassini and has since been termed the plunger mechanism. As a floret reaches anthesis the pollen is released introrsely and the elongating style pushes the pollen out with the sweeping hairs which are usually located at the apex of the style branches. Small (1915) was the first to elucidate the function the anther appendages played in the plunger mechanism. The apical appendages serve to guide the pollen out of the corolla while the basal appendages function in keeping it from falling to the bottom of the corolla. In the common ragweed (Bianchi et al., 1959), the anther appendages are well developed and the pistillodium has an expanded ring of hairs at the apex. In this Species, the pollen is eXpelled in clusters, falls to the foliage below 58 and is later separated and blown away in the wind. In the presumably anemOphilouS Eupatoria noted above, the reduction of the apical appendage would allow the gradual release of pollen grains. This condition would be advantageous for a wind—pollinated plant and, in fact, occurs in anemOphilous species of other families. Observations of post-anthesis florets of Eupatorium indicate that the pollen is released gradually in the presumably anemophilous Species. Species which are insect pollinated seldom have residual pollen in the corolla or anthers of post—anthesis florets and the few grains which remain are aborted. The three Species noted above, however, have a large number of grains remaining which are not aborted. Anther appendages are thought (Small, 22, gE£,) to represent a sterilization of Sporogenous tissue brought about by the economizing of pollen production which entomophily permits. The correlated trend in reduction of the anther appendages in the anemophilous species of Eupatorium is a feature which suggests that anemophily is a derived con- dition in the genus. The style branches of the presumed anemophilous species of Eupatorium are devoid of sweeping hairs but the clavate tips may function in this capacity. The increase in stigmatic surface area of the style branches is thought to represent another adaptation to anemophily. While the length of the style is not unusual in these Species, the effective pollen trapping area is considerably increased. 59 I have not had the Opportunity to observe these Species of EUQatorium in their native habitats and thus the evidence for anemOphily is indirect. Unfortunately, data on their breeding systems is also unavailable. The possibility of occasional insect visitation cannot be ruled out. The reten- tion of apparently well-developed nectaries might suggest occasional insect pollination, but the nectaries are also well-developed in some obligately apomictic species of Eupatorium. Species in several families are known to be entomophilous at one time in their life—cycle and anemOphilous at another (cf. Knuth, 1906-09, Vol. I: p. 71). It would appear that the adoption of the anemOphilous habit is gradual, and, during the incipient stages, both anemophily and entomo- phily are operative to some extent. Thus far only three Species of Eupatorium have been studied in detail in relation to possible anemophily. Five other Species may also have this method of reproduction. These are E, eriocarpum, E, hebebotryum, E, incomptum, E, morifolium and E, qgadrangulare. With the exception of E, morifolium all the Species have elongated inflorescences. All have a copious supply of relatively short-spined, powdery pollen. Eupatorium incomptum is in section Hebeclinum; the other species are in section Subimbricata. Chromosome counts ofug = 10 are reported for three of the species (see Table 1). 60 Great importance has been placed on the anther appendage by B. L. Robinson. Species with vestigial appendages were often considered by him to be generically distinct from those with well-developed appendages. The eight presumably anemophilous species of Mexican and Central American Eupatorium Show a gradation from fairly well-developed anther appendages to nearly no appendages. Most of these species are closely related and one must therefore question the soundness of any taxonomic decision based on the nature of the anther appendages alone. GENERAL DISCUSSION The first hypothesis relating hybridization to apomixis was postulated by Ernst (1918). He observed that meiosis is disturbed in apomicts much the same way as it is in some hybrids; polymorphism and accompanying diffusion of species limits are Similar in groups with predominant hybridization as well as in groups with predominant apomixis. Ernst's theory has now become accepted as fact although experiments attempting to produce apomixis by means of hybridization have failed. Hybridization causes new gene combinations which, through apomixis, can be perpetuated, resulting in large "populations” of individuals each of which contains the hybrid vigor of the original hybrid individual. These perpetual hybrids are frequently capable of a greater degree of tolerance to ecological conditions than the sexual par- ental Species (cf. Stebbins, 1950, p. 595). Variability can exist in agamic complexes either as a result of variability which was originally incorporated into the complex at its inception or as a result of genetic changes after the establishment of apomixis. Agamic com- plexes often contain as many biotypes and as much variation as is fOund in outcrossing Species (Clausen, 1960, Valentine, 61 62 1960). This is true for both facultative and obligate complexes because obligate apomicts were probably derived from facultative apomicts rather than directly from sexual progenitors (Babcock and Stebbins, 1958, Stebbins, 1941). Aneuploidy (the loss or gain of one or more chromosomes or parts of chromosomes) can produce new biotypes (SOrensen, 1958). Aneuploidy often causes severe weakness and/or sterility in sexual plants; however, apomicts are nearly always polyploid (Gustafsson, 1946), thus the overall de— leterious effect of aneuploidy is lessened, and they are capable of overcoming the sterility barrier through agamic reproduction. Mutation and autosegregation occur occasional- ly (Turesson, 1956). New biotypes may be formed by spon- taneous production of polyploid offspring (Kappert, 1956). The production of haploid offsPring can be induced in some highly polyploid apomicts (Gustafsson, 1947) but it is not known to occur naturally. Three conditions have been suggested as contributing to the success of apomicts in arctic and alpine regions (Davis and Heywood, 1965, p. 574). 1. Apomixis facilitates the duplication of favor- able gene combinations and thus allows for the rapid Spread of asexual plants with favorable genotypes into recently disturbed habitats. 2. Apomixis sometimes allows for more rapid maturation of seeds (embryos often form before anthesis) 65 and this permits the occupation of areas having short growing seasons. 5. Apomixis is one adaptation to a shortage of pollinating insects. None of these three conditions are restricted to high altitudes or high latitudes. It should, therefore, not be surprising to find apomicts in other parts of the world where one or more of these conditions exist. In Mexico and Central America apparently all three conditions occur. Man has altered vast eXpanses of the natural countryside making available newly disturbed habitats suitable for colonization of plants with the prOper genotype. The dry season effectively terminates the lives of many plants or at least limits their growing season. A Shortage of insect pollinators, while not proven empirically, is sug- gested by the several trends toward life cycles independent of insect pollinators (see below). In Mexico and Central America, species in two sections of Eupatorium have a high frequency of pollen abnormality. The contents of anthers of Species in section Cylindrocephala usually have a different appearance than the anther contents of species in section Eximbricata. In section Cylindro- cephala pollen grains are highly variable in Size, ranging from about twice the size of normal grains to about one—third the Size of normal grains. The larger grains often stain with aceto-carmine but the smaller ones nearly always appear 64 vacant. In section Eximbricata the abnormal pollen grains are not as variable in Size as those of section Cylindrogephala and they Seldom Stain. Microsporogenesis is frequently disrUpted resulting in the complete failure of pollen production. On the basis of the appearance of the pollen in these two sections, it appears that apomixis may have arisen inde- pendently in each section. Cytological investigation of megasporogenesis in species of section gylindrocephala are needed to support this view. Eight triploids (2§_= 50) in section Subimbricata were reported by Grant (1955). These species from the South- eastern United States may be the result of yet a third inde- pendent derivation of apomixis in Eupatorium. Three Species of Mexican Subimbricata occasionally have abnormal pollen: E, collinum, E, mendezii, and E, ortegae. These shrubby Species are very closely related to one another but they do not appear closely related to the triploids (all herbaceous) reported by Grant. No abnormal pollen was found in Species of sections Chromoleana, Conoclinum and Hebeclinum. Eupatorium odoratum (section Cylindrocephala), on the basis of the foregoing survey, has both sexual and apomictic plants. Four chromosome numbers have been reported for the Species (see Table 1) and each count is different. Voucher specimens of two of the counts were examined in this study. 65 The count oflg = 20 + fragment (unpublished) was made in a collection (King E Soderstrom 4814) in which pollen appeared normal. The count of 22.: 62 (unpublished) was made in a collection (Breedlove 8857) in which the pollen was irregu- lar in shape and size and frequently aborted. Of all the Species studied, E, odoratum is perhaps the most widespread and the most variable in both vegetative and floral struc- tures. The extreme morphological diversity and the various chromosome numbers suggest that Eupatorium odoratum may be facultatively apomictic (cf. Davis and Haywood, 1965, p. 565). No experimental studies have yet been done to test this hypothesis. Inbreeding has become established in species which are related to Eupatorium pycnocephalum (section Subimbricata) through the development of self-compatibility (Baker, 1965, 1967). In the Species studied, self-compatibility is corre- lated with a reduction of chromosome number and a capacity for weediness. The only known completely self-compatible species is E, microstemon which produces 20-40 pollen grains per anther lobe. This is a reduction from the 160-200 grains per lobe produced by the self-incompatible E, pycnocephalum. During the study of pollen quality I noted other related, minute-flowered Species (E, jejunum, E, macrum, and E, minarum) which also had an unusually low number of pollen grains. No quantitative data on the number of pollen grains per floret were obtained at that time, however. It is possible that these species are also self—compatible. 66 Because apomixis, anemOphily, and autogamy all occur in Species of Eupatorium from the Same geographical area, it appears that these reproductive adaptations may be responses to the same stimulus. One such stimulus might possibly be a decrease in the number of insect pollinators resulting from widespread post-Pleistocene drying in Mexico and Central America. Another possible, although currently uninvestigated, adaptation paralleling the three noted above may be non— selectivity of pollinators. EntomOphilous species must at- tract sufficient numbers of insect pollinators in order to maintain the plant pOpulation. Such Species may respond to a shortage of efficient pollinators by producing substances capable of attracting a greater number and variety of insects. Although many of the insects thus attracted may be inef- ficient pollinators, the greater numbers of attracted insects may offset any disadvantage caused by their inefficiency. Many Species of Eupatorium have brightly colored flowers and very odoriferous glandular secretions in both floral and vegetative structures. Insects of many kinds are attracted to the plants readily. I have observed plants of Eupatorium odoratum surrounded by myriad insects including beetles, waSps, bees, butterflies, and moths. Knuth (1907-9, Vol. 2, pp. 572-5) lists insects of the following orders which were observed to frequent Eupatorium species: Diptera, HymenOptera, LepidoPtera, NeurOptera, Coleoptera. Breeland and Pickard 67 (1961) reported mosquitoes feeding on Eupatorium, but these insects most likely play little, if any, role in pollination. BIBLIOGRAPHY BIBLIOGRAPHY Babcock, E. B. and G. L. Stebbins. 1958. The American species of Crepis: their relationships and distribu- tion as affected by polyploidy and apomixis. Publ. Carnegie Inst. Wash. No. 504. 200 pp. Baker, H. G. 1965. Characteristics and modes of origin of weeds. l2 Baker and Stebbins, ed. The genetics of colonizing Species. Academic Press, New York. Pp. 147-172. . 1967. The evolution of weedy taxa in the Eupatorium mgcrostemon species aggregate. Taxon 16: 295-500. Beaman, J. H. 1957. The systematics and evolution of Townsendia.’ Contr. Gray Herb.‘185: 1-151. Bentham, G. and J. D. Hooker. 1876. Genera Plantarum. Vol. 2, pp. 245-246. London. Bianchi, D. E., D. J. Schwemmin, and W. H. Wagner. 1959. Pollen release in the common ragweed (Ambrosia aremesiifolia). Bot. Gaz. (Crawfordsville) 120: 255- 245. Breeland, S. G. and E. Pickard. 1961. Observations on mosquito feeding activity on the flower heads of Eunatorium and Solidago (Compositae). Mosqu1to News 21: 52-54. Clausen, J. 1960. A Simple method for the sampling of natural pOpulations. Annual Rep. Scott. Pl. Breed. Sta. 1960: 69-75. Clausen, J., D. D. Keck, and W. M. Heisey. 1947. Heredity of geographically and ecologically isolated races. Amer. Naturalist. 81: 114-155. Coleman, J. R. 1968. Chromosome numbers in some Brazilian Compositae. Rhodora 70: 228-240. 68 69 1955. Phylogeny and taxonomy of the Compo- Cronquist, A. Amer. Midl. Naturalist 55: 478-511. sitae.' Cross, L. B. 1897. On the structure and pollenation of the flowers of Eupatorium ageratoides and Eupatorium coelestinum. Contr. Bot. Lab. Morris Arbor., Univ. Pennsylvania 1: 260-269. Davis, P. H. and V. H. Heywood. 1965. Principles of angiosperm taxonomy. Van Nostrand, Princeton, N. J. 558 pp. Prodromus De Candolle, A. P. and A. de Candolle. 1856. Vol. 5, pp. systematis naturalis regni vegetabilis. 141-186. The morphology of Artemesia tridentata. Diettert, R. A. 1958. Nutt. Lloydia 1: 5-74. Bastardierung als Ursache der Apogamie im Ernst, A. 1918. Jena, G. Fischer. 665 pp. Pflanzenreich. Anatomical and cytological studies on beet Esau, K. 1944. i 69: 95-117. mosaic. J. Agric. Res. 1966. The principles of K. and L. van der Pijl. New York. 248 pp. Faegri, Pergamon Press, pollination ecology. 1905. Beitrége zur Kenntniss der Ornithophilie Fries, R. E. Ark. Bot. 1: 589-440. in der sfidamericanischen Flora. Fryar, W. R. .1964. Natural hybridization between two perennial Species of Egpatorium (Compositae). Unpub- lished Master's Thesis. Florida State University. Ghosh, R. B. 1961. Chromosome numbers of some flowering plants. Curr. Sci. 50: 75. W. F. 1955. A cytotaxonomic study in the genus Grant, Amer. J. Bot. 40: 729-742. EUQatorium. . 1964. IOBP Chromosome numbers, Report II. Taxon 15: 201-209. Apomixis in Higher Plants I. The Acta Univ. Lund. N. F. Adv. 2, Gustafsson, A. 1946. mechanics of apomixis. 42 (5): 1-66. . 1947. Apomixis in higher plants III. Biotype 2. and Species fOrmation. -Acta Univ. Lund. N. F. Adv. 45 (12) 185-570. 7O Hoffmann, O. 1889. Compositae. Eg_Engler and Prantl, ed. Die naturlichen Pflanzenfamilien. Vol. 4. Pt. 5, pp. 87-587. Leipzig. Kappert, H. 1956. Experimentelle untersuchungen fiber die Variabilitét eineS Totalapomicten (Taraxacum officinale Web.). Ber. Deutsch. Bot. Ges. 67: 325-555. Holmgren, O. 1916. .Apogamie in der Gattung Eupatorium (Vorlaugige Mitteilung.).SvenSk Bot. Tidskr. 10. . 1919. Zytologische Studien fiber die Fortpflanzung bie den Gattungen.Erigeron und Eupatorium. Kongl. Svenska Vetenskapskad. Handl. 59: 1—72. Kingsbury, J. M. 1964. Poisonous plants of the United States and Canada. Prentice-Hall, Englewood, N. J. 626 pp. Knuth, P. 1906—09. Handbook of flower pollination. English translation. Clarindon Press, Oxford, Vol. 1, 1906, Vol. 2, 1908, Vol. 5, 1909. LOve, A. and D. Ldve. 1956. Cytotaxonomic conSpectus of the Icelandic flora. Acta Horti. Gothob. 20: 65-290. Mangenot, S. and G. Mangenot. 1962. Enqugte sur les nombres chomosomiques danS une collection d'especies trOpicales. Rev. Cytol. Biol. Veg. 25: 411—447. Ornduff, R. 1969. Reproductive biology in relation to systematics. Taxon 18: 121-155. Payne, W. W. 1965. The morphology and inflorescence of ragweeds (Ambrosia-—Franseria: Compositae). Amer. J. Bot. 50: 872-880. Powell, A. M. and R. M. King. 1969. Chromosome numbers in the Compositae: Columbian species. ~Amer. J. Bot. 56: 116—121. Powell, A. M. and B. L. Turner. 1965. Chromosome numbers in the Compositae. VII. Additional Species from the southwestern United States and Mexico. Madrofio 17: 128-140. Powers, L. 1945. Fertilization without reduction in guayule (Earthenium argentatum Gray) and a hypothesis as to the evolution of apomixis and polyploidy. Genetics 50: 525- 346 C 71 Robinson, B. L. 1926. EUpatorium. ;§_Standley. Trees and Shrubs of Mexico. Contr. U. S. Natl. Herb. 25 (5): 1452-1469. Small, J. 1915. The pollen-presentation mechanism in the Compositae. Ann. Bot. (London) 29: 457—470. Smith, J. 1841. Notice of a plant which produces seeds without any apparent action of pollen. Trans. Linn. Soc. London 18. SOrensen, T. 1958. Sexual chromosome-aberrants in triploid apomictic Taraxaca. Bot. Tidsskr. 54: 1—22. Sparvoli, E. 1958. Osservazioni cito-embriologiche in Eugetorium riparium Reg.--Nota I: Cariologia. Ann Bot. (Rome) 26: 120-127. . 1960. Osservazioni cito-embriologiche in Eupatorium riparium Reg.--Nota II. MegaSporogenesi e sviluppo del gametofito femminile. Ann. Bot. (Rome) 26: 481-504. Stebbins, G. L. 1941. Apomixis in angiOSperms. Bot. Rev. (Lancaster) 7: 507-542. . 1950. Variation and Evolution in Plants. Columbia Univ. Press, New York. 645 pp. Turesson, G. 1956. Variation in the apomictic microspecies of Alchemilla vulggris L. II. Bot. Not. 109: 400—404. Turner, B. L., J. H. Beaman and H. F. L. Rock. 1961a. Chromosome numbers in the Compositae V. Mexican and Guatemalan species. Rhodora 65: 121—129. Turner, B. L., W. L. Ellison and R. M. King. 1961b. Chromosome numbers in the Compositae. IV. North American species, with phyletic interpretations. Amer. J. Bot. 48: 216—225. Turner, B. L., M, Powell and R. M. King. 1962. Chromosome numbers in the Compositae VI. Additional Mexican and Guatemalan Species. Rhodora 64: 251-271. Turner, B. L. and R. M. King. 1964. Chromosome numbers in the Compositae. VIII. Mexican and Central American Species. Southw, Naturalist 9: 27-59. 72 Valentine, D. H. 1960. The treatment of apomictic groups in Flora Eur0paea. Feddes Repert. 65: 114-127. Whitehead, D. R. 1969. Wind pollination in the Angiosperms: Evolutionary and environmental considerations. Evolution 25: 28—55. Wodehouse, R. P. 1926. Pollen grain morphology in the classification of the Anthemideae. Bull. Torrey Bot. Club 55: 479-485. APPENDIX APPENDIX SPECIMENS USED AND POLLEN QUALITY OBSERVED IN THE INVESTIGATION OF POLLEN QUALITY OF MEXICAN AND CENTRAL AMERICAN SPECIES OF EUPATORIUM Listed below are the specimens which were used in the investigation of pollen quality with the evaluation of the quality of pollen observed in each Specimen. See text for an explanation of the terms employed. Eupatorium Species, , collector and number F MICH. MSC Good ? Bad adenachaenium Hernandez s.n. x x Rzedowski 19552 x x adenophorum Bourgeau 172 x x de Harre s.n. x x McVaugh 22684 x Palmer 1510-1891 x x Paray 1115bis x Pringle 2495 Pringle 6856 de Puga 171 Quijano s.n. Rzedowski 19552 XXXXXX X adenospermum Feddema 2444 X X de Harre s-n. x Hinton 5110 . X Langlassé 882 X McVaugh 14165 X XXXX 75 74 Eupatorium species, collector and number F MICH MSC Good ?, Bad albicaule Aguilar 17 Gentle 288 (2 spec.) Gentle 575 Gentle 809 Kenoyer A151 King 5057 Lundell 2791 Lundell 4011 Lundell 4016 .Matuda 1466 Matuda 5529 Palmer 256-1902 Pringle 5105 Steere 1056 Steere 1665 Steere 1979 Yuncker 4921 amblyolepis Pringle 8054 Pringle 9900 Pringle 9900 amestinum Tate s.n. amygdalinum Allen 1278 Dodge et al. 16841 Kellerman 7605 Skutch 2460 Standley 55856 Standley 56049 Williams & Merrill 17150 anchisteum Johnston 556 Molina 18554 Molina 21084 Morales 1222 Rodriguez 1458 Standley 577 Standley 27520 Standley 76558 Standley 77248 XXX XXX XXXXXXXXX XXXXXXXXXXX XXXX >¢X5< X>< >¢X>< X>< 94 Eupatorium Species, collector and number F MICH MSC Good ? Bad pansamalense Tfirckheim II 2145 x x parryi Pringle 155 Pringle 1595 Wynd & Mueller 254 XXX XX XXX pauciflorum Bourgeau 1100 x x pazcuarense DeJong 1794 x x Gentry 1945 x x Gunzman GH1158 Hinton et al. 15657 x Knoblock 5449 x Krauss 1067 (2 spec.) Krauss 1071 McGreggor et al. 264 McVaugh 14024 McVaugh 20600 McVaugh 20611 McVaugh 22292 McVaugh 25071 McVaugh 25428 Mendellin 287 x x Pringle 1265 x Pringle 6562 X X Pringle 8027 Pringle 8028 Pringle 11525 x x Rzedowski 8164 Rzedowski 9725 (2 spec.) Rzedowski 18799 Rzedowski 21410 Rzedowski 21426 Rzedowski 25069 XXXX X XXX XXXX XXXXXX XXXXX X XX XX XXXXXX X peninsulare Pringle 155 x X petiolare Gilly & Simpson 16A Gonzalez 421 Gonzalez 2040 Gonzalez 2065 XXXX XXXX 95 EUpatorium species, collector and number F MICH MSC Good ? Bad petiolare (cont.) Gonzalez 2087 Gonzalez 2129 Gonzalez 2159 Schnee s.n. phoenicolepis Breedlove 7515 Breedlove 7918 Breedlove 8017 Breedlove 14024 Ton 1496 photinum Pringle 8029 platyphyllum .Skutch 2561 plectranthifolium Rodriguez 506 porriginosum Pringle 6552 pringlei Breedlove 7987 Breedlove 9214 Pringle 6118 Ton 670 prunellaefolium Beaman 2458 Beaman 2526 Beaman 2799 Beaman 2860 Beaman 2905 Beaman 2952 Beaman 4049 Beaman 4114 Beaman 4290 Pringle 4286 Rzedowski 20492 Rzedowski 20517 Rzedowski 25876 XXX XXXX XXXXXXXXXXXXX XXXX XXXXX X XXXX XXXXXXX XXX X 96 Eupatorium species, collector and number F MICH MSC Good Bad pulchellum Bourgeau 948 Detling 8724 x Gilly 105 Pringle 2579 Rzedowski 25154 purpusii Feddema 2895A Gentry 7169 McVaugh 9954 McVaugh 22790 Rzedowski 18118 XXXXX pycnocephaloides Standley 66528 X pycnocephalum Bourgeau 1098 Bourgeau 1252 DeJong 1454 DeJong 1465 Gonzalez 5192 Grashoff 78 Grashoff 86 McGreggor 16571 Montgomery & Root 8617c Palmer 855-1894/5 Palmer 245-1902 Palmer 500-1902 Pringle 8077 Romero 15 Rzedowski 5557 Rzedowski 8501 Rzedowski 9726 Rzedowski 10091 Rzedowski 19205 Rzedowski 19579 Rzedowski 21472 Rzedowski 25297 quadrangulare Baker 2502 King 2448 King 5855 King 5964 McVaugh 22440 XXXX XXX XXXXXXXXXXXXXXXXXXXXXX X XXXXX XXXXXXXX XXXXXXX X XXXXX XXXXX XXXXX 97 Eupatorium species, collector and number F MICH MSC Good Bad quadrangulare Pringle 2669 wright 1555 rafaelense Conzatti 5247 Laughlin 501 Pringle 8051 Pringle 8040 Pringle 8056 rapunculoides McVaugh 15280 McVaugh 17450 McVaugh 18180 Palmer 545-1886 Pringle 2512 Pringle 11517 rhodochlamedeum Pringle 5570 Rzedowski 8520 rhomboideum McVaugh 15598 Pringle 6561 Pringle 15795 Rzedowski 19557 Rzedowski 19557 Rzedowski 21658 riparium Bourgeau 1928 rivale McVaugh 9986 rothrockii Pennell 19124 Pringle 1265 Townsend & Barber 525 White 4407 rubricaule Hinton et al. 8772 Pringle 2878 Pringle 4272 XX XXX XXXXX XXXX XXXXX X XX X XXX XXXX X X X X XXX XXXX 98 Eupatorium species, collector and number F MICH MSC Good Bad rUpicola Pringle 4970 sagittatum Carter et al. 5455 Carter & Chisaki 5465 Gentry 4712 Gentry 4848 Wiggins & Rollins 262 saltilense Muller 2802 Rozynski 195 schaffneri McVaugh 15571 McVaugh 20546 Pringle 15052 Pringle 15997 Pringle 15997 schultzii Bourgeau 1925 Breedlove 8822 Matuda 0752 Matuda 0774 Matuda 0799 scorodonioides DeJong 1518 Diaz 5 Gonzalez 1591 Gonzalez 2944 Gonzalez 5078 Gonzalez 5115 Gonzalez 5128 Lagunas & Castillo s.n. Pringle 8244 Rzedowski 20044 Shreve 9598 serotinum Palmer 424—1880 Sinclairii King 5545 Tucker 459 XXXX XXXXX XXXX XX XXXXXXXXX XXXXX XX XXXXX X X XX XXXXX X XXXX 99 Eupatorium Species, collector and number F MICH MSC Good ? Bad skutchii Breedlove 6725 Breedlove 9267 Breedlove 9508 Breedlove 11597 Rzedowski 16571 Standley 66552 Ton 719 Ton 758 solidaginifolium McVaugh & Koelz 1675 Pringle 942 Shreve 6659 White 4458 Wilkinson s.n. solidaginoides Bartlett 11462 Pringle 5956 Purpus 14255 Rozynski 641 Rzedowski & McVaugh 94 Standley 60112 Standley 60852 sordidum Breedlove 9408 Dodds & Simpson 17 Matuda 1157 Pringle 8048 Rosas 187 Rzedowski 11089 strictum Gentry 644M McVaugh 21705 subcordatum Rodriguez 516 Stork 2588 subintegrum Pringle 5511 Rzedowski 17760 X XXXXXXXX XXX XXXXXX X XX X XXXXX XXXXX XXXXX XX XXXX 100 Eupatorium species, collector and number F MICH MSC Good ? Bad subpenninervium Matuda 0854 Pringle 6275 tetragonum Rzedowski 25290 thyrsiflorum Cronquist 9824 McVaugh 21771 Ortega 4444 Pringle 11521 tomentellum Breedlove 7921 Breedlove 14019 Cronquist 9695 Pringle 4959 trinervium McVaugh 21229 McVaugh & Koelz 784 Palmer 565-1886 Pringle 1789 Pringle 11527 tuerckheimii Breedlove 881 Breedlove & Raven 8242 Carlson 2464 Raven & Breedlove 19794 Tfirckheim 77 turbinatum Rzedowski 5580 vernale Breedlove 7998 McClintock s.n. Skutch 295 viburnoides Gonzalez 5286 Pringle 10566 Pringle 15597 XXXXX XXXXX XXX XXX X X>